mktables

#!/usr/bin/perl -w

# !!!!!!!!!!!!!!       IF YOU MODIFY THIS FILE       !!!!!!!!!!!!!!!!!!!!!!!!!
# Any files created or read by this program should be listed in 'mktables.lst'
# Use -makelist to regenerate it.

# There was an attempt when this was first rewritten to make it 5.8
# compatible, but that has now been abandoned, and newer constructs are used
# as convenient.

# NOTE: this script can run quite slowly in older/slower systems.
# It can also consume a lot of memory (128 MB or more), you may need
# to raise your process resource limits (e.g. in bash, "ulimit -a"
# to inspect, and "ulimit -d ..." or "ulimit -m ..." to set)

my $start_time;
BEGIN { # Get the time the script started running; do it at compilation to
        # get it as close as possible
    $start_time= time;
}

require 5.010_001;
use strict;
use warnings;
use builtin qw(refaddr);
use Carp;
use Config;
use File::Find;
use File::Path;
use File::Spec;
use Text::Tabs;
use re "/aa";

use feature 'state';
use feature 'signatures';
no warnings qw( experimental::builtin );

sub DEBUG () { 0 }  # Set to 0 for production; 1 for development
$| = 1 if DEBUG;
my $debugging_build = $Config{"ccflags"} =~ /-DDEBUGGING/;

sub NON_ASCII_PLATFORM { ord("A") != 65 }

# When a new version of Unicode is published, unfortunately the algorithms for
# dealing with various bounds, like \b{gcb}, \b{lb} may have to be updated
# manually.  The changes may or may not be backward compatible with older
# releases.  The code is in regen/mk_invlist.pl and regexec.c.  Make the
# changes, then come back here and set the variable below to what version the
# code is expecting.  If a newer version of Unicode is being compiled than
# expected, a warning will be generated.  If an older version is being
# compiled, any bounds tests that fail in the generated test file (-maketest
# option) will be marked as TODO.
my $version_of_mk_invlist_bounds = v15.0.0;

##########################################################################
#
# mktables -- create the runtime Perl Unicode files (lib/unicore/.../*.pl),
# from the Unicode database files (lib/unicore/.../*.txt),  It also generates
# a pod file and .t files, depending on option parameters.
#
# The structure of this file is:
#   First these introductory comments; then
#   code needed for everywhere, such as debugging stuff; then
#   code to handle input parameters; then
#   data structures likely to be of external interest (some of which depend on
#       the input parameters, so follows them; then
#   more data structures and subroutine and package (class) definitions; then
#   the small actual loop to process the input files and finish up; then
#   a __DATA__ section, for the .t tests
#
# This program works on all releases of Unicode so far.  The outputs have been
# scrutinized most intently for release 5.1.  The others have been checked for
# somewhat more than just sanity.  It can handle all non-provisional Unicode
# character properties in those releases.
#
# This program is mostly about Unicode character (or code point) properties.
# A property describes some attribute or quality of a code point, like if it
# is lowercase or not, its name, what version of Unicode it was first defined
# in, or what its uppercase equivalent is.  Unicode deals with these disparate
# possibilities by making all properties into mappings from each code point
# into some corresponding value.  In the case of it being lowercase or not,
# the mapping is either to 'Y' or 'N' (or various synonyms thereof).  Each
# property maps each Unicode code point to a single value, called a "property
# value".  (Some more recently defined properties, map a code point to a set
# of values.)
#
# When using a property in a regular expression, what is desired isn't the
# mapping of the code point to its property's value, but the reverse (or the
# mathematical "inverse relation"): starting with the property value, "Does a
# code point map to it?"  These are written in a "compound" form:
# \p{property=value}, e.g., \p{category=punctuation}.  This program generates
# files containing the lists of code points that map to each such regular
# expression property value, one file per list
#
# There is also a single form shortcut that Perl adds for many of the commonly
# used properties.  This happens for all binary properties, plus script,
# general_category, and block properties.
#
# Thus the outputs of this program are files.  There are map files, mostly in
# the 'To' directory; and there are list files for use in regular expression
# matching, all in subdirectories of the 'lib' directory, with each
# subdirectory being named for the property that the lists in it are for.
# Bookkeeping, test, and documentation files are also generated.

my $matches_directory = 'lib';   # Where match (\p{}) files go.
my $map_directory = 'To';        # Where map files go.

# DATA STRUCTURES
#
# The major data structures of this program are Property, of course, but also
# Table.  There are two kinds of tables, very similar to each other.
# "Match_Table" is the data structure giving the list of code points that have
# a particular property value, mentioned above.  There is also a "Map_Table"
# data structure which gives the property's mapping from code point to value.
# There are two structures because the match tables need to be combined in
# various ways, such as constructing unions, intersections, complements, etc.,
# and the map ones don't.  And there would be problems, perhaps subtle, if
# a map table were inadvertently operated on in some of those ways.
# The use of separate classes with operations defined on one but not the other
# prevents accidentally confusing the two.
#
# At the heart of each table's data structure is a "Range_List", which is just
# an ordered list of "Ranges", plus ancillary information, and methods to
# operate on them.  A Range is a compact way to store property information.
# Each range has a starting code point, an ending code point, and a value that
# is meant to apply to all the code points between the two end points,
# inclusive.  For a map table, this value is the property value for those
# code points.  Two such ranges could be written like this:
#   0x41 .. 0x5A, 'Upper',
#   0x61 .. 0x7A, 'Lower'
#
# Each range also has a type used as a convenience to classify the values.
# Most ranges in this program will be Type 0, or normal, but there are some
# ranges that have a non-zero type.  These are used only in map tables, and
# are for mappings that don't fit into the normal scheme of things.  Mappings
# that require a hash entry to communicate with utf8.c are one example;
# another example is mappings for charnames.pm to use which indicate a name
# that is algorithmically determinable from its code point (and the reverse).
# These are used to significantly compact these tables, instead of listing
# each one of the tens of thousands individually.
#
# In a match table, the value of a range is irrelevant (and hence the type as
# well, which will always be 0), and arbitrarily set to the empty string.
# Using the example above, there would be two match tables for those two
# entries, one named Upper would contain the 0x41..0x5A range, and the other
# named Lower would contain 0x61..0x7A.
#
# Actually, there are two types of range lists, "Range_Map" is the one
# associated with map tables, and "Range_List" with match tables.
# Again, this is so that methods can be defined on one and not the others so
# as to prevent operating on them in incorrect ways.
#
# Eventually, most tables are written out to files to be read by Unicode::UCD.
# All tables could in theory be written, but some are suppressed because there
# is no current practical use for them.  It is easy to change which get
# written by changing various lists that are near the top of the actual code
# in this file.  The table data structures contain enough ancillary
# information to allow them to be treated as separate entities for writing,
# such as the path to each one's file.  There is a heading in each map table
# that gives the format of its entries, and what the map is for all the code
# points missing from it.  (This allows tables to be more compact.)
#
# The Property data structure contains one or more tables.  All properties
# contain a map table (except the $perl property which is a
# pseudo-property containing only match tables), and any properties that
# are usable in regular expression matches also contain various matching
# tables, one for each value the property can have.  A binary property can
# have two values, True and False (or Y and N, which are preferred by Unicode
# terminology).  Thus each of these properties will have a map table that
# takes every code point and maps it to Y or N (but having ranges cuts the
# number of entries in that table way down), and two match tables, one
# which has a list of all the code points that map to Y, and one for all the
# code points that map to N.  (For each binary property, a third table is also
# generated for the pseudo Perl property.  It contains the identical code
# points as the Y table, but can be written in regular expressions, not in the
# compound form, but in a "single" form like \p{IsUppercase}.)  Many
# properties are binary, but some properties have several possible values,
# some have many, and properties like Name have a different value for every
# named code point.  Those will not, unless the controlling lists are changed,
# have their match tables written out.  But all the ones which can be used in
# regular expression \p{} and \P{} constructs will.  Prior to 5.14, generally
# a property would have either its map table or its match tables written but
# not both.  Again, what gets written is controlled by lists which can easily
# be changed.  Starting in 5.14, advantage was taken of this, and all the map
# tables needed to reconstruct the Unicode db are now written out, while
# suppressing the Unicode .txt files that contain the data.  Our tables are
# much more compact than the .txt files, so a significant space savings was
# achieved.  Also, tables are not written out that are trivially derivable
# from tables that do get written.  So, there typically is no file containing
# the code points not matched by a binary property (the table for \P{} versus
# lowercase \p{}), since you just need to invert the True table to get the
# False table.

# Properties have a 'Type', like 'binary', or 'string', or 'enum' depending on
# how many match tables there are and the content of the maps.  This 'Type' is
# different than a range 'Type', so don't get confused by the two concepts
# having the same name.
#
# For information about the Unicode properties, see Unicode's UAX44 document:

my $unicode_reference_url = 'http://www.unicode.org/reports/tr44/';

# As stated earlier, this program will work on any release of Unicode so far.
# Most obvious problems in earlier data have NOT been corrected except when
# necessary to make Perl or this program work reasonably, and to keep out
# potential security issues.  For example, no folding information was given in
# early releases, so this program substitutes lower case instead, just so that
# a regular expression with the /i option will do something that actually
# gives the right results in many cases.  There are also a couple other
# corrections for version 1.1.5, commented at the point they are made.  As an
# example of corrections that weren't made (but could be) is this statement
# from DerivedAge.txt: "The supplementary private use code points and the
# non-character code points were assigned in version 2.0, but not specifically
# listed in the UCD until versions 3.0 and 3.1 respectively."  (To be precise
# it was 3.0.1 not 3.0.0)  More information on Unicode version glitches is
# further down in these introductory comments.
#
# This program works on all non-provisional properties as of the current
# Unicode release, though the files for some are suppressed for various
# reasons.  You can change which are output by changing lists in this program.
#
# The old version of mktables emphasized the term "Fuzzy" to mean Unicode's
# loose matchings rules (from Unicode TR18):
#
#    The recommended names for UCD properties and property values are in
#    PropertyAliases.txt [Prop] and PropertyValueAliases.txt
#    [PropValue]. There are both abbreviated names and longer, more
#    descriptive names. It is strongly recommended that both names be
#    recognized, and that loose matching of property names be used,
#    whereby the case distinctions, whitespace, hyphens, and underbar
#    are ignored.
#
# The program still allows Fuzzy to override its determination of if loose
# matching should be used, but it isn't currently used, as it is no longer
# needed; the calculations it makes are good enough.
#
# SUMMARY OF HOW IT WORKS:
#
#   Process arguments
#
#   A list is constructed containing each input file that is to be processed
#
#   Each file on the list is processed in a loop, using the associated handler
#   code for each:
#        The PropertyAliases.txt and PropValueAliases.txt files are processed
#            first.  These files name the properties and property values.
#            Objects are created of all the property and property value names
#            that the rest of the input should expect, including all synonyms.
#        The other input files give mappings from properties to property
#           values.  That is, they list code points and say what the mapping
#           is under the given property.  Some files give the mappings for
#           just one property; and some for many.  This program goes through
#           each file and populates the properties and their map tables from
#           them.  Some properties are listed in more than one file, and
#           Unicode has set up a precedence as to which has priority if there
#           is a conflict.  Thus the order of processing matters, and this
#           program handles the conflict possibility by processing the
#           overriding input files last, so that if necessary they replace
#           earlier values.
#        After this is all done, the program creates the property mappings not
#            furnished by Unicode, but derivable from what it does give.
#        The tables of code points that match each property value in each
#            property that is accessible by regular expressions are created.
#        The Perl-defined properties are created and populated.  Many of these
#            require data determined from the earlier steps
#        Any Perl-defined synonyms are created, and name clashes between Perl
#            and Unicode are reconciled and warned about.
#        All the properties are written to files
#        Any other files are written, and final warnings issued.
#
# For clarity, a number of operators have been overloaded to work on tables:
#   ~ means invert (take all characters not in the set).  The more
#       conventional '!' is not used because of the possibility of confusing
#       it with the actual boolean operation.
#   + means union
#   - means subtraction
#   & means intersection
# The precedence of these is the order listed.  Parentheses should be
# copiously used.  These are not a general scheme.  The operations aren't
# defined for a number of things, deliberately, to avoid getting into trouble.
# Operations are done on references and affect the underlying structures, so
# that the copy constructors for them have been overloaded to not return a new
# clone, but the input object itself.
#
# The bool operator is deliberately not overloaded to avoid confusion with
# "should it mean if the object merely exists, or also is non-empty?".
#
# WHY CERTAIN DESIGN DECISIONS WERE MADE
#
# This program needs to be able to run under miniperl.  Therefore, it uses a
# minimum of other modules, and hence implements some things itself that could
# be gotten from CPAN
#
# This program uses inputs published by the Unicode Consortium.  These can
# change incompatibly between releases without the Perl maintainers realizing
# it.  Therefore this program is now designed to try to flag these.  It looks
# at the directories where the inputs are, and flags any unrecognized files.
# It keeps track of all the properties in the files it handles, and flags any
# that it doesn't know how to handle.  It also flags any input lines that
# don't match the expected syntax, among other checks.
#
# It is also designed so if a new input file matches one of the known
# templates, one hopefully just needs to add it to a list to have it
# processed.
#
# As mentioned earlier, some properties are given in more than one file.  In
# particular, the files in the extracted directory are supposedly just
# reformattings of the others.  But they contain information not easily
# derivable from the other files, including results for Unihan (which isn't
# usually available to this program) and for unassigned code points.  They
# also have historically had errors or been incomplete.  In an attempt to
# create the best possible data, this program thus processes them first to
# glean information missing from the other files; then processes those other
# files to override any errors in the extracted ones.  Much of the design was
# driven by this need to store things and then possibly override them.
#
# It tries to keep fatal errors to a minimum, to generate something usable for
# testing purposes.  It always looks for files that could be inputs, and will
# warn about any that it doesn't know how to handle (the -q option suppresses
# the warning).
#
# Why is there more than one type of range?
#   This simplified things.  There are some very specialized code points that
#   have to be handled specially for output, such as Hangul syllable names.
#   By creating a range type (done late in the development process), it
#   allowed this to be stored with the range, and overridden by other input.
#   Originally these were stored in another data structure, and it became a
#   mess trying to decide if a second file that was for the same property was
#   overriding the earlier one or not.
#
# Why are there two kinds of tables, match and map?
#   (And there is a base class shared by the two as well.)  As stated above,
#   they actually are for different things.  Development proceeded much more
#   smoothly when I (khw) realized the distinction.  Map tables are used to
#   give the property value for every code point (actually every code point
#   that doesn't map to a default value).  Match tables are used for regular
#   expression matches, and are essentially the inverse mapping.  Separating
#   the two allows more specialized methods, and error checks so that one
#   can't just take the intersection of two map tables, for example, as that
#   is nonsensical.
#
# What about 'fate' and 'status'.  The concept of a table's fate was created
#   late when it became clear that something more was needed.  The difference
#   between this and 'status' is unclean, and could be improved if someone
#   wanted to spend the effort.
#
# DEBUGGING
#
# This program is written so it will run under miniperl.  Occasionally changes
# will cause an error where the backtrace doesn't work well under miniperl.
# To diagnose the problem, you can instead run it under regular perl, if you
# have one compiled.
#
# There is a good trace facility.  To enable it, first sub DEBUG must be set
# to return true.  Then a line like
#
# local $to_trace = 1 if main::DEBUG;
#
# can be added to enable tracing in its lexical scope (plus dynamic) or until
# you insert another line:
#
# local $to_trace = 0 if main::DEBUG;
#
# To actually trace, use a line like "trace $a, @b, %c, ...;
#
# Some of the more complex subroutines already have trace statements in them.
# Permanent trace statements should be like:
#
# trace ... if main::DEBUG && $to_trace;
#
# main::stack_trace() will display what its name implies
#
# If there is just one or a few files that you're debugging, you can easily
# cause most everything else to be skipped.  Change the line
#
# my $debug_skip = 0;
#
# to 1, and every file whose object is in @input_file_objects and doesn't have
# a, 'non_skip => 1,' in its constructor will be skipped.  However, skipping
# Jamo.txt or UnicodeData.txt will likely cause fatal errors.
#
# To compare the output tables, it may be useful to specify the -annotate
# flag.  (As of this writing, this can't be done on a clean workspace, due to
# requirements in Text::Tabs used in this option; so first run mktables
# without this option.)  This option adds comment lines to each table, one for
# each non-algorithmically named character giving, currently its code point,
# name, and graphic representation if printable (and you have a font that
# knows about it).  This makes it easier to see what the particular code
# points are in each output table.  Non-named code points are annotated with a
# description of their status, and contiguous ones with the same description
# will be output as a range rather than individually.  Algorithmically named
# characters are also output as ranges, except when there are just a few
# contiguous ones.
#
# FUTURE ISSUES
#
# The program would break if Unicode were to change its names so that
# interior white space, underscores, or dashes differences were significant
# within property and property value names.
#
# It might be easier to use the xml versions of the UCD if this program ever
# would need heavy revision, and the ability to handle old versions was not
# required.  Also, it turns out to be risky to rely on this, as in early 2024,
# Unicode decided to drop the xml version.  It was news to many that this was
# not considered to be an official product that needs to be maintained going
# forward.  Someone acceptable to the Unicode management volunteered to take
# over from the retiring volunteer, and so it continues, but beware.
#
# There is the potential for name collisions, in that Perl has chosen names
# that Unicode could decide it also likes.  There have been such collisions in
# the past, with mostly Perl deciding to adopt the Unicode definition of the
# name.  However in the 5.2 Unicode beta testing, there were a number of such
# collisions, which were withdrawn before the final release, because of Perl's
# and other's protests.  These all involved new properties which began with
# 'Is'.  Based on the protests, Unicode is unlikely to try that again.  Also,
# many of the Perl-defined synonyms, like Any, Word, etc, are listed in a
# Unicode document, so they are unlikely to be used by Unicode for another
# purpose.  However, they might try something beginning with 'In', or use any
# of the other Perl-defined properties.  This program will warn you of name
# collisions, and refuse to generate tables with them, but manual intervention
# will be required in this event.  One scheme that could be implemented, if
# necessary, would be to have this program generate another file, or add a
# field to mktables.lst that gives the date of first definition of a property.
# Each new release of Unicode would use that file as a basis for the next
# iteration.  And the Perl synonym addition code could sort based on the age
# of the property, so older properties get priority, and newer ones that clash
# would be refused; hence existing code would not be impacted, and some other
# synonym would have to be used for the new property.  This is ugly, and
# manual intervention would certainly be easier to do in the short run; lets
# hope it never comes to this.
#
# A NOTE ON UNIHAN
#
# This program can generate tables from the Unihan database.  But that DB
# isn't normally available, so it is marked as optional.  Prior to version
# 5.2, this database was in a single file, Unihan.txt.  In 5.2 the database
# was split into 8 different files, all beginning with the letters 'Unihan'.
# If you plunk those files down into the directory mktables ($0) is in, this
# program will read them and automatically create tables for the properties
# from it that are listed in PropertyAliases.txt and PropValueAliases.txt,
# plus any you add to the @cjk_properties array and the @cjk_property_values
# array, being sure to add necessary '# @missings' lines to the latter.  For
# Unicode versions earlier than 5.2, most of the Unihan properties are not
# listed at all in PropertyAliases nor PropValueAliases.  This program assumes
# for these early releases that you want the properties that are specified in
# the 5.2 release.
#
# You may need to adjust the entries to suit your purposes.  setup_unihan(),
# and filter_unihan_line() are the functions where this is done.  This program
# already does some adjusting to make the lines look more like the rest of the
# Unicode DB;  You can see what that is in filter_unihan_line()
#
# There is a bug in the 3.2 data file in which some values for the
# kPrimaryNumeric property have commas and an unexpected comment.  A filter
# could be added to correct these; or for a particular installation, the
# Unihan.txt file could be edited to fix them.
#
# HOW TO ADD A FILE TO BE PROCESSED
#
# A new file from Unicode needs to have an object constructed for it in
# @input_file_objects, probably at the end or at the end of the extracted
# ones.  The program should warn you if its name will clash with others on
# restrictive file systems, like DOS.  If so, figure out a better name, and
# add lines to the README.perl file giving that.  If the file is a character
# property, it should be in the format that Unicode has implicitly
# standardized for such files for the more recently introduced ones.
# If so, the Input_file constructor for @input_file_objects can just be the
# file name and release it first appeared in.  If not, then it should be
# possible to construct an each_line_handler() to massage the line into the
# standardized form.
#
# For non-character properties, more code will be needed.  You can look at
# the existing entries for clues.
#
# UNICODE VERSIONS NOTES
#
# The Unicode UCD has had a number of errors in it over the versions.  And
# these remain, by policy, in the standard for that version.  Therefore it is
# risky to correct them, because code may be expecting the error.  So this
# program doesn't generally make changes, unless the error breaks the Perl
# core.  As an example, some versions of 2.1.x Jamo.txt have the wrong value
# for U+1105, which causes real problems for the algorithms for Jamo
# calculations, so it is changed here.
#
# But it isn't so clear cut as to what to do about concepts that are
# introduced in a later release; should they extend back to earlier releases
# where the concept just didn't exist?  It was easier to do this than to not,
# so that's what was done.  For example, the default value for code points not
# in the files for various properties was probably undefined until changed by
# some version.  No_Block for blocks is such an example.  This program will
# assign No_Block even in Unicode versions that didn't have it.  This has the
# benefit that code being written doesn't have to special case earlier
# versions; and the detriment that it doesn't match the Standard precisely for
# the affected versions.
#
# Here are some observations about some of the issues in early versions:
#
# Prior to version 3.0, there were 3 character decompositions.  These are not
# handled by Unicode::Normalize, nor will it compile when presented a version
# that has them.  However, you can trivially get it to compile by simply
# ignoring those decompositions, by changing the croak to a carp.  At the time
# of this writing, the line (in dist/Unicode-Normalize/Normalize.pm or
# dist/Unicode-Normalize/mkheader) reads
#
#   croak("Weird Canonical Decomposition of U+$h");
#
# Simply comment it out.  It will compile, but will not know about any three
# character decompositions.

# The number of code points in \p{alpha=True} halved in 2.1.9.  It turns out
# that the reason is that the CJK block starting at 4E00 was removed from
# PropList, and was not put back in until 3.1.0.  The Perl extension (the
# single property name \p{alpha}) has the correct values.  But the compound
# form is simply not generated until 3.1, as it can be argued that prior to
# this release, this was not an official property.  The comments for
# filter_old_style_proplist() give more details.
#
# Unicode introduced the synonym Space for White_Space in 4.1.  Perl has
# always had a \p{Space}.  In release 3.2 only, they are not synonymous.  The
# reason is that 3.2 introduced U+205F=medium math space, which was not
# classed as white space, but Perl figured out that it should have been. 4.0
# reclassified it correctly.
#
# Another change between 3.2 and 4.0 is the CCC property value ATBL.  In 3.2
# this was erroneously a synonym for 202 (it should be 200).  In 4.0, ATB
# became 202, and ATBL was left with no code points, as all the ones that
# mapped to 202 stayed mapped to 202.  Thus if your program used the numeric
# name for the class, it would not have been affected, but if it used the
# mnemonic, it would have been.
#
# \p{Script=Hrkt} (Katakana_Or_Hiragana) came in 4.0.1.  Before that, code
# points which eventually came to have this script property value, instead
# mapped to "Unknown".  But in the next release all these code points were
# moved to \p{sc=common} instead.

# The tests furnished  by Unicode for testing WordBreak and SentenceBreak
# generate errors in 5.0 and earlier.
#
# The default for missing code points for BidiClass is complicated.  Starting
# in 3.1.1, the derived file DBidiClass.txt handles this, but this program
# tries to do the best it can for earlier releases.  It is done in
# process_PropertyAliases()
#
# In version 2.1.2, the entry in UnicodeData.txt:
#   0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;;019F;
# should instead be
#   0275;LATIN SMALL LETTER BARRED O;Ll;0;L;;;;;N;;;019F;;019F
# Without this change, there are casing problems for this character.
#
# Search for $string_compare_versions to see how to compare changes to
# properties between Unicode versions
#
##############################################################################

my $UNDEF = ':UNDEF:';  # String to print out for undefined values in tracing
                        # and errors
my $MAX_LINE_WIDTH = 78;

# Debugging aid to skip most files so as to not be distracted by them when
# concentrating on the ones being debugged.  Add
# non_skip => 1,
# to the constructor for those files you want processed when you set this.
# Files with a first version number of 0 are special: they are always
# processed regardless of the state of this flag.  Generally, Jamo.txt and
# UnicodeData.txt must not be skipped if you want this program to not die
# before normal completion.
my $debug_skip = 0;


# Normally these are suppressed.
my $write_Unicode_deprecated_tables = 0;

# Set to 1 to enable tracing.
our $to_trace = 0;

{ # Closure for trace: debugging aid
    my $print_caller = 1;        # ? Include calling subroutine name
    my $main_with_colon = 'main::';
    my $main_colon_length = length($main_with_colon);

    sub trace {
        return unless $to_trace;        # Do nothing if global flag not set

        my @input = @_;

        local $DB::trace = 0;
        $DB::trace = 0;          # Quiet 'used only once' message

        my $line_number;

        # Loop looking up the stack to get the first non-trace caller
        my $caller_line;
        my $caller_name;
        my $i = 0;
        do {
            $line_number = $caller_line;
            (my $pkg, my $file, $caller_line, my $caller) = caller $i++;
            $caller = $main_with_colon unless defined $caller;

            $caller_name = $caller;

            # get rid of pkg
            $caller_name =~ s/.*:://;
            if (substr($caller_name, 0, $main_colon_length)
                eq $main_with_colon)
            {
                $caller_name = substr($caller_name, $main_colon_length);
            }

        } until ($caller_name ne 'trace');

        # If the stack was empty, we were called from the top level
        $caller_name = 'main' if ($caller_name eq ""
                                    || $caller_name eq 'trace');

        my $output = "";
        #print STDERR __LINE__, ": ", join ", ", @input, "\n";
        foreach my $string (@input) {
            if (ref $string eq 'ARRAY' || ref $string eq 'HASH') {
                $output .= simple_dumper($string);
            }
            else {
                $string = "$string" if ref $string;
                $string = $UNDEF unless defined $string;
                chomp $string;
                $string = '""' if $string eq "";
                $output .= " " if $output ne ""
                                && $string ne ""
                                && substr($output, -1, 1) ne " "
                                && substr($string, 0, 1) ne " ";
                $output .= $string;
            }
        }

        print STDERR sprintf "%4d: ", $line_number if defined $line_number;
        print STDERR "$caller_name: " if $print_caller;
        print STDERR $output, "\n";
        return;
    }
}

sub stack_trace() {
    local $to_trace = 1 if main::DEBUG;
    my $line = (caller(0))[2];
    my $i = 1;

    # Accumulate the stack trace
    while (1) {
        my ($pkg, $file, $caller_line, $caller) = caller $i++;

        last unless defined $caller;

        trace "called from $caller() at line $line";
        $line = $caller_line;
    }
}

# This is for a rarely used development feature that allows you to compare two
# versions of the Unicode standard without having to deal with changes caused
# by the code points introduced in the later version.  You probably also want
# to use the -annotate option when using this.  Run this program on a unicore
# containing the starting release you want to compare.  Save that output
# structure.  Then, switching to a unicore with the ending release, change the
# "" in the $string_compare_versions definition just below to a string
# containing a SINGLE dotted Unicode release number (e.g. "2.1") corresponding
# to the starting release.  This program will then compile, but throw away all
# code points introduced after the starting release.  Finally use a diff tool
# to compare the two directory structures.  They include only the code points
# common to both releases, and you can see the changes caused just by the
# underlying release semantic changes.  For versions earlier than 3.2, you
# must copy a version of DAge.txt into the directory.
my $string_compare_versions = DEBUG && "";
my $compare_versions = DEBUG
                       && $string_compare_versions
                       && pack "C*", split /\./, $string_compare_versions;

sub uniques {
    # Returns non-duplicated input values.  From "Perl Best Practices:
    # Encapsulated Cleverness".  p. 455 in first edition.

    my %seen;
    # Arguably this breaks encapsulation, if the goal is to permit multiple
    # distinct objects to stringify to the same value, and be interchangeable.
    # However, for this program, no two objects stringify identically, and all
    # lists passed to this function are either objects or strings. So this
    # doesn't affect correctness, but it does give a couple of percent speedup.
    no overloading;
    return grep { ! $seen{$_}++ } @_;
}

$0 = File::Spec->canonpath($0);

my $make_test_script = 0;      # ? Should we output a test script
my $make_norm_test_script = 0; # ? Should we output a normalization test script
my $write_unchanged_files = 0; # ? Should we update the output files even if
                               #    we don't think they have changed
my $use_directory = "";        # ? Should we chdir somewhere.
my $pod_directory;             # input directory to store the pod file.
my $pod_file = 'perluniprops';
my $t_path;                     # Path to the .t test file
my $file_list = 'mktables.lst'; # File to store input and output file names.
                               # This is used to speed up the build, by not
                               # executing the main body of the program if
                               # nothing on the list has changed since the
                               # previous build
my $make_list = 1;             # ? Should we write $file_list.  Set to always
                               # make a list so that when the release manager
                               # is preparing a release, they won't have to do
                               # special things
my $glob_list = 0;             # ? Should we try to include unknown .txt files
                               # in the input.
my $output_range_counts = $debugging_build;   # ? Should we include the number
                                              # of code points in ranges in
                                              # the output
my $annotate = 0;              # ? Should character names be in the output

# Verbosity levels; 0 is quiet
my $NORMAL_VERBOSITY = 1;
my $PROGRESS = 2;
my $VERBOSE = 3;

my $verbosity = $NORMAL_VERBOSITY;

# Stored in mktables.lst so that if this program is called with different
# options, will regenerate even if the files otherwise look like they're
# up-to-date.
my $command_line_arguments = join " ", @ARGV;

# Process arguments
while (@ARGV) {
    my $arg = shift @ARGV;
    if ($arg eq '-v') {
        $verbosity = $VERBOSE;
    }
    elsif ($arg eq '-p') {
        $verbosity = $PROGRESS;
        $| = 1;     # Flush buffers as we go.
    }
    elsif ($arg eq '-q') {
        $verbosity = 0;
    }
    elsif ($arg eq '-w') {
        # update the files even if they haven't changed
        $write_unchanged_files = 1;
    }
    elsif ($arg eq '-check') {
        my $this = shift @ARGV;
        my $ok = shift @ARGV;
        if ($this ne $ok) {
            print "Skipping as check params are not the same.\n";
            exit(0);
        }
    }
    elsif ($arg eq '-P' && defined ($pod_directory = shift)) {
        -d $pod_directory or croak "Directory '$pod_directory' doesn't exist";
    }
    elsif ($arg eq '-maketest' || ($arg eq '-T' && defined ($t_path = shift)))
    {
        $make_test_script = 1;
    }
    elsif ($arg eq '-makenormtest')
    {
        $make_norm_test_script = 1;
    }
    elsif ($arg eq '-makelist') {
        $make_list = 1;
    }
    elsif ($arg eq '-C' && defined ($use_directory = shift)) {
        -d $use_directory or croak "Unknown directory '$use_directory'";
    }
    elsif ($arg eq '-L') {

        # Existence not tested until have chdir'd
        $file_list = shift;
    }
    elsif ($arg eq '-globlist') {
        $glob_list = 1;
    }
    elsif ($arg eq '-c') {
        $output_range_counts = ! $output_range_counts
    }
    elsif ($arg eq '-annotate') {
        $annotate = 1;
        $debugging_build = 1;
        $output_range_counts = 1;
    }
    else {
        my $with_c = 'with';
        $with_c .= 'out' if $output_range_counts;   # Complements the state
        croak <<END;
usage: $0 [-c|-p|-q|-v|-w] [-C dir] [-L filelist] [ -P pod_dir ]
          [ -T test_file_path ] [-globlist] [-makelist] [-maketest]
          [-check A B ]
  -c          : Output comments $with_c number of code points in ranges
  -q          : Quiet Mode: Only output serious warnings.
  -p          : Set verbosity level to normal plus show progress.
  -v          : Set Verbosity level high:  Show progress and non-serious
                warnings
  -w          : Write files regardless
  -C dir      : Change to this directory before proceeding. All relative paths
                except those specified by the -P and -T options will be done
                with respect to this directory.
  -P dir      : Output $pod_file file to directory 'dir'.
  -T path     : Create a test script as 'path'; overrides -maketest
  -L filelist : Use alternate 'filelist' instead of standard one
  -globlist   : Take as input all non-Test *.txt files in current and sub
                directories
  -maketest   : Make test script 'TestProp.pl' in current (or -C directory),
                overrides -T
  -makelist   : Rewrite the file list $file_list based on current setup
  -annotate   : Output an annotation for each character in the table files;
                useful for debugging mktables, looking at diffs; but is slow
                and memory intensive
  -check A B  : Executes $0 only if A and B are the same
END
    }
}

# Stores the most-recently changed file.  If none have changed, can skip the
# build
my $most_recent = (stat $0)[9];   # Do this before the chdir!

# Change directories now, because need to read 'version' early.
if ($use_directory) {
    if ($pod_directory && ! File::Spec->file_name_is_absolute($pod_directory)) {
        $pod_directory = File::Spec->rel2abs($pod_directory);
    }
    if ($t_path && ! File::Spec->file_name_is_absolute($t_path)) {
        $t_path = File::Spec->rel2abs($t_path);
    }
    chdir $use_directory or croak "Failed to chdir to '$use_directory':$!";
    if ($pod_directory && File::Spec->file_name_is_absolute($pod_directory)) {
        $pod_directory = File::Spec->abs2rel($pod_directory);
    }
    if ($t_path && File::Spec->file_name_is_absolute($t_path)) {
        $t_path = File::Spec->abs2rel($t_path);
    }
}

# Get Unicode version into regular and v-string.  This is done now because
# various tables below get populated based on it.  These tables are populated
# here to be near the top of the file, and so easily seeable by those needing
# to modify things.
open my $VERSION, "<", "version"
                    or croak "$0: can't open required file 'version': $!\n";
my $string_version = <$VERSION>;
close $VERSION;
chomp $string_version;
my $v_version = pack "C*", split /\./, $string_version;        # v string

my $unicode_version = ($compare_versions)
                      ? (  "$string_compare_versions (using "
                         . "$string_version rules)")
                      : $string_version;

# The following are the complete names of properties with property values that
# are known to not match any code points in some versions of Unicode, but that
# may change in the future so they should be matchable, hence an empty file is
# generated for them.
my @tables_that_may_be_empty;
push @tables_that_may_be_empty, 'Joining_Type=Left_Joining'
                                                    if $v_version lt v6.3.0;
push @tables_that_may_be_empty, 'Script=Common' if $v_version le v4.0.1;
push @tables_that_may_be_empty, 'Title' if $v_version lt v2.0.0;
push @tables_that_may_be_empty, 'Script=Katakana_Or_Hiragana'
                                                    if $v_version ge v4.1.0;
push @tables_that_may_be_empty, 'Script_Extensions=Katakana_Or_Hiragana'
                                                    if $v_version ge v6.0.0;
push @tables_that_may_be_empty, 'Grapheme_Cluster_Break=Prepend'
                                                    if $v_version ge v6.1.0;
push @tables_that_may_be_empty, 'Canonical_Combining_Class=CCC133'
                                                    if $v_version ge v6.2.0;

# These properties of Egyptian hieroglyphs are not yet handled by Perl.  Their
# intended audience is only specialist Egyptologists
push @tables_that_may_be_empty, qw(kEH_Cat kEH_Desc kEH_HG kEH_IFAO
                                   kEH_JSesh
                                   kEH_NoMirror kEH_NoMirror=Yes
                                   kEH_NoMirror=No
                                   kEH_NoRotate kEH_NoRotate=Yes)
                                                    if $v_version ge v16.0.0;

# The lists below are hashes, so the key is the item in the list, and the
# value is the reason why it is in the list.  This makes generation of
# documentation easier.

my %why_suppressed;  # No file generated for these.

# Files aren't generated for empty extraneous properties.  This is arguable.
# Extraneous properties generally come about because a property is no longer
# used in a newer version of Unicode.  If we generated a file without code
# points, programs that used to work on that property will still execute
# without errors.  It just won't ever match (or will always match, with \P{}).
# This means that the logic is now likely wrong.  I (khw) think its better to
# find this out by getting an error message.  Just move them to the table
# above to change this behavior
my %why_suppress_if_empty_warn_if_not = (

   # It is the only property that has ever officially been removed from the
   # Standard.  The database never contained any code points for it.
   'Special_Case_Condition' => 'Obsolete',

   # Apparently never official, but there were code points in some versions of
   # old-style PropList.txt
   'Non_Break' => 'Obsolete',
);

# These would normally go in the warn table just above, but they were changed
# a long time before this program was written, so warnings about them are
# moot.
if ($v_version gt v3.2.0) {
    push @tables_that_may_be_empty,
                                'Canonical_Combining_Class=Attached_Below_Left'
}

# Obsoleted
if ($v_version ge v11.0.0) {
    push @tables_that_may_be_empty, qw(
                                       Grapheme_Cluster_Break=E_Base
                                       Grapheme_Cluster_Break=E_Base_GAZ
                                       Grapheme_Cluster_Break=E_Modifier
                                       Grapheme_Cluster_Break=Glue_After_Zwj
                                       Word_Break=E_Base
                                       Word_Break=E_Base_GAZ
                                       Word_Break=E_Modifier
                                       Word_Break=Glue_After_Zwj);
}

# Enum values for to_output_map() method in the Map_Table package. (0 is don't
# output)
my $EXTERNAL_MAP = 1;
my $INTERNAL_MAP = 2;
my $OUTPUT_ADJUSTED = 3;

# To override computed values for writing the map tables for these properties.
# The default for enum map tables is to write them out, so that the Unicode
# .txt files can be removed, but all the data to compute any property value
# for any code point is available in a more compact form.
my %global_to_output_map = (
    # Needed by UCD.pm, but don't want to publicize that it exists, so won't
    # get stuck supporting it if things change.  Since it is a STRING
    # property, it normally would be listed in the pod, but INTERNAL_MAP
    # suppresses that.
    Unicode_1_Name => $INTERNAL_MAP,

    Present_In => 0,                # Suppress, as easily computed from Age
    Block => (NON_ASCII_PLATFORM) ? 1 : 0,  # Suppress, as Blocks.txt is
                                            # retained, but needed for
                                            # non-ASCII

    # Suppress, as mapping can be found instead from the
    # Perl_Decomposition_Mapping file
    Decomposition_Type => 0,
);

# There are several types of obsolete properties defined by Unicode.  These
# must be hand-edited for every new Unicode release.
my %why_deprecated;  # Generates a deprecated warning message if used.
my %why_stabilized;  # Documentation only
my %why_obsolete;    # Documentation only

{   # Closure
    my $simple = 'Perl uses the more complete version';
    my $unihan = 'Unihan properties are by default not enabled in the Perl core.';

    my $other_properties = 'other properties';
    my $contributory = "Used by Unicode internally for generating $other_properties and not intended to be used stand-alone";
    my $why_no_expand  = "Deprecated by Unicode.  These are characters that expand to more than one character in the specified normalization form, but whether they actually take up more bytes or not depends on the encoding being used.  For example, a UTF-8 encoded character may expand to a different number of bytes than a UTF-32 encoded character.";

    %why_deprecated = (
        'Grapheme_Link' => 'Duplicates ccc=vr (Canonical_Combining_Class=Virama)',
        'Jamo_Short_Name' => $contributory,
        'Line_Break=Surrogate' => 'Surrogates should never appear in well-formed text, and therefore shouldn\'t be the basis for line breaking',
        'Other_Alphabetic' => $contributory,
        'Other_Default_Ignorable_Code_Point' => $contributory,
        'Other_Grapheme_Extend' => $contributory,
        'Other_ID_Continue' => $contributory,
        'Other_ID_Start' => $contributory,
        'Other_Lowercase' => $contributory,
        'Other_Math' => $contributory,
        'Other_Uppercase' => $contributory,
        'Expands_On_NFC' => $why_no_expand,
        'Expands_On_NFD' => $why_no_expand,
        'Expands_On_NFKC' => $why_no_expand,
        'Expands_On_NFKD' => $why_no_expand,
    );

    %why_suppressed = (
        # There is a lib/unicore/Decomposition.pl (used by Normalize.pm) which
        # contains the same information, but without the algorithmically
        # determinable Hangul syllables'.  This file is not published, so it's
        # existence is not noted in the comment.
        'Decomposition_Mapping' => 'Accessible via Unicode::Normalize or prop_invmap() or charprop() in Unicode::UCD::',

        # Don't suppress ISO_Comment, as otherwise special handling is needed
        # to differentiate between it and gc=c, which can be written as 'isc',
        # which is the same characters as ISO_Comment's short name.

        'Name' => "Accessible via \\N{...} or 'use charnames;' or charprop() or prop_invmap() in Unicode::UCD::",

        'Simple_Case_Folding' => "$simple.  Can access this through casefold(), charprop(), or prop_invmap() in Unicode::UCD",
        'Simple_Lowercase_Mapping' => "$simple.  Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",
        'Simple_Titlecase_Mapping' => "$simple.  Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",
        'Simple_Uppercase_Mapping' => "$simple.  Can access this through charinfo(), charprop(), or prop_invmap() in Unicode::UCD",

        FC_NFKC_Closure => 'Deprecated by Unicode, and supplanted in usage by NFKC_Casefold; otherwise not useful',
    );

    foreach my $property (

            # The following are suppressed because they were made contributory
            # or deprecated by Unicode before Perl ever thought about
            # supporting them.
            'Jamo_Short_Name',
            'Grapheme_Link',
            'Expands_On_NFC',
            'Expands_On_NFD',
            'Expands_On_NFKC',
            'Expands_On_NFKD',

            # The following are suppressed because they have been marked
            # as deprecated for a sufficient amount of time
            'Other_Alphabetic',
            'Other_Default_Ignorable_Code_Point',
            'Other_Grapheme_Extend',
            'Other_ID_Continue',
            'Other_ID_Start',
            'Other_Lowercase',
            'Other_Math',
            'Other_Uppercase',
    ) {
        $why_suppressed{$property} = $why_deprecated{$property};
    }

    # Customize the message for all the 'Other_' properties
    foreach my $property (keys %why_deprecated) {
        next if (my $main_property = $property) !~ s/^Other_//;
        $why_deprecated{$property} =~ s/$other_properties/the $main_property property (which should be used instead)/;
    }
}

if ($write_Unicode_deprecated_tables) {
    foreach my $property (keys %why_suppressed) {
        delete $why_suppressed{$property} if $property =~
                                                    / ^ Other | Grapheme /x;
    }
}

if ($v_version ge 4.0.0) {
    $why_stabilized{'Hyphen'} = 'Use the Line_Break property instead; see www.unicode.org/reports/tr14';
    if ($v_version ge 6.0.0) {
        $why_deprecated{'Hyphen'} = 'Supplanted by Line_Break property values; see www.unicode.org/reports/tr14';
    }
}
if ($v_version ge 5.2.0 && $v_version lt 6.0.0) {
    $why_obsolete{'ISO_Comment'} = 'Code points for it have been removed';
    if ($v_version ge 6.0.0) {
        $why_deprecated{'ISO_Comment'} = 'No longer needed for Unicode\'s internal chart generation; otherwise not useful, and code points for it have been removed';
    }
}

# Probably obsolete forever
if ($v_version ge v4.1.0) {
    $why_suppressed{'Script=Katakana_Or_Hiragana'} = 'Obsolete.  All code points previously matched by this have been moved to "Script=Common".';
}
if ($v_version ge v6.0.0) {
    $why_suppressed{'Script=Katakana_Or_Hiragana'} .= '  Consider instead using "Script_Extensions=Katakana" or "Script_Extensions=Hiragana" (or both)';
    $why_suppressed{'Script_Extensions=Katakana_Or_Hiragana'} = 'All code points that would be matched by this are matched by either "Script_Extensions=Katakana" or "Script_Extensions=Hiragana"';
}

# This program can create files for enumerated-like properties, such as
# 'Numeric_Type'.  This file would be the same format as for a string
# property, with a mapping from code point to its value, so you could look up,
# for example, the script a code point is in.  But no one so far wants this
# mapping, or they have found another way to get it since this is a new
# feature.  So no file is generated except if it is in this list.
my @output_mapped_properties = split "\n", <<END;
END

# If you want more Unihan properties than the default, you need to add them to
# these arrays.  Depending on the property type, @missing lines might have to
# be added to the second array.  A sample entry would be (including the '#'):
# @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
my @cjk_properties = split "\n", <<'END';
END
my @cjk_property_values = split "\n", <<'END';
END

# The input files don't list every code point.  Those not listed are to be
# defaulted to some value.  Below are hard-coded what those values are for
# non-binary properties as of 5.1.  Starting in 5.0, there are
# machine-parsable comment lines in the files that give the defaults; so this
# list shouldn't have to be extended.  The claim is that all missing entries
# for binary properties will default to 'N'.  Unicode tried to change that in
# 5.2, but the beta period produced enough protest that they backed off.
#
# The defaults for the fields that appear in UnicodeData.txt in this hash must
# be in the form that it expects.  The others may be synonyms.
my $CODE_POINT = '<code point>';
my %default_mapping = (
    Age => "Unassigned",
    # Bidi_Class => Complicated; set in code
    Bidi_Mirroring_Glyph => "",
    Block => 'No_Block',
    Canonical_Combining_Class => 0,
    Case_Folding => $CODE_POINT,
    Decomposition_Mapping => $CODE_POINT,
    Decomposition_Type => 'None',
    East_Asian_Width => "Neutral",
    FC_NFKC_Closure => $CODE_POINT,
    General_Category => ($v_version le 6.3.0) ? 'Cn' : 'Unassigned',
    Grapheme_Cluster_Break => 'Other',
    Hangul_Syllable_Type => 'NA',
    ISO_Comment => "",
    Jamo_Short_Name => "",
    Joining_Group => "No_Joining_Group",
    # Joining_Type => Complicated; set in code
    kIICore => 'N',   #                       Is converted to binary
    #Line_Break => Complicated; set in code
    Lowercase_Mapping => $CODE_POINT,
    Name => "",
    Name_Alias => "",
    NFC_QC => 'Yes',
    NFD_QC => 'Yes',
    NFKC_QC => 'Yes',
    NFKD_QC => 'Yes',
    Numeric_Type => 'None',
    Numeric_Value => 'NaN',
    Script => ($v_version le 4.1.0) ? 'Common' : 'Unknown',
    Sentence_Break => 'Other',
    Simple_Case_Folding => $CODE_POINT,
    Simple_Lowercase_Mapping => $CODE_POINT,
    Simple_Titlecase_Mapping => $CODE_POINT,
    Simple_Uppercase_Mapping => $CODE_POINT,
    Titlecase_Mapping => $CODE_POINT,
    Unicode_1_Name => "",
    Unicode_Radical_Stroke => "",
    Uppercase_Mapping => $CODE_POINT,
    Word_Break => 'Other',
);

### End of externally interesting definitions, except for @input_file_objects

my $HEADER=<<"EOF";
# !!!!!!!   DO NOT EDIT THIS FILE   !!!!!!!
# This file is machine-generated by $0 from the Unicode
# database, Version $unicode_version.  Any changes made here will be lost!
EOF

my $INTERNAL_ONLY_HEADER = <<"EOF";

# !!!!!!!   INTERNAL PERL USE ONLY   !!!!!!!
# This file is for internal use by core Perl only.  The format and even the
# name or existence of this file are subject to change without notice.  Don't
# use it directly.  Use Unicode::UCD to access the Unicode character data
# base.
EOF

my $DEVELOPMENT_ONLY=<<"EOF";
# !!!!!!!   DEVELOPMENT USE ONLY   !!!!!!!
# This file contains information artificially constrained to code points
# present in Unicode release $string_compare_versions.
# IT CANNOT BE RELIED ON.  It is for use during development only and should
# not be used for production.

EOF

my $MAX_UNICODE_CODEPOINT_STRING = ($v_version ge v2.0.0)
                                   ? "10FFFF"
                                   : "FFFF";
my $MAX_UNICODE_CODEPOINT = hex $MAX_UNICODE_CODEPOINT_STRING;
my $MAX_UNICODE_CODEPOINTS = $MAX_UNICODE_CODEPOINT + 1;

# We work with above-Unicode code points, up to IV_MAX, but we may want to use
# sentinels above that number.  Therefore for internal use, we use a much
# smaller number, translating it to IV_MAX only for output.  The exact number
# is immaterial (all above-Unicode code points are treated exactly the same),
# but the algorithm requires it to be at least
# 2 * $MAX_UNICODE_CODEPOINTS + 1
my $MAX_WORKING_CODEPOINTS= $MAX_UNICODE_CODEPOINT * 8;
my $MAX_WORKING_CODEPOINT = $MAX_WORKING_CODEPOINTS - 1;
my $MAX_WORKING_CODEPOINT_STRING = sprintf("%X", $MAX_WORKING_CODEPOINT);

my $MAX_PLATFORM_CODEPOINT = ~0 >> 1;

# Matches legal code point.  4-6 hex numbers, If there are 6, the first
# two must be 10; if there are 5, the first must not be a 0.  Written this way
# to decrease backtracking.  The first regex allows the code point to be at
# the end of a word, but to work properly, the word shouldn't end with a valid
# hex character.  The second one won't match a code point at the end of a
# word, and doesn't have the run-on issue
my $run_on_code_point_re =
            qr/ (?: 10[0-9A-F]{4} | [1-9A-F][0-9A-F]{4} | [0-9A-F]{4} ) \b/x;
my $code_point_re = qr/\b$run_on_code_point_re/;

# This matches the beginning of the line in the Unicode DB files that give the
# defaults for code points not listed (i.e., missing) in the file.  The code
# depends on this ending with a semi-colon, so it can assume it is a valid
# field when the line is split() by semi-colons
my $missing_defaults_prefix = qr/ ^ \# \s+ \@missing: \s+
                                           ($code_point_re)
                                           \.\.
                                           ($code_point_re)
                                       \s* ;
                                /x;

# Property types.  Unicode has more types, but these are sufficient for our
# purposes.
my $UNKNOWN = -1;   # initialized to illegal value
my $NON_STRING = 1; # Either binary or enum
my $BINARY = 2;
my $FORCED_BINARY = 3; # Not a binary property, but, besides its normal
                       # tables, additional true and false tables are
                       # generated so that false is anything matching the
                       # default value, and true is everything else.
my $ENUM = 4;       # Include catalog
my $STRING = 5;     # Anything else: string or misc

# Some input files have lines that give default values for code points not
# contained in the file.  Sometimes these should be ignored.
my $NO_DEFAULTS = 0;        # Must evaluate to false
my $NOT_IGNORED = 1;
my $IGNORED = 2;

# Range types.  Each range has a type.  Most ranges are type 0, for normal,
# and will appear in the main body of the tables in the output files, but
# there are other types of ranges as well, listed below, that are specially
# handled.   There are pseudo-types as well that will never be stored as a
# type, but will affect the calculation of the type.

# 0 is for normal, non-specials
my $MULTI_CP = 1;           # Sequence of more than code point
my $HANGUL_SYLLABLE = 2;
my $CP_IN_NAME = 3;         # The NAME contains the code point appended to it.
my $NULL = 4;               # The map is to the null string; utf8.c can't
                            # handle these, nor is there an accepted syntax
                            # for them in \p{} constructs
my $COMPUTE_NO_MULTI_CP = 5; # Pseudo-type; means that ranges that would
                             # otherwise be $MULTI_CP type are instead type 0

# process_generic_property_file() can accept certain overrides in its input.
# Each of these must begin AND end with $CMD_DELIM.
my $CMD_DELIM = "\a";
my $REPLACE_CMD = 'replace';    # Override the Replace
my $MAP_TYPE_CMD = 'map_type';  # Override the Type

my $NO = 0;
my $YES = 1;

# Values for the Replace argument to add_range.
# $NO                      # Don't replace; add only the code points not
                           # already present.
my $IF_NOT_EQUIVALENT = 1; # Replace only under certain conditions; details in
                           # the comments at the subroutine definition.
my $UNCONDITIONALLY = 2;   # Replace without conditions.
my $MULTIPLE_BEFORE = 4;   # Don't replace, but add a duplicate record if
                           # already there
my $MULTIPLE_AFTER = 5;    # Don't replace, but add a duplicate record if
                           # already there
my $CROAK = 6;             # Die with an error if is already there

# Flags to give property statuses.  The phrases are to remind maintainers that
# if the flag is changed, the indefinite article referring to it in the
# documentation may need to be as well.
my $NORMAL = "";
my $DEPRECATED = 'D';
my $a_bold_deprecated = "a 'B<$DEPRECATED>'";
my $A_bold_deprecated = "A 'B<$DEPRECATED>'";
my $DISCOURAGED = 'X';
my $a_bold_discouraged = "an 'B<$DISCOURAGED>'";
my $A_bold_discouraged = "An 'B<$DISCOURAGED>'";
my $STRICTER = 'T';
my $a_bold_stricter = "a 'B<$STRICTER>'";
my $A_bold_stricter = "A 'B<$STRICTER>'";
my $STABILIZED = 'S';
my $a_bold_stabilized = "an 'B<$STABILIZED>'";
my $A_bold_stabilized = "An 'B<$STABILIZED>'";
my $OBSOLETE = 'O';
my $a_bold_obsolete = "an 'B<$OBSOLETE>'";
my $A_bold_obsolete = "An 'B<$OBSOLETE>'";

# Aliases can also have an extra status:
my $INTERNAL_ALIAS = 'P';

my %status_past_participles = (
    $DISCOURAGED => 'discouraged',
    $STABILIZED => 'stabilized',
    $OBSOLETE => 'obsolete',
    $DEPRECATED => 'deprecated',
    $INTERNAL_ALIAS => 'reserved for Perl core internal use only',
);

# Table fates.  These are somewhat ordered, so that fates < $MAP_PROXIED should be
# externally documented.
my $ORDINARY = 0;       # The normal fate.
my $MAP_PROXIED = 1;    # The map table for the property isn't written out,
                        # but there is a file written that can be used to
                        # reconstruct this table
my $INTERNAL_ONLY = 2;  # The file for this table is written out, but it is
                        # for Perl's internal use only
my $SUPPRESSED = 3;     # The file for this table is not written out, and as a
                        # result, we don't bother to do many computations on
                        # it.
my $PLACEHOLDER = 4;    # Like $SUPPRESSED, but we go through all the
                        # computations anyway, as the values are needed for
                        # things to work.  This happens when we have Perl
                        # extensions that depend on Unicode tables that
                        # wouldn't normally be in a given Unicode version.

# The format of the values of the tables:
my $EMPTY_FORMAT = "";
my $BINARY_FORMAT = 'b';
my $DECIMAL_FORMAT = 'd';
my $FLOAT_FORMAT = 'f';
my $INTEGER_FORMAT = 'i';
my $HEX_FORMAT = 'x';
my $RATIONAL_FORMAT = 'r';
my $STRING_FORMAT = 's';
my $ADJUST_FORMAT = 'a';
my $HEX_ADJUST_FORMAT = 'ax';
my $DECOMP_STRING_FORMAT = 'c';
my $STRING_WHITE_SPACE_LIST = 'sw';

my %map_table_formats = (
    $BINARY_FORMAT => 'binary',
    $DECIMAL_FORMAT => 'single decimal digit',
    $FLOAT_FORMAT => 'floating point number',
    $INTEGER_FORMAT => 'integer',
    $HEX_FORMAT => 'non-negative hex whole number; a code point',
    $RATIONAL_FORMAT => 'rational: an integer or a fraction',
    $STRING_FORMAT => 'string',
    $ADJUST_FORMAT => 'some entries need adjustment',
    $HEX_ADJUST_FORMAT => 'mapped value in hex; some entries need adjustment',
    $DECOMP_STRING_FORMAT => 'Perl\'s internal (Normalize.pm) decomposition mapping',
    $STRING_WHITE_SPACE_LIST => 'string, but some elements are interpreted as a list; white space occurs only as list item separators'
);

# Unicode didn't put such derived files in a separate directory at first.
my $EXTRACTED_DIR = (-d 'extracted') ? 'extracted' : "";
my $EXTRACTED = ($EXTRACTED_DIR) ? "$EXTRACTED_DIR/" : "";
my $AUXILIARY = 'auxiliary';
my $EMOJI = 'emoji';

# Hashes and arrays that will eventually go into UCD.pl for the use of UCD.pm
my %loose_to_file_of;       # loosely maps table names to their respective
                            # files
my %stricter_to_file_of;    # same; but for stricter mapping.
my %loose_property_to_file_of; # Maps a loose property name to its map file
my %strict_property_to_file_of; # Same, but strict
my @inline_definitions = "V0"; # Each element gives a definition of a unique
                            # inversion list.  When a definition is inlined,
                            # its value in the hash it's in (one of the two
                            # defined just above) will include an index into
                            # this array.  The 0th element is initialized to
                            # the definition for a zero length inversion list
my %file_to_swash_name;     # Maps the file name to its corresponding key name
                            # in the hash %Unicode::UCD::SwashInfo
my %nv_floating_to_rational; # maps numeric values floating point numbers to
                             # their rational equivalent
my %loose_property_name_of; # Loosely maps (non_string) property names to
                            # standard form
my %strict_property_name_of; # Strictly maps (non_string) property names to
                            # standard form
my %string_property_loose_to_name; # Same, for string properties.
my %loose_defaults;         # keys are of form "prop=value", where 'prop' is
                            # the property name in standard loose form, and
                            # 'value' is the default value for that property,
                            # also in standard loose form.
my %loose_to_standard_value; # loosely maps table names to the canonical
                            # alias for them
my %ambiguous_names;        # keys are alias names (in standard form) that
                            # have more than one possible meaning.
my %combination_property;   # keys are alias names (in standard form) that
                            # have both a map table, and a binary one that
                            # yields true for all non-null maps.
my %prop_aliases;           # Keys are standard property name; values are each
                            # one's aliases
my %prop_value_aliases;     # Keys of top level are standard property name;
                            # values are keys to another hash,  Each one is
                            # one of the property's values, in standard form.
                            # The values are that prop-val's aliases.
my %skipped_files;          # List of files that we skip
my %ucd_pod;    # Holds entries that will go into the UCD section of the pod

# Most properties are immune to caseless matching, otherwise you would get
# nonsensical results, as properties are a function of a code point, not
# everything that is caselessly equivalent to that code point.  For example,
# Changes_When_Case_Folded('s') should be false, whereas caselessly it would
# be true because 's' and 'S' are equivalent caselessly.  However,
# traditionally, [:upper:] and [:lower:] are equivalent caselessly, so we
# extend that concept to those very few properties that are like this.  Each
# such property will match the full range caselessly.  They are hard-coded in
# the program; it's not worth trying to make it general as it's extremely
# unlikely that they will ever change.
my %caseless_equivalent_to;

# This is the range of characters that were in Release 1 of Unicode, and
# removed in Release 2 (replaced with the current Hangul syllables starting at
# U+AC00).  The range was reused starting in Release 3 for other purposes.
my $FIRST_REMOVED_HANGUL_SYLLABLE = 0x3400;
my $FINAL_REMOVED_HANGUL_SYLLABLE = 0x4DFF;

# These constants names and values were taken from the Unicode standard,
# version 5.1, section 3.12.  They are used in conjunction with Hangul
# syllables.  The '_string' versions are so generated tables can retain the
# hex format, which is the more familiar value
my $SBase_string = "0xAC00";
my $SBase = CORE::hex $SBase_string;
my $LBase_string = "0x1100";
my $LBase = CORE::hex $LBase_string;
my $VBase_string = "0x1161";
my $VBase = CORE::hex $VBase_string;
my $TBase_string = "0x11A7";
my $TBase = CORE::hex $TBase_string;
my $SCount = 11172;
my $LCount = 19;
my $VCount = 21;
my $TCount = 28;
my $NCount = $VCount * $TCount;

# For Hangul syllables;  These store the numbers from Jamo.txt in conjunction
# with the above published constants.
my %Jamo;
my %Jamo_L;     # Leading consonants
my %Jamo_V;     # Vowels
my %Jamo_T;     # Trailing consonants

# For code points whose name contains its ordinal as a '-ABCD' suffix.
# The key is the base name of the code point, and the value is an
# array giving all the ranges that use this base name.  Each range
# is actually a hash giving the 'low' and 'high' values of it.
my %names_ending_in_code_point;
my %loose_names_ending_in_code_point;   # Same as above, but has blanks, dashes
                                        # removed from the names
# Inverse mapping.  The list of ranges that have these kinds of
# names.  Each element contains the low, high, and base names in an
# anonymous hash.
my @code_points_ending_in_code_point;

# To hold Unicode's normalization test suite
my @normalization_tests;

# Boolean: does this Unicode version have the hangul syllables, and are we
# writing out a table for them?
my $has_hangul_syllables = 0;

# Does this Unicode version have code points whose names end in their
# respective code points, and are we writing out a table for them?  0 for no;
# otherwise points to first property that a table is needed for them, so that
# if multiple tables are needed, we don't create duplicates
my $needing_code_points_ending_in_code_point = 0;

my @backslash_X_tests;     # List of tests read in for testing \X
my @LB_tests;              # List of tests read in for testing \b{lb}
my @SB_tests;              # List of tests read in for testing \b{sb}
my @WB_tests;              # List of tests read in for testing \b{wb}
my @unhandled_properties;  # Will contain a list of properties found in
                           # the input that we didn't process.
my @match_properties;      # Properties that have match tables, to be
                           # listed in the pod
my @map_properties;        # Properties that get map files written
my @named_sequences;       # NamedSequences.txt contents.
my %potential_files;       # Generated list of all .txt files in the directory
                           # structure so we can warn if something is being
                           # ignored.
my @missing_early_files;   # Generated list of absent files that we need to
                           # proceed in compiling this early Unicode version
my @files_actually_output; # List of files we generated.
my @more_Names;            # Some code point names are compound; this is used
                           # to store the extra components of them.
my $E_FLOAT_PRECISION = 3; # The minimum number of digits after the decimal
                           # point of a normalized floating point number
                           # needed to match before we consider it equivalent
                           # to a candidate rational

# These store references to certain commonly used property objects
my $age;
my $ccc;
my $gc;
my $perl;
my $block;
my $perl_charname;
my $print;
my $All;
my $Assigned;   # All assigned characters in this Unicode release
my $DI;         # Default_Ignorable_Code_Point property
my $NChar;      # Noncharacter_Code_Point property
my $script;
my $scx;        # Script_Extensions property
my $idt;        # Identifier_Type property

# Are there conflicting names because of beginning with 'In_', or 'Is_'
my $has_In_conflicts = 0;
my $has_Is_conflicts = 0;

sub internal_file_to_platform ($file=undef) {
    # Convert our file paths which have '/' separators to those of the
    # platform.

    return undef unless defined $file;

    return File::Spec->join(split '/', $file);
}

sub file_exists ($file=undef) {   # platform independent '-e'.  This program internally
                        # uses slash as a path separator.
    return 0 unless defined $file;
    return -e internal_file_to_platform($file);
}

sub objaddr($addr) {
    # Returns the address of the blessed input object.
    # It doesn't check for blessedness because that would do a string eval
    # every call, and the program is structured so that this is never called
    # for a non-blessed object.

    return pack 'J', refaddr $addr;
}

# These are used only if $annotate is true.
# The entire range of Unicode characters is examined to populate these
# after all the input has been processed.  But most can be skipped, as they
# have the same descriptive phrases, such as being unassigned
my @viacode;            # Contains the 1 million character names
my @age;                # And their ages ("" if none)
my @printable;          # boolean: And are those characters printable?
my @annotate_char_type; # Contains a type of those characters, specifically
                        # for the purposes of annotation.
my $annotate_ranges;    # A map of ranges of code points that have the same
                        # name for the purposes of annotation.  They map to the
                        # upper edge of the range, so that the end point can
                        # be immediately found.  This is used to skip ahead to
                        # the end of a range, and avoid processing each
                        # individual code point in it.
my $unassigned_sans_noncharacters; # A Range_List of the unassigned
                                   # characters, but excluding those which are
                                   # also noncharacter code points

# The annotation types are an extension of the regular range types, though
# some of the latter are folded into one.  Make the new types negative to
# avoid conflicting with the regular types
my $SURROGATE_TYPE = -1;
my $UNASSIGNED_TYPE = -2;
my $PRIVATE_USE_TYPE = -3;
my $NONCHARACTER_TYPE = -4;
my $CONTROL_TYPE = -5;
my $ABOVE_UNICODE_TYPE = -6;
my $UNKNOWN_TYPE = -7;  # Used only if there is a bug in this program

sub populate_char_info ($i) {
    # Used only with the $annotate option.  Populates the arrays with the
    # input code point's info that are needed for outputting more detailed
    # comments.  If calling context wants a return, it is the end point of
    # any contiguous range of characters that share essentially the same info

    $viacode[$i] = $perl_charname->value_of($i) || "";
    $age[$i] = (defined $age)
               ? (($age->value_of($i) =~ / ^ \d+ \. \d+ $ /x)
                  ? $age->value_of($i)
                  : "")
               : "";

    # A character is generally printable if Unicode says it is,
    # but below we make sure that most Unicode general category 'C' types
    # aren't.
    $printable[$i] = $print->contains($i);

    # But the characters in this range were removed in v2.0 and replaced by
    # different ones later.  Modern fonts will be for the replacement
    # characters, so suppress printing them.
    if (($v_version lt v2.0
         || ($compare_versions && $compare_versions lt v2.0))
        && (   $i >= $FIRST_REMOVED_HANGUL_SYLLABLE
            && $i <= $FINAL_REMOVED_HANGUL_SYLLABLE))
    {
        $printable[$i] = 0;
    }

    $annotate_char_type[$i] = $perl_charname->type_of($i) || 0;

    # Only these two regular types are treated specially for annotations
    # purposes
    $annotate_char_type[$i] = 0 if $annotate_char_type[$i] != $CP_IN_NAME
                                && $annotate_char_type[$i] != $HANGUL_SYLLABLE;

    # Give a generic name to all code points that don't have a real name.
    # We output ranges, if applicable, for these.  Also calculate the end
    # point of the range.
    my $end;
    if (! $viacode[$i]) {
        if ($i > $MAX_UNICODE_CODEPOINT) {
            $viacode[$i] = 'Above-Unicode';
            $annotate_char_type[$i] = $ABOVE_UNICODE_TYPE;
            $printable[$i] = 0;
            $end = $MAX_WORKING_CODEPOINT;
        }
        elsif ($gc-> table('Private_use')->contains($i)) {
            $viacode[$i] = 'Private Use';
            $annotate_char_type[$i] = $PRIVATE_USE_TYPE;
            $printable[$i] = 0;
            $end = $gc->table('Private_Use')->containing_range($i)->end;
        }
        elsif ($NChar->contains($i)) {
            $viacode[$i] = 'Noncharacter';
            $annotate_char_type[$i] = $NONCHARACTER_TYPE;
            $printable[$i] = 0;
            $end = $NChar->containing_range($i)->end;
        }
        elsif ($gc-> table('Control')->contains($i)) {
            my $name_ref = property_ref('Name_Alias');
            $name_ref = property_ref('Unicode_1_Name') if ! defined $name_ref;
            $viacode[$i] = (defined $name_ref)
                           ? $name_ref->value_of($i)
                           : 'Control';
            $annotate_char_type[$i] = $CONTROL_TYPE;
            $printable[$i] = 0;
        }
        elsif ($gc-> table('Unassigned')->contains($i)) {
            $annotate_char_type[$i] = $UNASSIGNED_TYPE;
            $printable[$i] = 0;
            $viacode[$i] = 'Unassigned';

            if (defined $block) { # No blocks in earliest releases
                $viacode[$i] .= ', block=' . $block-> value_of($i);
                $end = $gc-> table('Unassigned')->containing_range($i)->end;

                # Because we name the unassigned by the blocks they are in, it
                # can't go past the end of that block, and it also can't go
                # past the unassigned range it is in.  The special table makes
                # sure that the non-characters, which are unassigned, are
                # separated out.
                $end = min($block->containing_range($i)->end,
                           $unassigned_sans_noncharacters->
                                                    containing_range($i)->end);
            }
            else {
                $end = $i + 1;
                while ($unassigned_sans_noncharacters->contains($end)) {
                    $end++;
                }
                $end--;
            }
        }
        elsif ($perl->table('_Perl_Surrogate')->contains($i)) {
            $viacode[$i] = 'Surrogate';
            $annotate_char_type[$i] = $SURROGATE_TYPE;
            $printable[$i] = 0;
            $end = $gc->table('Surrogate')->containing_range($i)->end;
        }
        else {
            Carp::my_carp_bug("Can't figure out how to annotate "
                              . sprintf("U+%04X", $i)
                              . ".  Proceeding anyway.");
            $viacode[$i] = 'UNKNOWN';
            $annotate_char_type[$i] = $UNKNOWN_TYPE;
            $printable[$i] = 0;
        }
    }

    # Here, has a name, but if it's one in which the code point number is
    # appended to the name, do that.
    elsif ($annotate_char_type[$i] == $CP_IN_NAME) {
        $viacode[$i] .= sprintf("-%04X", $i);

        my $limit = $perl_charname->containing_range($i)->end;
        if (defined $age) {
            # Do all these as groups of the same age, instead of individually,
            # because their names are so meaningless, and there are typically
            # large quantities of them.
            $end = $i + 1;
            while ($end <= $limit && $age->value_of($end) == $age[$i]) {
                $end++;
            }
            $end--;
        }
        else {
            $end = $limit;
        }
    }

    # And here, has a name, but if it's a hangul syllable one, replace it with
    # the correct name from the Unicode algorithm
    elsif ($annotate_char_type[$i] == $HANGUL_SYLLABLE) {
        use integer;
        my $SIndex = $i - $SBase;
        my $L = $LBase + $SIndex / $NCount;
        my $V = $VBase + ($SIndex % $NCount) / $TCount;
        my $T = $TBase + $SIndex % $TCount;
        $viacode[$i] = "HANGUL SYLLABLE $Jamo{$L}$Jamo{$V}";
        $viacode[$i] .= $Jamo{$T} if $T != $TBase;
        $end = $perl_charname->containing_range($i)->end;
    }

    return if ! defined wantarray;
    return $i if ! defined $end;    # If not a range, return the input

    # Save this whole range so can find the end point quickly
    $annotate_ranges->add_map($i, $end, $end);

    return $end;
}

sub max($a, $b) {
    return $a >= $b ? $a : $b;
}

sub min($a, $b) {
    return $a <= $b ? $a : $b;
}

sub clarify_number ($number) {
    # This returns the input number with underscores inserted every 3 digits
    # in large (5 digits or more) numbers.  Input must be entirely digits, not
    # checked.

    my $pos = length($number) - 3;
    return $number if $pos <= 1;
    while ($pos > 0) {
        substr($number, $pos, 0) = '_';
        $pos -= 3;
    }
    return $number;
}

sub clarify_code_point_count ($number) {
    # This is like clarify_number(), but the input is assumed to be a count of
    # code points, rather than a generic number.

    my $append = "";

    if ($number > $MAX_UNICODE_CODEPOINTS) {
        $number -= ($MAX_WORKING_CODEPOINTS - $MAX_UNICODE_CODEPOINTS);
        return "All above-Unicode code points" if $number == 0;
        $append = " + all above-Unicode code points";
    }
    return clarify_number($number) . $append;
}

package Carp;

# These routines give a uniform treatment of messages in this program.  They
# are placed in the Carp package to cause the stack trace to not include them,
# although an alternative would be to use another package and set @CARP_NOT
# for it.

our $Verbose = 1 if main::DEBUG;  # Useful info when debugging

# This is a work-around suggested by Nicholas Clark to fix a problem with Carp
# and overload trying to load Scalar:Util under miniperl.  See
# http://www.xray.mpe.mpg.de/mailing-lists/perl5-porters/2009-11/msg01057.html
undef $overload::VERSION;

sub my_carp($message="", $nofold=0) {

    if ($message) {
        $message = main::join_lines($message);
        $message =~ s/^$0: *//;     # Remove initial program name
        $message =~ s/[.;,]+$//;    # Remove certain ending punctuation
        $message = "\n$0: $message;";

        # Fold the message with program name, semi-colon end punctuation
        # (which looks good with the message that carp appends to it), and a
        # hanging indent for continuation lines.
        $message = main::simple_fold($message, "", 4) unless $nofold;
        $message =~ s/\n$//;        # Remove the trailing nl so what carp
                                    # appends is to the same line
    }

    return $message if defined wantarray;   # If a caller just wants the msg

    carp $message;
    return;
}

sub my_carp_bug($message="") {
    # This is called when it is clear that the problem is caused by a bug in
    # this program.
    $message =~ s/^$0: *//;
    $message = my_carp("Bug in $0.  Please report it by running perlbug or if that is unavailable, by sending email to perbug\@perl.org:\n$message");
    carp $message;
    return;
}

sub carp_too_few_args($args_ref, $count) {
    my_carp_bug("Need at least $count arguments to "
        . (caller 1)[3]
        . ".  Instead got: '"
        . join ', ', @$args_ref
        . "'.  No action taken.");
    return;
}

sub carp_extra_args($args_ref) {
    unless (ref $args_ref) {
        my_carp_bug("Argument to 'carp_extra_args' ($args_ref) must be a ref.  Not checking arguments.");
        return;
    }
    my ($package, $file, $line) = caller;
    my $subroutine = (caller 1)[3];

    my $list;
    if (ref $args_ref eq 'HASH') {
        foreach my $key (keys %$args_ref) {
            $args_ref->{$key} = $UNDEF unless defined $args_ref->{$key};
        }
        $list = join ', ', each %{$args_ref};
    }
    elsif (ref $args_ref eq 'ARRAY') {
        foreach my $arg (@$args_ref) {
            $arg = $UNDEF unless defined $arg;
        }
        $list = join ', ', @$args_ref;
    }
    else {
        my_carp_bug("Can't cope with ref "
                . ref($args_ref)
                . " . argument to 'carp_extra_args'.  Not checking arguments.");
        return;
    }

    my_carp_bug("Unrecognized parameters in options: '$list' to $subroutine.  Skipped.");
    return;
}

package main;

{ # Closure

    # This program uses the inside-out method for objects, as recommended in
    # "Perl Best Practices".  (This is the best solution still, since this has
    # to run under miniperl.)  This closure aids in generating those.  There
    # are two routines.  setup_package() is called once per package to set
    # things up, and then set_access() is called for each hash representing a
    # field in the object.  These routines arrange for the object to be
    # properly destroyed when no longer used, and for standard accessor
    # functions to be generated.  If you need more complex accessors, just
    # write your own and leave those accesses out of the call to set_access().
    # More details below.

    my %constructor_fields; # fields that are to be used in constructors; see
                            # below

    # The values of this hash will be the package names as keys to other
    # hashes containing the name of each field in the package as keys, and
    # references to their respective hashes as values.
    my %package_fields;

    sub setup_package {
        # Sets up the package, creating standard DESTROY and dump methods
        # (unless already defined).  The dump method is used in debugging by
        # simple_dumper().
        # The optional parameters are:
        #   a)  a reference to a hash, that gets populated by later
        #       set_access() calls with one of the accesses being
        #       'constructor'.  The caller can then refer to this, but it is
        #       not otherwise used by these two routines.
        #   b)  a reference to a callback routine to call during destruction
        #       of the object, before any fields are actually destroyed

        my %args = @_;
        my $constructor_ref = delete $args{'Constructor_Fields'};
        my $destroy_callback = delete $args{'Destroy_Callback'};
        Carp::carp_extra_args(\@_) if main::DEBUG && %args;

        my %fields;
        my $package = (caller)[0];

        $package_fields{$package} = \%fields;
        $constructor_fields{$package} = $constructor_ref;

        unless ($package->can('DESTROY')) {
            my $destroy_name = "${package}::DESTROY";
            no strict "refs";

            # Use typeglob to give the anonymous subroutine the name we want
            *$destroy_name = sub ($self) {
                my $addr = pack 'J', refaddr $self;

                $self->$destroy_callback if $destroy_callback;
                foreach my $field (keys %{$package_fields{$package}}) {
                    #print STDERR __LINE__, ": Destroying ", ref $self, " ", sprintf("%04X", $addr), ": ", $field, "\n";
                    delete $package_fields{$package}{$field}{$addr};
                }
                return;
            }
        }

        unless ($package->can('dump')) {
            my $dump_name = "${package}::dump";
            no strict "refs";
            *$dump_name = sub ($self, @_args) {
                return dump_inside_out($self, $package_fields{$package}, @_args);
            }
        }
        return;
    }

    sub set_access($name, $field, @accessors) {
        # Arrange for the input field to be garbage collected when no longer
        # needed.  Also, creates standard accessor functions for the field
        # based on the optional parameters-- none if none of these parameters:
        #   'addable'    creates an 'add_NAME()' accessor function.
        #   'readable' or 'readable_array'   creates a 'NAME()' accessor
        #                function.
        #   'settable'   creates a 'set_NAME()' accessor function.
        #   'constructor' doesn't create an accessor function, but adds the
        #                field to the hash that was previously passed to
        #                setup_package();
        # Any of the accesses can be abbreviated down, so that 'a', 'ad',
        # 'add' etc. all mean 'addable'.
        # The read accessor function will work on both array and scalar
        # values.  If another accessor in the parameter list is 'a', the read
        # access assumes an array.  You can also force it to be array access
        # by specifying 'readable_array' instead of 'readable'
        #
        # A sort-of 'protected' access can be set-up by preceding the addable,
        # readable or settable with some initial portion of 'protected_' (but,
        # the underscore is required), like 'p_a', 'pro_set', etc.  The
        # "protection" is only by convention.  All that happens is that the
        # accessor functions' names begin with an underscore.  So instead of
        # calling set_foo, the call is _set_foo.  (Real protection could be
        # accomplished by having a new subroutine, end_package, called at the
        # end of each package, and then storing the __LINE__ ranges and
        # checking them on every accessor.  But that is way overkill.)

        # We create anonymous subroutines as the accessors and then use
        # typeglobs to assign them to the proper package and name

        # $name 	Name of the field
        # $field 	Reference to the inside-out hash containing the
		# 			field

        my $package = (caller)[0];

        if (! exists $package_fields{$package}) {
            croak "$0: Must call 'setup_package' before 'set_access'";
        }

        # Stash the field so DESTROY can get it.
        $package_fields{$package}{$name} = $field;

        # Remaining arguments are the accessors.  For each...
        foreach my $access (@accessors) {
            my $access = lc $access;

            my $protected = "";

            # Match the input as far as it goes.
            if ($access =~ /^(p[^_]*)_/) {
                $protected = $1;
                if (substr('protected_', 0, length $protected)
                    eq $protected)
                {

                    # Add 1 for the underscore not included in $protected
                    $access = substr($access, length($protected) + 1);
                    $protected = '_';
                }
                else {
                    $protected = "";
                }
            }

            if (substr('addable', 0, length $access) eq $access) {
                my $subname = "${package}::${protected}add_$name";
                no strict "refs";

                # add_ accessor.  Don't add if already there, which we
                # determine using 'eq' for scalars and '==' otherwise.
                *$subname = sub ($self, $value) {
                    use strict "refs";
                    my $addr = pack 'J', refaddr $self;
                    if (ref $value) {
                        return if grep { $value == $_ } @{$field->{$addr}};
                    }
                    else {
                        return if grep { $value eq $_ } @{$field->{$addr}};
                    }
                    push @{$field->{$addr}}, $value;
                    return;
                }
            }
            elsif (substr('constructor', 0, length $access) eq $access) {
                if ($protected) {
                    Carp::my_carp_bug("Can't set-up 'protected' constructors")
                }
                else {
                    $constructor_fields{$package}{$name} = $field;
                }
            }
            elsif (substr('readable_array', 0, length $access) eq $access) {

                # Here has read access.  If one of the other parameters for
                # access is array, or this one specifies array (by being more
                # than just 'readable_'), then create a subroutine that
                # assumes the data is an array.  Otherwise just a scalar
                my $subname = "${package}::${protected}$name";
                if (grep { /^a/i } (@accessors)
                    or length($access) > length('readable_'))
                {
                    no strict "refs";
                    *$subname = sub ($_addr) {
                        use strict "refs";
                        my $addr = pack 'J', refaddr $_addr;
                        if (ref $field->{$addr} ne 'ARRAY') {
                            my $type = ref $field->{$addr};
                            $type = 'scalar' unless $type;
                            Carp::my_carp_bug("Trying to read $name as an array when it is a $type.  Big problems.");
                            return;
                        }
                        return scalar @{$field->{$addr}} unless wantarray;

                        # Make a copy; had problems with caller modifying the
                        # original otherwise
                        my @return = @{$field->{$addr}};
                        return @return;
                    }
                }
                else {

                    # Here not an array value, a simpler function.
                    no strict "refs";
                    *$subname = sub ($addr) {
                        use strict "refs";
                        return $field->{pack 'J', refaddr $addr};
                    }
                }
            }
            elsif (substr('settable', 0, length $access) eq $access) {
                my $subname = "${package}::${protected}set_$name";
                no strict "refs";
                *$subname = sub ($self, $value) {
                    use strict "refs";
                    # $self is $_[0]; $value is $_[1]
                    $field->{pack 'J', refaddr $self} = $value;
                    return;
                }
            }
            else {
                Carp::my_carp_bug("Unknown accessor type $access.  No accessor set.");
            }
        }
        return;
    }
}

package Input_file;

# All input files use this object, which stores various attributes about them,
# and provides for convenient, uniform handling.  The run method wraps the
# processing.  It handles all the bookkeeping of opening, reading, and closing
# the file, returning only significant input lines.
#
# Each object gets a handler which processes the body of the file, and is
# called by run().  All character property files must use the generic,
# default handler, which has code scrubbed to handle things you might not
# expect, including automatic EBCDIC handling.  For files that don't deal with
# mapping code points to a property value, such as test files,
# PropertyAliases, PropValueAliases, and named sequences, you can override the
# handler to be a custom one.  Such a handler should basically be a
# while(next_line()) {...} loop.
#
# You can also set up handlers to
#   0) call during object construction time, after everything else is done
#   1) call before the first line is read, for pre processing
#   2) call to adjust each line of the input before the main handler gets
#      them.  This can be automatically generated, if appropriately simple
#      enough, by specifying a Properties parameter in the constructor.
#   3) call upon EOF before the main handler exits its loop
#   4) call at the end, for post processing
#
# $_ is used to store the input line, and is to be filtered by the
# each_line_handler()s.  So, if the format of the line is not in the desired
# format for the main handler, these are used to do that adjusting.  They can
# be stacked (by enclosing them in an [ anonymous array ] in the constructor,
# so the $_ output of one is used as the input to the next.  The EOF handler
# is also stackable, but none of the others are, but could easily be changed
# to be so.
#
# Some properties are used by the Perl core but aren't defined until later
# Unicode releases.  The perl interpreter would have problems working when
# compiled with an earlier Unicode version that doesn't have them, so we need
# to define them somehow for those releases.  The 'Early' constructor
# parameter can be used to automatically handle this.  It is essentially
# ignored if the Unicode version being compiled has a data file for this
# property.  Either code to execute or a file to read can be specified.
# Details are at the %early definition.
#
# Most of the handlers can call insert_lines() or insert_adjusted_lines()
# which insert the parameters as lines to be processed before the next input
# file line is read.  This allows the EOF handler(s) to flush buffers, for
# example.  The difference between the two routines is that the lines inserted
# by insert_lines() are subjected to the each_line_handler()s.  (So if you
# called it from such a handler, you would get infinite recursion without some
# mechanism to prevent that.)  Lines inserted by insert_adjusted_lines() go
# directly to the main handler without any adjustments.  If the
# post-processing handler calls any of these, there will be no effect.  Some
# error checking for these conditions could be added, but it hasn't been done.
#
# carp_bad_line() should be called to warn of bad input lines, which clears $_
# to prevent further processing of the line.  This routine will output the
# message as a warning once, and then keep a count of the lines that have the
# same message, and output that count at the end of the file's processing.
# This keeps the number of messages down to a manageable amount.
#
# get_missings() should be called to retrieve any @missing input lines.
# Messages will be raised if this isn't done if the options aren't to ignore
# missings.

sub trace { return main::trace(@_); }

{ # Closure
    # Keep track of fields that are to be put into the constructor.
    my %constructor_fields;

    main::setup_package(Constructor_Fields => \%constructor_fields);

    my %file; # Input file name, required
    main::set_access('file', \%file, qw{ c r });

    my %first_released; # Unicode version file was first released in, required
    main::set_access('first_released', \%first_released, qw{ c r });

    my %handler;    # Subroutine to process the input file, defaults to
                    # 'process_generic_property_file'
    main::set_access('handler', \%handler, qw{ c });

    my %property;
    # name of property this file is for.  defaults to none, meaning not
    # applicable, or is otherwise determinable, for example, from each line.
    main::set_access('property', \%property, qw{ c r });

    my %optional;
    # This is either an unsigned number, or a list of property names.  In the
    # former case, if it is non-zero, it means the file is optional, so if the
    # file is absent, no warning about that is output.  In the latter case, it
    # is a list of properties that the file (exclusively) defines.  If the
    # file is present, tables for those properties will be produced; if
    # absent, none will, even if they are listed elsewhere (namely
    # PropertyAliases.txt and PropValueAliases.txt) as being in this release,
    # and no warnings will be raised about them not being available.  (And no
    # warning about the file itself will be raised.)
    main::set_access('optional', \%optional, qw{ c readable_array } );

    my %non_skip;
    # This is used for debugging, to skip processing of all but a few input
    # files.  Add 'non_skip => 1' to the constructor for those files you want
    # processed when you set the $debug_skip global.
    main::set_access('non_skip', \%non_skip, 'c');

    my %skip;
    # This is used to skip processing of this input file (semi-) permanently.
    # The value should be the reason the file is being skipped.  It is used
    # for files that we aren't planning to process anytime soon, but want to
    # allow to be in the directory and be checked for their names not
    # conflicting with any other files on a DOS 8.3 name filesystem, but to
    # not otherwise be processed, and to not raise a warning about not being
    # handled.  In the constructor call, any value that evaluates to a numeric
    # 0 or undef means don't skip.  Any other value is a string giving the
    # reason it is being skipped, and this will appear in generated pod.
    # However, an empty string reason will suppress the pod entry.
    # Internally, calls that evaluate to numeric 0 are changed into undef to
    # distinguish them from an empty string call.
    main::set_access('skip', \%skip, 'c', 'r');

    my %each_line_handler;
    # list of subroutines to look at and filter each non-comment line in the
    # file.  defaults to none.  The subroutines are called in order, each is
    # to adjust $_ for the next one, and the final one adjusts it for
    # 'handler'
    main::set_access('each_line_handler', \%each_line_handler, 'c');

    my %retain_trailing_comments;
    # This is used to not discard the comments that end data lines.  This
    # would be used only for files with non-typical syntax, and most code here
    # assumes that comments have been stripped, so special handlers would have
    # to be written.  It is assumed that the code will use these in
    # single-quoted contexts, and so any "'" marks in the comment will be
    # prefixed by a backslash.
    main::set_access('retain_trailing_comments', \%retain_trailing_comments, 'c');

    my %properties; # Optional ordered list of the properties that occur in each
    # meaningful line of the input file.  If present, an appropriate
    # each_line_handler() is automatically generated and pushed onto the stack
    # of such handlers.  This is useful when a file contains multiple
    # properties per line, but no other special considerations are necessary.
    # The special value "<ignored>" means to discard the corresponding input
    # field.
    # Any @missing lines in the file should also match this syntax; no such
    # files exist as of 6.3.  But if it happens in a future release, the code
    # could be expanded to properly parse them.
    main::set_access('properties', \%properties, qw{ c r });

    my %has_missings_defaults;
    # ? Are there lines in the file giving default values for code points
    # missing from it?.  Defaults to NO_DEFAULTS.  Otherwise NOT_IGNORED is
    # the norm, but IGNORED means it has such lines, but the handler doesn't
    # use them.  Having these three states allows us to catch changes to the
    # UCD that this program should track.  XXX This could be expanded to
    # specify the syntax for such lines, like %properties above.
    main::set_access('has_missings_defaults',
                                        \%has_missings_defaults, qw{ c r });

    my %construction_time_handler;
    # Subroutine to call at the end of the new method.  If undef, no such
    # handler is called.
    main::set_access('construction_time_handler',
                                        \%construction_time_handler, qw{ c });

    my %pre_handler;
    # Subroutine to call before doing anything else in the file.  If undef, no
    # such handler is called.
    main::set_access('pre_handler', \%pre_handler, qw{ c });

    my %eof_handler;
    # Subroutines to call upon getting an EOF on the input file, but before
    # that is returned to the main handler.  This is to allow buffers to be
    # flushed.  The handler is expected to call insert_lines() or
    # insert_adjusted() with the buffered material
    main::set_access('eof_handler', \%eof_handler, qw{ c });

    my %post_handler;
    # Subroutine to call after all the lines of the file are read in and
    # processed.  If undef, no such handler is called.  Note that this cannot
    # add lines to be processed; instead use eof_handler
    main::set_access('post_handler', \%post_handler, qw{ c });

    my %progress_message;
    # Message to print to display progress in lieu of the standard one
    main::set_access('progress_message', \%progress_message, qw{ c });

    my %handle;
    # cache open file handle, internal.  Is undef if file hasn't been
    # processed at all, empty if has;
    main::set_access('handle', \%handle);

    my %added_lines;
    # cache of lines added virtually to the file, internal
    main::set_access('added_lines', \%added_lines);

    my %remapped_lines;
    # cache of lines added virtually to the file, internal
    main::set_access('remapped_lines', \%remapped_lines);

    my %errors;
    # cache of errors found, internal
    main::set_access('errors', \%errors);

    my %missings;
    # storage of '@missing' defaults lines
    main::set_access('missings', \%missings);

    my %early;
    # Used for properties that must be defined (for Perl's purposes) on
    # versions of Unicode earlier than Unicode itself defines them.  The
    # parameter is an array (it would be better to be a hash, but not worth
    # bothering about due to its rare use).
    #
    # The first element is either a code reference to call when in a release
    # earlier than the Unicode file is available in, or it is an alternate
    # file to use instead of the non-existent one.  This file must have been
    # plunked down in the same directory as mktables.  Should you be compiling
    # on a release that needs such a file, mktables will abort the
    # compilation, and tell you where to get the necessary file(s), and what
    # name(s) to use to store them as.
    # In the case of specifying an alternate file, the array must contain two
    # further elements:
    #
    # [1] is the name of the property that will be generated by this file.
    # The class automatically takes the input file and excludes any code
    # points in it that were not assigned in the Unicode version being
    # compiled.  It then uses this result to define the property in the given
    # version.  Since the property doesn't actually exist in the Unicode
    # version being compiled, this should be a name accessible only by core
    # perl.  If it is the same name as the regular property, the constructor
    # will mark the output table as a $PLACEHOLDER so that it doesn't actually
    # get output, and so will be unusable by non-core code.  Otherwise it gets
    # marked as $INTERNAL_ONLY.
    #
    # [2] is a property value to assign (only when compiling Unicode 1.1.5) to
    # the Hangul syllables in that release (which were ripped out in version
    # 2) for the given property .  (Hence it is ignored except when compiling
    # version 1.  You only get one value that applies to all of them, which
    # may not be the actual reality, but probably nobody cares anyway for
    # these obsolete characters.)
    #
    # [3] if present is the default value for the property to assign for code
    # points not given in the input.  If not present, the default from the
    # normal property is used
    #
    # [-1] If there is an extra final element that is the string 'ONLY_EARLY'.
    # it means to not add the name in [1] as an alias to the property name
    # used for these.  Normally, when compiling Unicode versions that don't
    # invoke the early handling, the name is added as a synonym.
    #
    # Not all files can be handled in the above way, and so the code ref
    # alternative is available.  It can do whatever it needs to.  The other
    # array elements are optional in this case, and the code is free to use or
    # ignore them if they are present.
    #
    # Internally, the constructor unshifts a 0 or 1 onto this array to
    # indicate if an early alternative is actually being used or not.  This
    # makes for easier testing later on.
    main::set_access('early', \%early, 'c');

    my %only_early;
    main::set_access('only_early', \%only_early, 'c');

    my %required_even_in_debug_skip;
    # debug_skip is used to speed up compilation during debugging by skipping
    # processing files that are not needed for the task at hand.  However,
    # some files pretty much can never be skipped, and this is used to specify
    # that this is one of them.  In order to skip this file, the call to the
    # constructor must be edited to comment out this parameter.
    main::set_access('required_even_in_debug_skip',
                     \%required_even_in_debug_skip, 'c');

    my %withdrawn;
    # Some files get removed from the Unicode DB.  This is a version object
    # giving the first release without this file.
    main::set_access('withdrawn', \%withdrawn, 'c');

    my %ucd;
    # Some files are not actually part of the Unicode Character Database.
    # These typically have a different way of indicating their version
    main::set_access('ucd', \%ucd, 'c');

    my %in_this_release;
    # Calculated value from %first_released and %withdrawn.  Are we compiling
    # a Unicode release which includes this file?
    main::set_access('in_this_release', \%in_this_release);

    sub _next_line;
    sub _next_line_with_remapped_range;

    sub new {
        my $class = shift;

        my $self = bless \do{ my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        # Set defaults
        $handler{$addr} = \&main::process_generic_property_file;
        $retain_trailing_comments{$addr} = 0;
        $non_skip{$addr} = 0;
        $skip{$addr} = undef;
        $has_missings_defaults{$addr} = $NO_DEFAULTS;
        $handle{$addr} = undef;
        $added_lines{$addr} = [ ];
        $remapped_lines{$addr} = [ ];
        $each_line_handler{$addr} = [ ];
        $eof_handler{$addr} = [ ];
        $errors{$addr} = { };
        $missings{$addr} = [ ];
        $early{$addr} = [ ];
        $optional{$addr} = [ ];
        $ucd{$addr} = 1;

        # Two positional parameters.
        return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;
        $file{$addr} = main::internal_file_to_platform(shift);
        $first_released{$addr} = shift;

        # The rest of the arguments are key => value pairs
        # %constructor_fields has been set up earlier to list all possible
        # ones.  Either set or push, depending on how the default has been set
        # up just above.
        my %args = @_;
        foreach my $key (keys %args) {
            my $argument = $args{$key};

            # Note that the fields are the lower case of the constructor keys
            my $hash = $constructor_fields{lc $key};
            if (! defined $hash) {
                Carp::my_carp_bug("Unrecognized parameters '$key => $argument' to new() for $self.  Skipped");
                next;
            }
            if (ref $hash->{$addr} eq 'ARRAY') {
                if (ref $argument eq 'ARRAY') {
                    foreach my $argument (@{$argument}) {
                        next if ! defined $argument;
                        push @{$hash->{$addr}}, $argument;
                    }
                }
                else {
                    push @{$hash->{$addr}}, $argument if defined $argument;
                }
            }
            else {
                $hash->{$addr} = $argument;
            }
            delete $args{$key};
        };

        $non_skip{$addr} = 1 if $required_even_in_debug_skip{$addr};

        # Convert 0 (meaning don't skip) to undef
        undef $skip{$addr} unless $skip{$addr};

        # Handle the case where this file is optional
        my $pod_message_for_non_existent_optional = "";
        if ($optional{$addr}->@*) {

            # First element is the pod message
            $pod_message_for_non_existent_optional
                                                = shift $optional{$addr}->@*;
            # Convert a 0 'Optional' argument to an empty list to make later
            # code more concise.
            if (   $optional{$addr}->@*
                && $optional{$addr}->@* == 1
                && $optional{$addr}[0] ne ""
                && $optional{$addr}[0] !~ /\D/
                && $optional{$addr}[0] == 0)
            {
                $optional{$addr} = [ ];
            }
            else {  # But if the only element doesn't evaluate to 0, make sure
                    # that this file is indeed considered optional below.
                unshift $optional{$addr}->@*, 1;
            }
        }

        my $progress;
        my $function_instead_of_file = 0;

        if ($early{$addr}->@* && $early{$addr}[-1] eq 'ONLY_EARLY') {
            $only_early{$addr} = 1;
            pop $early{$addr}->@*;
        }

        # If we are compiling a Unicode release earlier than the file became
        # available, the constructor may have supplied a substitute
        if ($first_released{$addr} gt $v_version && $early{$addr}->@*) {

            # Yes, we have a substitute, that we will use; mark it so
            unshift $early{$addr}->@*, 1;

            # See the definition of %early for what the array elements mean.
            # Note that we have just unshifted onto the array, so the numbers
            # below are +1 of those in the %early description.
            # If we have a property this defines, create a table and default
            # map for it now (at essentially compile time), so that it will be
            # available for the whole of run time.  (We will want to add this
            # name as an alias when we are using the official property name;
            # but this must be deferred until run(), because at construction
            # time the official names have yet to be defined.)
            if ($early{$addr}[2]) {
                my $fate = ($property{$addr}
                            && $property{$addr} eq $early{$addr}[2])
                          ? $PLACEHOLDER
                          : $INTERNAL_ONLY;
                my $prop_object = Property->new($early{$addr}[2],
                                                Fate => $fate,
                                                Perl_Extension => 1,
                                                );

                # If not specified by the constructor, use the default mapping
                # for the regular property for this substitute one.
                if ($early{$addr}[4]) {
                    $prop_object->set_default_map($early{$addr}[4]);
                }
                elsif (    defined $property{$addr}
                       &&  defined $default_mapping{$property{$addr}})
                {
                    $prop_object
                        ->set_default_map($default_mapping{$property{$addr}});
                }
            }

            if (ref $early{$addr}[1] eq 'CODE') {
                $function_instead_of_file = 1;

                # If the first element of the array is a code ref, the others
                # are optional.
                $handler{$addr} = $early{$addr}[1];
                $property{$addr} = $early{$addr}[2]
                                                if defined $early{$addr}[2];
                $progress = "substitute $file{$addr}";

                undef $file{$addr};
            }
            else {  # Specifying a substitute file

                if (! main::file_exists($early{$addr}[1])) {

                    # If we don't see the substitute file, generate an error
                    # message giving the needed things, and add it to the list
                    # of such to output before actual processing happens
                    # (hence the user finds out all of them in one run).
                    # Instead of creating a general method for NameAliases,
                    # hard-code it here, as there is unlikely to ever be a
                    # second one which needs special handling.
                    my $string_version = ($file{$addr} eq "NameAliases.txt")
                                    ? 'at least 6.1 (the later, the better)'
                                    : sprintf "%vd", $first_released{$addr};
                    push @missing_early_files, <<END;
'$file{$addr}' version $string_version should be copied to '$early{$addr}[1]'.
END
                    ;
                    return;
                }
                $progress = $early{$addr}[1];
                $progress .= ", substituting for $file{$addr}" if $file{$addr};
                $file{$addr} = $early{$addr}[1];
                $property{$addr} = $early{$addr}[2];

                # Ignore code points not in the version being compiled
                push $each_line_handler{$addr}->@*, \&_exclude_unassigned;

                if (   $v_version lt v2.0        # Hanguls in this release ...
                    && defined $early{$addr}[3]) # ... need special treatment
                {
                    push $eof_handler{$addr}->@*, \&_fixup_obsolete_hanguls;
                }
            }

            # And this substitute is valid for all releases.
            $first_released{$addr} = v0;
        }
        else {  # Normal behavior
            $progress = $file{$addr};
            unshift $early{$addr}->@*, 0; # No substitute
        }

        my $file = $file{$addr};
        $progress_message{$addr} = "Processing $progress"
                                            unless $progress_message{$addr};

        # A file should be there if it is within the window of versions for
        # which Unicode supplies it
        if ($withdrawn{$addr} && $withdrawn{$addr} le $v_version) {
            $in_this_release{$addr} = 0;
            $skip{$addr} = "";
        }
        else {
            $in_this_release{$addr} = $first_released{$addr} le $v_version;

            # Check that the file for this object (possibly using a substitute
            # for early releases) exists or we have a function alternative
            if (   ! $function_instead_of_file
                && ! main::file_exists($file))
            {
                # Here there is nothing available for this release.  This is
                # fine if we aren't expecting anything in this release.
                if (! $in_this_release{$addr}) {
                    $skip{$addr} = "";  # Don't remark since we expected
                                        # nothing and got nothing
                }
                elsif ($optional{$addr}->@*) {

                    # Here the file is optional in this release; Use the
                    # passed in text to document this case in the pod.
                    $skip{$addr} = $pod_message_for_non_existent_optional;
                }
                elsif (   $in_this_release{$addr}
                       && ! defined $skip{$addr}
                       && defined $file)
                { # Doesn't exist but should.
                    $skip{$addr} = "'$file' not found.  Possibly Big problems";
                    Carp::my_carp($skip{$addr});
                }
            }
            elsif ($debug_skip && ! defined $skip{$addr} && ! $non_skip{$addr})
            {

                # The file exists; if not skipped for another reason, and we are
                # skipping most everything during debugging builds, use that as
                # the skip reason.
                $skip{$addr} = '$debug_skip is on'
            }
        }

        if (   ! $debug_skip
            && $non_skip{$addr}
            && ! $required_even_in_debug_skip{$addr}
            && $verbosity)
        {
            print "Warning: " . __PACKAGE__ . " constructor for $file has useless 'non_skip' in it\n";
        }

        # Here, we have figured out if we will be skipping this file or not.
        # If so, we add any single property it defines to any passed in
        # optional property list.  These will be dealt with at run time.
        if (defined $skip{$addr}) {
            if ($property{$addr}) {
                push $optional{$addr}->@*, $property{$addr};
            }
        } # Otherwise, are going to process the file.
        elsif ($property{$addr}) {

            # If the file has a property defined in the constructor for it, it
            # means that the property is not listed in the file's entries.  So
            # add a handler (to the list of line handlers) to insert the
            # property name into the lines, to provide a uniform interface to
            # the final processing subroutine.
            push @{$each_line_handler{$addr}}, \&_insert_property_into_line;
        }
        elsif ($properties{$addr}) {

            # Similarly, there may be more than one property represented on
            # each line, with no clue but the constructor input what those
            # might be.  Add a handler for each line in the input so that it
            # creates a separate input line for each property in those input
            # lines, thus making them suitable to handle generically.

            push @{$each_line_handler{$addr}},
                 sub {
                    my $file = shift;
                    Carp::carp_extra_args(\@_) if main::DEBUG && @_;
                    my @fields = split /\s*;\s*/, $_, -1;

                    if (@fields - 1 > @{$properties{$addr}}) {
                        $file->carp_bad_line('Extra fields');
                        $_ = "";
                        return;
                    }
                    my $range = shift @fields;  # 0th element is always the
                                                # range

                    # The next fields in the input line correspond
                    # respectively to the stored properties.
                    for my $i (0 ..  @{$properties{$addr}} - 1) {
                        my $property_name = $properties{$addr}[$i];
                        next if $property_name eq '<ignored>';
                        $file->insert_adjusted_lines(
                              "$range; $property_name; $fields[$i]");
                    }
                    $_ = "";

                    return;
                };
        }

        {   # On non-ascii platforms, we use a special pre-handler
            no strict;
            no warnings 'once';
            *next_line = (main::NON_ASCII_PLATFORM)
                         ? *_next_line_with_remapped_range
                         : *_next_line;
        }

        &{$construction_time_handler{$addr}}($self)
                                        if $construction_time_handler{$addr};

        return $self;
    }


    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
        "." => \&main::_operator_dot,
        ".=" => \&main::_operator_dot_equal,
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {
        return __PACKAGE__ . " object for " . $self->file;
    }

    sub run($self) {
        # Process the input object $self.  This opens and closes the file and
        # calls all the handlers for it.  Currently,  this can only be called
        # once per file, as it destroy's the EOF handlers

        # flag to make sure extracted files are processed early
        state $seen_non_extracted = 0;

        my $addr = pack 'J', refaddr $self;

        my $file = $file{$addr};

        if (! $file) {
            $handle{$addr} = 'pretend_is_open';
        }
        else {
            if ($seen_non_extracted) {
                if ($file =~ /$EXTRACTED/i) # Some platforms may change the
                                            # case of the file's name
                {
                    Carp::my_carp_bug(main::join_lines(<<END
$file should be processed just after the 'Prop...Alias' files, and before
anything not in the $EXTRACTED_DIR directory.  Proceeding, but the results may
have subtle problems
END
                    ));
                }
            }
            elsif ($EXTRACTED_DIR

                    # We only do this check for generic property files
                    && $handler{$addr} == \&main::process_generic_property_file

                    && $file !~ /$EXTRACTED/i)
            {
                # We don't set this (by the 'if' above) if we have no
                # extracted directory, so if running on an early version,
                # this test won't work.  Not worth worrying about.
                $seen_non_extracted = 1;
            }

            # Mark the file as having being processed, and warn if it
            # isn't a file we are expecting.  As we process the files,
            # they are deleted from the hash, so any that remain at the
            # end of the program are files that we didn't process.
            my $fkey = File::Spec->rel2abs($file);
            my $exists = delete $potential_files{lc($fkey)};

            Carp::my_carp("Was not expecting '$file'.")
                                    if $exists && ! $in_this_release{$addr};

            # If there is special handling for compiling Unicode releases
            # earlier than the first one in which Unicode defines this
            # property ...
            if ($early{$addr}->@* > 1) {

                # Mark as processed any substitute file that would be used in
                # such a release
                $fkey = File::Spec->rel2abs($early{$addr}[1]);
                delete $potential_files{lc($fkey)};

                # As commented in the constructor code, when using the
                # official property, we still have to allow the publicly
                # inaccessible early name so that the core code which uses it
                # will work regardless.
                if (   ! $only_early{$addr}
                    && ! $early{$addr}[0]
                    && $early{$addr}->@* > 2)
                {
                    my $early_property_name = $early{$addr}[2];
                    if ($property{$addr} ne $early_property_name) {
                        main::property_ref($property{$addr})
                                            ->add_alias($early_property_name);
                    }
                }
            }

            # We may be skipping this file ...
            if (defined $skip{$addr}) {

                # If the file isn't supposed to be in this release, there is
                # nothing to do
                if ($in_this_release{$addr}) {

                    # But otherwise, we may print a message
                    if ($debug_skip) {
                        print STDERR "Skipping input file '$file'",
                                     " because '$skip{$addr}'\n";
                    }

                    # And add it to the list of skipped files, which is later
                    # used to make the pod
                    $skipped_files{$file} = $skip{$addr};

                    # The 'optional' list contains properties that are also to
                    # be skipped along with the file.  (There may also be
                    # digits which are just placeholders to make sure it isn't
                    # an empty list
                    foreach my $property ($optional{$addr}->@*) {
                        next unless $property =~ /\D/;
                        my $prop_object = main::property_ref($property);
                        next unless defined $prop_object;
                        $prop_object->set_fate($SUPPRESSED, $skip{$addr});
                    }
                }

                return;
            }

            # Here, we are going to process the file.  Open it, converting the
            # slashes used in this program into the proper form for the OS
            my $file_handle;
            if (not open $file_handle, "<", $file) {
                Carp::my_carp("Can't open $file.  Skipping: $!");
                return;
            }
            $handle{$addr} = $file_handle; # Cache the open file handle

            # If possible, make sure that the file is the correct version.
            # (This data isn't available on early Unicode releases or in
            # UnicodeData.txt.)  We don't do this check if we are using a
            # substitute file instead of the official one (though the code
            # could be extended to do so).
            if ($in_this_release{$addr}
                && ! $early{$addr}[0]
                && lc($file) ne 'unicodedata.txt')
            {
                my $this_version;

                if ($file !~ /^Unihan/i) {

                    # The non-Unihan files started getting version numbers in
                    # 3.2, but some files in 4.0 are unchanged from 3.2, and
                    # marked as 3.2.  4.0.1 is the first version where there
                    # are no files marked as being from less than 4.0, though
                    # some are marked as 4.0.  In versions after that, the
                    # numbers are correct.
                    if ($v_version ge v4.0.1) {
                        $_ = <$file_handle>;    # The version number is in the
                                                # very first line if it is a
                                                # UCD file; otherwise, it
                                                # might be
                        goto valid_version if $_ =~ / - $string_version \. /x;
                        chomp;
                        if ($ucd{$addr}) {
                            $_ =~ s/^#\s*//;

                            # 4.0.1 had some valid files that weren't updated.
                            goto valid_version
                                    if $v_version eq v4.0.1 && $_ =~ /4\.0\.0/;
                            $this_version = $_;
                            goto wrong_version;
                        }
                        else {
                            my $BOM = "\x{FEFF}";
                            utf8::encode($BOM);
                            my $BOM_re = qr/ ^ (?:$BOM)? /x;

                            do {
                                chomp;

                                # BOM; seems to be on many lines in some
                                # files!!
                                $_ =~ s/$BOM_re//;

                                if (/./) {

                                    # Only look for the version if in the
                                    # first comment block.
                                    goto no_version unless $_ =~ /^#/;

                                    if ($_ =~ /Version:? (\S*)/) {
                                        $this_version = $1;
                                        goto valid_version
                                          if  $this_version eq $string_version;
                                        goto valid_version
                                            if  "$this_version.0"
                                                            eq $string_version;
                                    }
                                }
                            } while (<$file_handle>);

                            goto no_version;
                        }
                    }
                }
                elsif ($v_version ge v6.0.0) { # Unihan

                    # Unihan files didn't get accurate version numbers until
                    # 6.0.  The version is somewhere in the first comment
                    # block
                    while (<$file_handle>) {
                        goto no_version if $_ !~ /^#/;
                        chomp;
                        $_ =~ s/^#\s*//;
                        next if $_ !~ / version: /x;
                        goto valid_version if $_ =~ /$string_version/;
                        goto wrong_version;
                    }
                    goto no_version;
                }
                else {  # Old Unihan; have to assume is valid
                    goto valid_version;
                }

              wrong_version:
                die Carp::my_carp("File '$file' is version "
                                . "'$this_version'.  It should be "
                                . "version $string_version");
              no_version:
                Carp::my_carp_bug("Could not find the expected "
                                . "version info in file '$file'");
            }
        }

      valid_version:
        print "$progress_message{$addr}\n" if $verbosity >= $PROGRESS;

        # Call any special handler for before the file.
        &{$pre_handler{$addr}}($self) if $pre_handler{$addr};

        # Then the main handler
        &{$handler{$addr}}($self);

        # Then any special post-file handler.
        &{$post_handler{$addr}}($self) if $post_handler{$addr};

        # If any errors have been accumulated, output the counts (as the first
        # error message in each class was output when it was encountered).
        if ($errors{$addr}) {
            my $total = 0;
            my $types = 0;
            foreach my $error (keys %{$errors{$addr}}) {
                $total += $errors{$addr}->{$error};
                delete $errors{$addr}->{$error};
                $types++;
            }
            if ($total > 1) {
                my $message
                        = "A total of $total lines had errors in $file.  ";

                $message .= ($types == 1)
                            ? '(Only the first one was displayed.)'
                            : '(Only the first of each type was displayed.)';
                Carp::my_carp($message);
            }
        }

        if (@{$missings{$addr}}) {
            Carp::my_carp_bug("Handler for $file didn't look at all the \@missing lines.  Generated tables likely are wrong");
        }

        # If a real file handle, close it.
        close $handle{$addr} or Carp::my_carp("Can't close $file: $!") if
                                                        ref $handle{$addr};
        $handle{$addr} = "";   # Uses empty to indicate that has already seen
                               # the file, as opposed to undef
        return;
    }

    sub _next_line($self) {
        # Sets $_ to be the next logical input line, if any.  Returns non-zero
        # if such a line exists.  'logical' means that any lines that have
        # been added via insert_lines() will be returned in $_ before the file
        # is read again.

        my $addr = pack 'J', refaddr $self;

        # Here the file is open (or if the handle is not a ref, is an open
        # 'virtual' file).  Get the next line; any inserted lines get priority
        # over the file itself.
        my $adjusted;

        LINE:
        while (1) { # Loop until find non-comment, non-empty line
            #local $to_trace = 1 if main::DEBUG;
            my $inserted_ref = shift @{$added_lines{$addr}};
            if (defined $inserted_ref) {
                ($adjusted, $_) = @{$inserted_ref};
                trace $adjusted, $_ if main::DEBUG && $to_trace;
                return 1 if $adjusted;
            }
            else {
                last if ! ref $handle{$addr}; # Don't read unless is real file
                last if ! defined ($_ = readline $handle{$addr});
            }
            chomp;
            trace $_ if main::DEBUG && $to_trace;

            # See if this line is the comment line that defines what property
            # value that code points that are not listed in the file should
            # have.  The format or existence of these lines is not guaranteed
            # by Unicode since they are comments, but the documentation says
            # that this was added for machine-readability, so probably won't
            # change.  This works starting in Unicode Version 5.0.  They look
            # like:
            #
            # @missing: 0000..10FFFF; Not_Reordered
            # @missing: 0000..10FFFF; Decomposition_Mapping; <code point>
            # @missing: 0000..10FFFF; ; NaN
            #
            # Save the line for a later get_missings() call.
            if (/$missing_defaults_prefix/) {
                if ($has_missings_defaults{$addr} == $NO_DEFAULTS) {
                    $self->carp_bad_line("Unexpected \@missing line.  Assuming no missing entries");
                }
                elsif ($has_missings_defaults{$addr} == $NOT_IGNORED) {
                    my $start = $1;     # The pattern saves the beginning and
                    my $end = $2;       # end points of the range the default
                                        # is for
                    my @defaults = split /\s* ; \s*/x, $_;

                    # The first field is the @missing, which ends in a
                    # semi-colon, so can safely shift.
                    shift @defaults;

                    # Some of these lines may have empty field placeholders
                    # which get in the way.  An example is:
                    # @missing: 0000..10FFFF; ; NaN
                    # Remove them.  Process starting from the top so the
                    # splice doesn't affect things still to be looked at.
                    for (my $i = @defaults - 1; $i >= 0; $i--) {
                        next if $defaults[$i] ne "";
                        splice @defaults, $i, 1;
                    }

                    # What's left should be just the property (maybe) and the
                    # default.  Having only one element means it doesn't have
                    # the property.
                    my $default;
                    my $property;
                    if (@defaults >= 1) {
                        if (@defaults == 1) {
                            $default = $defaults[0];
                        }
                        else {
                            $property = $defaults[0];
                            $default = $defaults[1];
                        }
                    }

                    if (@defaults < 1
                        || @defaults > 2
                        || ($default =~ /^</
                            && $default !~ /^<code *point>$/i
                            && $default !~ /^<none>$/i
                            && $default !~ /^<script>$/i))
                    {
                        $self->carp_bad_line("Unrecognized \@missing line: $_.  Assuming no missing entries");
                    }
                    else {

                        # If the property is missing from the line, it should
                        # be the one for the whole file
                        $property = $property{$addr} if ! defined $property;

                        # Change <none> to the null string, which is what it
                        # really means.  If the default is the code point
                        # itself, set it to <code point>, which is what
                        # Unicode uses (but sometimes they've forgotten the
                        # space)
                        if ($default =~ /^<none>$/i) {
                            $default = "";
                        }
                        elsif ($default =~ /^<code *point>$/i) {
                            $default = $CODE_POINT;
                        }
                        elsif ($default =~ /^<script>$/i) {

                            # Special case this one.  Currently is from
                            # ScriptExtensions.txt, and means for all unlisted
                            # code points, use their Script property values.
                            # For the code points not listed in that file, the
                            # default value is 'Unknown'.
                            $default = "Unknown";
                        }

                        # Store them as a sub-hash as part of an array, with
                        # both components.
                        push @{$missings{$addr}}, { start    => hex $start,
                                                    end      => hex $end,
                                                    default  => $default,
                                                    property => $property
                                                  };
                    }
                }

                # There is nothing for the caller to process on this comment
                # line.
                next;
            }

            # Unless to keep, remove comments.  If to keep, ignore
            # comment-only lines
            if ($retain_trailing_comments{$addr}) {
                next if / ^ \s* \# /x;

                # But escape any single quotes (done in both the comment and
                # non-comment portion; this could be a bug someday, but not
                # likely)
                s/'/\\'/g;
            }
            else {
                s/#.*//;
            }

            # Remove trailing space, and skip this line if the result is empty
            s/\s+$//;
            next if /^$/;

            # Call any handlers for this line, and skip further processing of
            # the line if the handler sets the line to null.
            foreach my $sub_ref (@{$each_line_handler{$addr}}) {
                &{$sub_ref}($self);
                next LINE if /^$/;
            }

            # Here the line is ok.  return success.
            return 1;
        } # End of looping through lines.

        # If there are EOF handlers, call each (only once) and if it generates
        # more lines to process go back in the loop to handle them.
        while ($eof_handler{$addr}->@*) {
            &{$eof_handler{$addr}[0]}($self);
            shift $eof_handler{$addr}->@*;   # Currently only get one shot at it.
            goto LINE if $added_lines{$addr};
        }

        # Return failure -- no more lines.
        return 0;

    }

    sub _next_line_with_remapped_range($self) {
        # like _next_line(), but for use on non-ASCII platforms.  It sets $_
        # to be the next logical input line, if any.  Returns non-zero if such
        # a line exists.  'logical' means that any lines that have been added
        # via insert_lines() will be returned in $_ before the file is read
        # again.
        #
        # The difference from _next_line() is that this remaps the Unicode
        # code points in the input to those of the native platform.  Each
        # input line contains a single code point, or a single contiguous
        # range of them  This routine splits each range into its individual
        # code points and caches them.  It returns the cached values,
        # translated into their native equivalents, one at a time, for each
        # call, before reading the next line.  Since native values can only be
        # a single byte wide, no translation is needed for code points above
        # 0xFF, and ranges that are entirely above that number are not split.
        # If an input line contains the range 254-1000, it would be split into
        # three elements: 254, 255, and 256-1000.  (The downstream table
        # insertion code will sort and coalesce the individual code points
        # into appropriate ranges.)

        my $addr = pack 'J', refaddr $self;

        while (1) {

            # Look in cache before reading the next line.  Return any cached
            # value, translated
            my $inserted = shift @{$remapped_lines{$addr}};
            if (defined $inserted) {
                trace $inserted if main::DEBUG && $to_trace;
                $_ = $inserted =~ s/^ ( \d+ ) /sprintf("%04X", utf8::unicode_to_native($1))/xer;
                trace $_ if main::DEBUG && $to_trace;
                return 1;
            }

            # Get the next line.
            return 0 unless _next_line($self);

            # If there is a special handler for it, return the line,
            # untranslated.  This should happen only for files that are
            # special, not being code-point related, such as property names.
            return 1 if $handler{$addr}
                                    != \&main::process_generic_property_file;

            my ($range, $property_name, $map, @remainder)
                = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields

            if (@remainder
                || ! defined $property_name
                || $range !~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
            {
                Carp::my_carp_bug("Unrecognized input line '$_'.  Ignored");
            }

            my $low = hex $1;
            my $high = (defined $2) ? hex $2 : $low;

            # If the input maps the range to another code point, remap the
            # target if it is between 0 and 255.
            my $tail;
            if (defined $map) {
                $map =~ s/\b 00 ( [0-9A-F]{2} ) \b/sprintf("%04X", utf8::unicode_to_native(hex $1))/gxe;
                $tail = "$property_name; $map";
                $_ = "$range; $tail";
            }
            else {
                $tail = $property_name;
            }

            # If entire range is above 255, just return it, unchanged (except
            # any mapped-to code point, already changed above)
            return 1 if $low > 255;

            # Cache an entry for every code point < 255.  For those in the
            # range above 255, return a dummy entry for just that portion of
            # the range.  Note that this will be out-of-order, but that is not
            # a problem.
            foreach my $code_point ($low .. $high) {
                if ($code_point > 255) {
                    $_ = sprintf "%04X..%04X; $tail", $code_point, $high;
                    return 1;
                }
                push @{$remapped_lines{$addr}}, "$code_point; $tail";
            }
        } # End of looping through lines.

        # NOTREACHED
    }

#   Not currently used, not fully tested.
#    sub peek {
#        # Non-destructive lookahead one non-adjusted, non-comment, non-blank
#        # record.  Not callable from an each_line_handler(), nor does it call
#        # an each_line_handler() on the line.
#
#        my $self = shift;
#        my $addr = pack 'J', refaddr $self;
#
#        foreach my $inserted_ref (@{$added_lines{$addr}}) {
#            my ($adjusted, $line) = @{$inserted_ref};
#            next if $adjusted;
#
#            # Remove comments and trailing space, and return a non-empty
#            # resulting line
#            $line =~ s/#.*//;
#            $line =~ s/\s+$//;
#            return $line if $line ne "";
#        }
#
#        return if ! ref $handle{$addr}; # Don't read unless is real file
#        while (1) { # Loop until find non-comment, non-empty line
#            local $to_trace = 1 if main::DEBUG;
#            trace $_ if main::DEBUG && $to_trace;
#            return if ! defined (my $line = readline $handle{$addr});
#            chomp $line;
#            push @{$added_lines{$addr}}, [ 0, $line ];
#
#            $line =~ s/#.*//;
#            $line =~ s/\s+$//;
#            return $line if $line ne "";
#        }
#
#        return;
#    }


    sub insert_lines($self, @lines) {
        # Lines can be inserted so that it looks like they were in the input
        # file at the place it was when this routine is called.  See also
        # insert_adjusted_lines().  Lines inserted via this routine go through
        # any each_line_handler()

        # Each inserted line is an array, with the first element being 0 to
        # indicate that this line hasn't been adjusted, and needs to be
        # processed.
        push @{$added_lines{pack 'J', refaddr $self}}, map { [ 0, $_ ] } @lines;
        return;
    }

    sub insert_adjusted_lines($self, @lines) {
        # Lines can be inserted so that it looks like they were in the input
        # file at the place it was when this routine is called.  See also
        # insert_lines().  Lines inserted via this routine are already fully
        # adjusted, ready to be processed; each_line_handler()s handlers will
        # not be called.  This means this is not a completely general
        # facility, as only the last each_line_handler on the stack should
        # call this.  It could be made more general, by passing to each of the
        # line_handlers their position on the stack, which they would pass on
        # to this routine, and that would replace the boolean first element in
        # the anonymous array pushed here, so that the next_line routine could
        # use that to call only those handlers whose index is after it on the
        # stack.  But this is overkill for what is needed now.

        trace $self if main::DEBUG && $to_trace;

        # Each inserted line is an array, with the first element being 1 to
        # indicate that this line has been adjusted
        push @{$added_lines{pack 'J', refaddr $self}}, map { [ 1, $_ ] } @lines;
        return;
    }

    sub get_missings($self) {
        # Returns the stored up @missings lines' values, and clears the list.
        # The values are in a hash, consisting of 'default' and 'property'.
        # However, since these lines can be stacked up, the return is an array
        # of all these hashes.

        my $addr = pack 'J', refaddr $self;

        # If not accepting a list return, just return the first one.
        return shift @{$missings{$addr}} unless wantarray;

        my @return = @{$missings{$addr}};
        undef @{$missings{$addr}};
        return @return;
    }

    sub _exclude_unassigned($self) {

        # Takes the range in $_ and excludes code points that aren't assigned
        # in this release

        state $skip_inserted_count = 0;

        # Ignore recursive calls.
        if ($skip_inserted_count) {
            $skip_inserted_count--;
            return;
        }

        # Find what code points are assigned in this release
        main::calculate_Assigned() if ! defined $Assigned;

        my ($range, @remainder)
            = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields

        # Examine the range.
        if ($range =~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
        {
            my $low = hex $1;
            my $high = (defined $2) ? hex $2 : $low;

            # Split the range into subranges of just those code points in it
            # that are assigned.
            my @ranges = (Range_List->new(Initialize
                              => Range->new($low, $high)) & $Assigned)->ranges;

            # Do nothing if nothing in the original range is assigned in this
            # release; handle normally if everything is in this release.
            if (! @ranges) {
                $_ = "";
            }
            elsif (@ranges != 1) {

                # Here, some code points in the original range aren't in this
                # release; @ranges gives the ones that are.  Create fake input
                # lines for each of the ranges, and set things up so that when
                # this routine is called on that fake input, it will do
                # nothing.
                $skip_inserted_count = @ranges;
                my $remainder = join ";", @remainder;
                for my $range (@ranges) {
                    $self->insert_lines(sprintf("%04X..%04X;%s",
                                    $range->start, $range->end, $remainder));
                }
                $_ = "";    # The original range is now defunct.
            }
        }

        return;
    }

    sub _fixup_obsolete_hanguls($self) {

        # This is called only when compiling Unicode version 1.  All Unicode
        # data for subsequent releases assumes that the code points that were
        # Hangul syllables in this release only are something else, so if
        # using such data, we have to override it

        my $addr = pack 'J', refaddr $self;

        my $object = main::property_ref($property{$addr});
        $object->add_map($FIRST_REMOVED_HANGUL_SYLLABLE,
                         $FINAL_REMOVED_HANGUL_SYLLABLE,
                         $early{$addr}[3],  # Passed-in value for these
                         Replace => $UNCONDITIONALLY);
    }

    sub _insert_property_into_line($self) {
        # Add a property field to $_, if this file requires it.

        my $property = $property{pack 'J', refaddr $self};
        $_ =~ s/(;|$)/; $property$1/;
        return;
    }

    sub carp_bad_line($self, $message="") {
        # Output consistent error messages, using either a generic one, or the
        # one given by the optional parameter.  To avoid gazillions of the
        # same message in case the syntax of a  file is way off, this routine
        # only outputs the first instance of each message, incrementing a
        # count so the totals can be output at the end of the file.

        my $addr = pack 'J', refaddr $self;

        $message = 'Unexpected line' unless $message;

        # No trailing punctuation so as to fit with our addenda.
        $message =~ s/[.:;,]$//;

        # If haven't seen this exact message before, output it now.  Otherwise
        # increment the count of how many times it has occurred
        unless ($errors{$addr}->{$message}) {
            Carp::my_carp("$message in '$_' in "
                            . $file{$addr}
                            . " at line $..  Skipping this line;");
            $errors{$addr}->{$message} = 1;
        }
        else {
            $errors{$addr}->{$message}++;
        }

        # Clear the line to prevent any further (meaningful) processing of it.
        $_ = "";

        return;
    }
} # End closure

package Multi_Default;

sub trace { return main::trace(@_); }

# Certain properties in early versions of Unicode had more than one possible
# default for code points missing from the files.  In these cases, one
# default applies to everything left over after all the others are applied,
# and for each of the others, there is a description of which class of code
# points applies to it.  This object helps implement this by storing the
# defaults, and for all but that final default, an eval string that generates
# the class that it applies to.  That class must be a Range_List, or contains
# a Range_List that the overloaded operators recognize as to be operated on.
# A string is used because this is called early when we know symbolically what
# needs to be done, but typically before any data is gathered.  Thus the
# evaluation gets delayed until we have at hand all the needed information.

{   # Closure

    main::setup_package();

    my %class_defaults;
    # The defaults structure for the classes
    main::set_access('class_defaults', \%class_defaults, 'readable_array');

    my %other_default;
    # The default that applies to everything left over.
    main::set_access('other_default', \%other_default, 'r');

    my %iterator;

    sub new {
        # The constructor is called with default => eval pairs, terminated by
        # the left-over default. e.g.
        # Multi_Default->new(
        #        'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C
        #               -  0x200D',
        #        'R' => 'some other expression that evaluates to code points',
        #        .
        #        .
        #        .
        #        'U'));
        # It is best to leave the final value be the one that matches the
        # above-Unicode code points.

        my $class = shift;

        my $self = bless \do{my $anonymous_scalar}, $class;
        my $addr = pack 'J', refaddr $self;
        $iterator{$addr} = 0;

        return $self unless @_;

        while (@_ > 1) {
            $self->append_default(shift, shift);
        }

        $self->set_final_default(shift);

        return $self;
    }

    sub append_default($self, $new_default, $eval) {
        my $addr = pack 'J', refaddr $self;

        # Pushes a default setting to the current list
        push $class_defaults{$addr}->@*, [ $new_default, $eval ];
    }

    sub set_final_default($self, $new_default) {
        my $addr = pack 'J', refaddr $self;
        $other_default{$addr} = $new_default;
    }

    sub get_next_defaults($self) {
        # Iterates and returns the next class of defaults.

        my $addr = pack 'J', refaddr $self;
        if ($iterator{$addr}++ < $class_defaults{$addr}->@*) {
            return $class_defaults{$addr}->[$iterator{$addr}-1]->@*;
        }

        $iterator{$addr} = 0;
        return undef;
    }
}

package Alias;

# An alias is one of the names that a table goes by.  This class defines them
# including some attributes.  Everything is currently setup in the
# constructor.


{   # Closure

    main::setup_package();

    my %name;
    main::set_access('name', \%name, 'r');

    my %loose_match;
    # Should this name match loosely or not.
    main::set_access('loose_match', \%loose_match, 'r');

    my %make_re_pod_entry;
    # Some aliases should not get their own entries in the re section of the
    # pod, because they are covered by a wild-card, and some we want to
    # discourage use of.  Binary
    main::set_access('make_re_pod_entry', \%make_re_pod_entry, 'r', 's');

    my %ucd;
    # Is this documented to be accessible via Unicode::UCD
    main::set_access('ucd', \%ucd, 'r', 's');

    my %status;
    # Aliases have a status, like deprecated, or even suppressed (which means
    # they don't appear in documentation).  Enum
    main::set_access('status', \%status, 'r');

    my %ok_as_filename;
    # Similarly, some aliases should not be considered as usable ones for
    # external use, such as file names, or we don't want documentation to
    # recommend them.  Boolean
    main::set_access('ok_as_filename', \%ok_as_filename, 'r');

    sub new {
        my $class = shift;

        my $self = bless \do { my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        $name{$addr} = shift;
        $loose_match{$addr} = shift;
        $make_re_pod_entry{$addr} = shift;
        $ok_as_filename{$addr} = shift;
        $status{$addr} = shift;
        $ucd{$addr} = shift;

        Carp::carp_extra_args(\@_) if main::DEBUG && @_;

        # Null names are never ok externally
        $ok_as_filename{$addr} = 0 if $name{$addr} eq "";

        return $self;
    }
}

package Range;

# A range is the basic unit for storing code points, and is described in the
# comments at the beginning of the program.  Each range has a starting code
# point; an ending code point (not less than the starting one); a value
# that applies to every code point in between the two end-points, inclusive;
# and an enum type that applies to the value.  The type is for the user's
# convenience, and has no meaning here, except that a non-zero type is
# considered to not obey the normal Unicode rules for having standard forms.
#
# The same structure is used for both map and match tables, even though in the
# latter, the value (and hence type) is irrelevant and could be used as a
# comment.  In map tables, the value is what all the code points in the range
# map to.  Type 0 values have the standardized version of the value stored as
# well, so as to not have to recalculate it a lot.

sub trace { return main::trace(@_); }

{   # Closure

    main::setup_package();

    my %start;
    main::set_access('start', \%start, 'r', 's');

    my %end;
    main::set_access('end', \%end, 'r', 's');

    my %value;
    main::set_access('value', \%value, 'r', 's');

    my %type;
    main::set_access('type', \%type, 'r');

    my %standard_form;
    # The value in internal standard form.  Defined only if the type is 0.
    main::set_access('standard_form', \%standard_form);

    # Note that if these fields change, the dump() method should as well

    sub new($class, $_addr, $_end, @_args) {
        my $self = bless \do { my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        $start{$addr} = $_addr;
        $end{$addr}   = $_end;

        my %args = @_args;

        my $value = delete $args{'Value'};  # Can be 0
        $value = "" unless defined $value;
        $value{$addr} = $value;

        $type{$addr} = delete $args{'Type'} || 0;

        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        return $self;
    }

    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
        "." => \&main::_operator_dot,
        ".=" => \&main::_operator_dot_equal,
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {
        my $addr = pack 'J', refaddr $self;

        # Output it like '0041..0065 (value)'
        my $return = sprintf("%04X", $start{$addr})
                        .  '..'
                        . sprintf("%04X", $end{$addr});
        my $value = $value{$addr};
        my $type = $type{$addr};
        $return .= ' (';
        $return .= "$value";
        $return .= ", Type=$type" if $type != 0;
        $return .= ')';

        return $return;
    }

    sub standard_form($self) {
        # Calculate the standard form only if needed, and cache the result.
        # The standard form is the value itself if the type is special.
        # This represents a considerable CPU and memory saving - at the time
        # of writing there are 368676 non-special objects, but the standard
        # form is only requested for 22047 of them - ie about 6%.

        my $addr = pack 'J', refaddr $self;

        return $standard_form{$addr} if defined $standard_form{$addr};

        my $value = $value{$addr};
        return $value if $type{$addr};
        return $standard_form{$addr} = main::standardize($value);
    }

    sub dump($self, $indent) {
        # Human, not machine readable.  For machine readable, comment out this
        # entire routine and let the standard one take effect.
        my $addr = pack 'J', refaddr $self;

        my $return = $indent
                    . sprintf("%04X", $start{$addr})
                    . '..'
                    . sprintf("%04X", $end{$addr})
                    . " '$value{$addr}';";
        if (! defined $standard_form{$addr}) {
            $return .= "(type=$type{$addr})";
        }
        elsif ($standard_form{$addr} ne $value{$addr}) {
            $return .= "(standard '$standard_form{$addr}')";
        }
        return $return;
    }
} # End closure

package _Range_List_Base;

# Base class for range lists.  A range list is simply an ordered list of
# ranges, so that the ranges with the lowest starting numbers are first in it.
#
# When a new range is added that is adjacent to an existing range that has the
# same value and type, it merges with it to form a larger range.
#
# Ranges generally do not overlap, except that there can be multiple entries
# of single code point ranges.  This is because of NameAliases.txt.
#
# In this program, there is a standard value such that if two different
# values, have the same standard value, they are considered equivalent.  This
# value was chosen so that it gives correct results on Unicode data

# There are a number of methods to manipulate range lists, and some operators
# are overloaded to handle them.

sub trace { return main::trace(@_); }

{ # Closure

    our $addr;

    # Max is initialized to a negative value that isn't adjacent to 0, for
    # simpler tests
    my $max_init = -2;

    main::setup_package();

    my %ranges;
    # The list of ranges
    main::set_access('ranges', \%ranges, 'readable_array');

    my %max;
    # The highest code point in the list.  This was originally a method, but
    # actual measurements said it was used a lot.
    main::set_access('max', \%max, 'r');

    my %each_range_iterator;
    # Iterator position for each_range()
    main::set_access('each_range_iterator', \%each_range_iterator);

    my %owner_name_of;
    # Name of parent this is attached to, if any.  Solely for better error
    # messages.
    main::set_access('owner_name_of', \%owner_name_of, 'p_r');

    my %_search_ranges_cache;
    # A cache of the previous result from _search_ranges(), for better
    # performance
    main::set_access('_search_ranges_cache', \%_search_ranges_cache);

    sub new {
        my $class = shift;
        my %args = @_;

        # Optional initialization data for the range list.  NOTE: For large
        # ranges, it is better to use Range object rather than
        #   [ low .. high ]
        # as it iterates through each one individually in the latter case.
        my $initialize = delete $args{'Initialize'};

        my $self;

        # Use _union() to initialize.  _union() returns an object of this
        # class, which means that it will call this constructor recursively.
        # But it won't have this $initialize parameter so that it won't
        # infinitely loop on this.
        return _union($class, $initialize, %args) if defined $initialize;

        $self = bless \do { my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        # Optional parent object, only for debug info.
        $owner_name_of{$addr} = delete $args{'Owner'};
        $owner_name_of{$addr} = "" if ! defined $owner_name_of{$addr};

        # Stringify, in case it is an object.
        $owner_name_of{$addr} = "$owner_name_of{$addr}";

        # This is used only for error messages, and so a colon is added
        $owner_name_of{$addr} .= ": " if $owner_name_of{$addr} ne "";

        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        $max{$addr} = $max_init;

        $_search_ranges_cache{$addr} = 0;
        $ranges{$addr} = [];

        return $self;
    }

    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
        "." => \&main::_operator_dot,
        ".=" => \&main::_operator_dot_equal,
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {
        my $addr = pack 'J', refaddr $self;

        return "Range_List attached to '$owner_name_of{$addr}'"
                                                if $owner_name_of{$addr};
        return "anonymous Range_List " . \$self;
    }

    sub _union {
        # Returns the union of the input code points.  It can be called as
        # either a constructor or a method.  If called as a method, the result
        # will be a new() instance of the calling object, containing the union
        # of that object with the other parameter's code points;  if called as
        # a constructor, the first parameter gives the class that the new object
        # should be, and the second parameter gives the code points to go into
        # it.
        # In either case, there are two parameters looked at by this routine;
        # any additional parameters are passed to the new() constructor.
        #
        # The code points can come in the form of some object that contains
        # ranges, and has a conventionally named method to access them; or
        # they can be an array of individual code points (as integers); or
        # just a single code point.
        #
        # If they are ranges, this routine doesn't make any effort to preserve
        # the range values and types of one input over the other.  Therefore
        # this base class should not allow _union to be called from other than
        # initialization code, so as to prevent two tables from being added
        # together where the range values matter.  The general form of this
        # routine therefore belongs in a derived class, but it was moved here
        # to avoid duplication of code.  The failure to overload this in this
        # class keeps it safe.
        #
        # It does make the effort during initialization to accept tables with
        # multiple values for the same code point, and to preserve the order
        # of these.  If there is only one input range or range set, it doesn't
        # sort (as it should already be sorted to the desired order), and will
        # accept multiple values per code point.  Otherwise it will merge
        # multiple values into a single one.

        my $self;
        my @args;   # Arguments to pass to the constructor

        my $class = shift;

        # If a method call, will start the union with the object itself, and
        # the class of the new object will be the same as self.
        if (ref $class) {
            $self = $class;
            $class = ref $self;
            push @args, $self;
        }

        # Add the other required parameter.
        push @args, shift;
        # Rest of parameters are passed on to the constructor

        # Accumulate all records from both lists.
        my @records;
        my $input_count = 0;
        for my $arg (@args) {
            #local $to_trace = 0 if main::DEBUG;
            trace "argument = $arg" if main::DEBUG && $to_trace;
            if (! defined $arg) {
                my $message = "";
                if (defined $self) {
                    $message .= $owner_name_of{pack 'J', refaddr $self};
                }
                Carp::my_carp_bug($message . "Undefined argument to _union.  No union done.");
                return;
            }

            $arg = [ $arg ] if ! ref $arg;
            my $type = ref $arg;
            if ($type eq 'ARRAY') {
                foreach my $element (@$arg) {
                    push @records, Range->new($element, $element);
                    $input_count++;
                }
            }
            elsif ($arg->isa('Range')) {
                push @records, $arg;
                $input_count++;
            }
            elsif ($arg->can('ranges')) {
                push @records, $arg->ranges;
                $input_count++;
            }
            else {
                my $message = "";
                if (defined $self) {
                    $message .= $owner_name_of{pack 'J', refaddr $self};
                }
                Carp::my_carp_bug($message . "Cannot take the union of a $type.  No union done.");
                return;
            }
        }

        # Sort with the range containing the lowest ordinal first, but if
        # two ranges start at the same code point, sort with the bigger range
        # of the two first, because it takes fewer cycles.
        if ($input_count > 1) {
            @records = sort { ($a->start <=> $b->start)
                                      or
                                    # if b is shorter than a, b->end will be
                                    # less than a->end, and we want to select
                                    # a, so want to return -1
                                    ($b->end <=> $a->end)
                                   } @records;
        }

        my $new = $class->new(@_);

        # Fold in records so long as they add new information.
        for my $set (@records) {
            my $start = $set->start;
            my $end   = $set->end;
            my $value = $set->value;
            my $type  = $set->type;
            if ($start > $new->max) {
                $new->_add_delete('+', $start, $end, $value, Type => $type);
            }
            elsif ($end > $new->max) {
                $new->_add_delete('+', $new->max +1, $end, $value,
                                                                Type => $type);
            }
            elsif ($input_count == 1) {
                # Here, overlaps existing range, but is from a single input,
                # so preserve the multiple values from that input.
                $new->_add_delete('+', $start, $end, $value, Type => $type,
                                                Replace => $MULTIPLE_AFTER);
            }
        }

        return $new;
    }

    sub range_count($self) {        # Return the number of ranges in the range list
        return scalar @{$ranges{pack 'J', refaddr $self}};
    }

    sub min($self) {
        # Returns the minimum code point currently in the range list, or if
        # the range list is empty, 2 beyond the max possible.  This is a
        # method because used so rarely, that not worth saving between calls,
        # and having to worry about changing it as ranges are added and
        # deleted.

        my $addr = pack 'J', refaddr $self;

        # If the range list is empty, return a large value that isn't adjacent
        # to any that could be in the range list, for simpler tests
        return $MAX_WORKING_CODEPOINT + 2 unless scalar @{$ranges{$addr}};
        return $ranges{$addr}->[0]->start;
    }

    sub contains($self, $codepoint) {
        # Boolean: Is argument in the range list?  If so returns $i such that:
        #   range[$i]->end < $codepoint <= range[$i+1]->end
        # which is one beyond what you want; this is so that the 0th range
        # doesn't return false

        my $i = $self->_search_ranges($codepoint);
        return 0 unless defined $i;

        # The search returns $i, such that
        #   range[$i-1]->end < $codepoint <= range[$i]->end
        # So is in the table if and only iff it is at least the start position
        # of range $i.
        return 0 if $ranges{pack 'J', refaddr $self}->[$i]->start > $codepoint;
        return $i + 1;
    }

    sub containing_range($self, $codepoint) {
        # Returns the range object that contains the code point, undef if none
        my $i = $self->contains($codepoint);
        return unless $i;

        # contains() returns 1 beyond where we should look
        return $ranges{pack 'J', refaddr $self}->[$i-1];
    }

    sub value_of($self, $codepoint) {
        # Returns the value associated with the code point, undef if none
        my $range = $self->containing_range($codepoint);
        return unless defined $range;

        return $range->value;
    }

    sub type_of($self, $codepoint) {
        # Returns the type of the range containing the code point, undef if
        # the code point is not in the table
        my $range = $self->containing_range($codepoint);
        return unless defined $range;

        return $range->type;
    }

    sub _search_ranges($self, $code_point) {
        # Find the range in the list which contains a code point, or where it
        # should go if were to add it.  That is, it returns $i, such that:
        #   range[$i-1]->end < $codepoint <= range[$i]->end
        # Returns undef if no such $i is possible (e.g. at end of table), or
        # if there is an error.
        my $addr = pack 'J', refaddr $self;

        return if $code_point > $max{$addr};
        my $r = $ranges{$addr};                # The current list of ranges
        my $range_list_size = scalar @$r;
        my $i;

        use integer;        # want integer division

        # Use the cached result as the starting guess for this one, because,
        # an experiment on 5.1 showed that 90% of the time the cache was the
        # same as the result on the next call (and 7% it was one less).
        $i = $_search_ranges_cache{$addr};
        $i = 0 if $i >= $range_list_size;   # Reset if no longer valid (prob.
                                            # from an intervening deletion
        #local $to_trace = 1 if main::DEBUG;
        trace "previous \$i is still valid: $i" if main::DEBUG && $to_trace && $code_point <= $r->[$i]->end && ($i == 0 || $r->[$i-1]->end < $code_point);
        return $i if $code_point <= $r->[$i]->end
                     && ($i == 0 || $r->[$i-1]->end < $code_point);

        # Here the cache doesn't yield the correct $i.  Try adding 1.
        if ($i < $range_list_size - 1
            && $r->[$i]->end < $code_point &&
            $code_point <= $r->[$i+1]->end)
        {
            $i++;
            trace "next \$i is correct: $i" if main::DEBUG && $to_trace;
            $_search_ranges_cache{$addr} = $i;
            return $i;
        }

        # Here, adding 1 also didn't work.  We do a binary search to
        # find the correct position, starting with current $i
        my $lower = 0;
        my $upper = $range_list_size - 1;
        while (1) {
            trace "top of loop i=$i:", sprintf("%04X", $r->[$lower]->start), "[$lower] .. ", sprintf("%04X", $r->[$i]->start), "[$i] .. ", sprintf("%04X", $r->[$upper]->start), "[$upper]" if main::DEBUG && $to_trace;

            if ($code_point <= $r->[$i]->end) {

                # Here we have met the upper constraint.  We can quit if we
                # also meet the lower one.
                last if $i == 0 || $r->[$i-1]->end < $code_point;

                $upper = $i;        # Still too high.

            }
            else {

                # Here, $r[$i]->end < $code_point, so look higher up.
                $lower = $i;
            }

            # Split search domain in half to try again.
            my $temp = ($upper + $lower) / 2;

            # No point in continuing unless $i changes for next time
            # in the loop.
            if ($temp == $i) {

                # We can't reach the highest element because of the averaging.
                # So if one below the upper edge, force it there and try one
                # more time.
                if ($i == $range_list_size - 2) {

                    trace "Forcing to upper edge" if main::DEBUG && $to_trace;
                    $i = $range_list_size - 1;

                    # Change $lower as well so if fails next time through,
                    # taking the average will yield the same $i, and we will
                    # quit with the error message just below.
                    $lower = $i;
                    next;
                }
                Carp::my_carp_bug("$owner_name_of{$addr}Can't find where the range ought to go.  No action taken.");
                return;
            }
            $i = $temp;
        } # End of while loop

        if (main::DEBUG && $to_trace) {
            trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i;
            trace "i=  [ $i ]", $r->[$i];
            trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < $range_list_size - 1;
        }

        # Here we have found the offset.  Cache it as a starting point for the
        # next call.
        $_search_ranges_cache{$addr} = $i;
        return $i;
    }

    sub _add_delete {
        # Add, replace or delete ranges to or from a list.  The $type
        # parameter gives which:
        #   '+' => insert or replace a range, returning a list of any changed
        #          ranges.
        #   '-' => delete a range, returning a list of any deleted ranges.
        #
        # The next three parameters give respectively the start, end, and
        # value associated with the range.  'value' should be null unless the
        # operation is '+';
        #
        # The range list is kept sorted so that the range with the lowest
        # starting position is first in the list, and generally, adjacent
        # ranges with the same values are merged into a single larger one (see
        # exceptions below).
        #
        # There are more parameters; all are key => value pairs:
        #   Type    gives the type of the value.  It is only valid for '+'.
        #           All ranges have types; if this parameter is omitted, 0 is
        #           assumed.  Ranges with type 0 are assumed to obey the
        #           Unicode rules for casing, etc; ranges with other types are
        #           not.  Otherwise, the type is arbitrary, for the caller's
        #           convenience, and looked at only by this routine to keep
        #           adjacent ranges of different types from being merged into
        #           a single larger range, and when Replace =>
        #           $IF_NOT_EQUIVALENT is specified (see just below).
        #   Replace  determines what to do if the range list already contains
        #            ranges which coincide with all or portions of the input
        #            range.  It is only valid for '+':
        #       => $NO            means that the new value is not to replace
        #                         any existing ones, but any empty gaps of the
        #                         range list coinciding with the input range
        #                         will be filled in with the new value.
        #       => $UNCONDITIONALLY  means to replace the existing values with
        #                         this one unconditionally.  However, if the
        #                         new and old values are identical, the
        #                         replacement is skipped to save cycles
        #       => $IF_NOT_EQUIVALENT means to replace the existing values
        #          (the default)  with this one if they are not equivalent.
        #                         Ranges are equivalent if their types are the
        #                         same, and they are the same string; or if
        #                         both are type 0 ranges, if their Unicode
        #                         standard forms are identical.  In this last
        #                         case, the routine chooses the more "modern"
        #                         one to use.  This is because some of the
        #                         older files are formatted with values that
        #                         are, for example, ALL CAPs, whereas the
        #                         derived files have a more modern style,
        #                         which looks better.  By looking for this
        #                         style when the pre-existing and replacement
        #                         standard forms are the same, we can move to
        #                         the modern style
        #       => $MULTIPLE_BEFORE means that if this range duplicates an
        #                         existing one, but has a different value,
        #                         don't replace the existing one, but insert
        #                         this one so that the same range can occur
        #                         multiple times.  They are stored LIFO, so
        #                         that the final one inserted is the first one
        #                         returned in an ordered search of the table.
        #                         If this is an exact duplicate, including the
        #                         value, the original will be moved to be
        #                         first, before any other duplicate ranges
        #                         with different values.
        #       => $MULTIPLE_AFTER is like $MULTIPLE_BEFORE, but is stored
        #                         FIFO, so that this one is inserted after all
        #                         others that currently exist.  If this is an
        #                         exact duplicate, including value, of an
        #                         existing range, this one is discarded
        #                         (leaving the existing one in its original,
        #                         higher priority position
        #       => $CROAK         Die with an error if is already there
        #       => anything else  is the same as => $IF_NOT_EQUIVALENT
        #
        # "same value" means identical for non-type-0 ranges, and it means
        # having the same standard forms for type-0 ranges.

        return Carp::carp_too_few_args(\@_, 5) if main::DEBUG && @_ < 5;

        my $self = shift;
        my $operation = shift;   # '+' for add/replace; '-' for delete;
        my $start = shift;
        my $end   = shift;
        my $value = shift;

        my %args = @_;

        $value = "" if not defined $value;        # warning: $value can be "0"

        my $replace = delete $args{'Replace'};
        $replace = $IF_NOT_EQUIVALENT unless defined $replace;

        my $type = delete $args{'Type'};
        $type = 0 unless defined $type;

        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        my $addr = pack 'J', refaddr $self;

        if ($operation ne '+' && $operation ne '-') {
            Carp::my_carp_bug("$owner_name_of{$addr}First parameter to _add_delete must be '+' or '-'.  No action taken.");
            return;
        }
        unless (defined $start && defined $end) {
            Carp::my_carp_bug("$owner_name_of{$addr}Undefined start and/or end to _add_delete.  No action taken.");
            return;
        }
        unless ($end >= $start) {
            Carp::my_carp_bug("$owner_name_of{$addr}End of range (" . sprintf("%04X", $end) . ") must not be before start (" . sprintf("%04X", $start) . ").  No action taken.");
            return;
        }
        #local $to_trace = 1 if main::DEBUG;

        if ($operation eq '-') {
            if ($replace != $IF_NOT_EQUIVALENT) {
                Carp::my_carp_bug("$owner_name_of{$addr}Replace => \$IF_NOT_EQUIVALENT is required when deleting a range from a range list.  Assuming Replace => \$IF_NOT_EQUIVALENT.");
                $replace = $IF_NOT_EQUIVALENT;
            }
            if ($type) {
                Carp::my_carp_bug("$owner_name_of{$addr}Type => 0 is required when deleting a range from a range list.  Assuming Type => 0.");
                $type = 0;
            }
            if ($value ne "") {
                Carp::my_carp_bug("$owner_name_of{$addr}Value => \"\" is required when deleting a range from a range list.  Assuming Value => \"\".");
                $value = "";
            }
        }

        my $r = $ranges{$addr};               # The current list of ranges
        my $range_list_size = scalar @$r;     # And its size
        my $max = $max{$addr};                # The current high code point in
                                              # the list of ranges

        # Do a special case requiring fewer machine cycles when the new range
        # starts after the current highest point.  The Unicode input data is
        # structured so this is common.
        if ($start > $max) {

            trace "$owner_name_of{$addr} $operation", sprintf("%04X..%04X (%s) type=%d; prev max=%04X", $start, $end, $value, $type, $max) if main::DEBUG && $to_trace;
            return if $operation eq '-'; # Deleting a non-existing range is a
                                         # no-op

            # If the new range doesn't logically extend the current final one
            # in the range list, create a new range at the end of the range
            # list.  (max cleverly is initialized to a negative number not
            # adjacent to 0 if the range list is empty, so even adding a range
            # to an empty range list starting at 0 will have this 'if'
            # succeed.)
            if ($start > $max + 1        # non-adjacent means can't extend.
                || @{$r}[-1]->value ne $value # values differ, can't extend.
                || @{$r}[-1]->type != $type # types differ, can't extend.
            ) {
                push @$r, Range->new($start, $end,
                                     Value => $value,
                                     Type => $type);
            }
            else {

                # Here, the new range starts just after the current highest in
                # the range list, and they have the same type and value.
                # Extend the existing range to incorporate the new one.
                @{$r}[-1]->set_end($end);
            }

            # This becomes the new maximum.
            $max{$addr} = $end;

            return;
        }
        #local $to_trace = 0 if main::DEBUG;

        trace "$owner_name_of{$addr} $operation", sprintf("%04X", $start) . '..' . sprintf("%04X", $end) . " ($value) replace=$replace" if main::DEBUG && $to_trace;

        # Here, the input range isn't after the whole rest of the range list.
        # Most likely 'splice' will be needed.  The rest of the routine finds
        # the needed splice parameters, and if necessary, does the splice.
        # First, find the offset parameter needed by the splice function for
        # the input range.  Note that the input range may span multiple
        # existing ones, but we'll worry about that later.  For now, just find
        # the beginning.  If the input range is to be inserted starting in a
        # position not currently in the range list, it must (obviously) come
        # just after the range below it, and just before the range above it.
        # Slightly less obviously, it will occupy the position currently
        # occupied by the range that is to come after it.  More formally, we
        # are looking for the position, $i, in the array of ranges, such that:
        #
        # r[$i-1]->start <= r[$i-1]->end < $start < r[$i]->start <= r[$i]->end
        #
        # (The ordered relationships within existing ranges are also shown in
        # the equation above).  However, if the start of the input range is
        # within an existing range, the splice offset should point to that
        # existing range's position in the list; that is $i satisfies a
        # somewhat different equation, namely:
        #
        #r[$i-1]->start <= r[$i-1]->end < r[$i]->start <= $start <= r[$i]->end
        #
        # More briefly, $start can come before or after r[$i]->start, and at
        # this point, we don't know which it will be.  However, these
        # two equations share these constraints:
        #
        #   r[$i-1]->end < $start <= r[$i]->end
        #
        # And that is good enough to find $i.

        my $i = $self->_search_ranges($start);
        if (! defined $i) {
            Carp::my_carp_bug("Searching $self for range beginning with $start unexpectedly returned undefined.  Operation '$operation' not performed");
            return;
        }

        # The search function returns $i such that:
        #
        # r[$i-1]->end < $start <= r[$i]->end
        #
        # That means that $i points to the first range in the range list
        # that could possibly be affected by this operation.  We still don't
        # know if the start of the input range is within r[$i], or if it
        # points to empty space between r[$i-1] and r[$i].
        trace "[$i] is the beginning splice point.  Existing range there is ", $r->[$i] if main::DEBUG && $to_trace;

        # Special case the insertion of data that is not to replace any
        # existing data.
        if ($replace == $NO) {  # If $NO, has to be operation '+'
            #local $to_trace = 1 if main::DEBUG;
            trace "Doesn't replace" if main::DEBUG && $to_trace;

            # Here, the new range is to take effect only on those code points
            # that aren't already in an existing range.  This can be done by
            # looking through the existing range list and finding the gaps in
            # the ranges that this new range affects, and then calling this
            # function recursively on each of those gaps, leaving untouched
            # anything already in the list.  Gather up a list of the changed
            # gaps first so that changes to the internal state as new ranges
            # are added won't be a problem.
            my @gap_list;

            # First, if the starting point of the input range is outside an
            # existing one, there is a gap from there to the beginning of the
            # existing range -- add a span to fill the part that this new
            # range occupies
            if ($start < $r->[$i]->start) {
                push @gap_list, Range->new($start,
                                           main::min($end,
                                                     $r->[$i]->start - 1),
                                           Type => $type);
                trace "gap before $r->[$i] [$i], will add", $gap_list[-1] if main::DEBUG && $to_trace;
            }

            # Then look through the range list for other gaps until we reach
            # the highest range affected by the input one.
            my $j;
            for ($j = $i+1; $j < $range_list_size; $j++) {
                trace "j=[$j]", $r->[$j] if main::DEBUG && $to_trace;
                last if $end < $r->[$j]->start;

                # If there is a gap between when this range starts and the
                # previous one ends, add a span to fill it.  Note that just
                # because there are two ranges doesn't mean there is a
                # non-zero gap between them.  It could be that they have
                # different values or types
                if ($r->[$j-1]->end + 1 != $r->[$j]->start) {
                    push @gap_list,
                        Range->new($r->[$j-1]->end + 1,
                                   $r->[$j]->start - 1,
                                   Type => $type);
                    trace "gap between $r->[$j-1] and $r->[$j] [$j], will add: $gap_list[-1]" if main::DEBUG && $to_trace;
                }
            }

            # Here, we have either found an existing range in the range list,
            # beyond the area affected by the input one, or we fell off the
            # end of the loop because the input range affects the whole rest
            # of the range list.  In either case, $j is 1 higher than the
            # highest affected range.  If $j == $i, it means that there are no
            # affected ranges, that the entire insertion is in the gap between
            # r[$i-1], and r[$i], which we already have taken care of before
            # the loop.
            # On the other hand, if there are affected ranges, it might be
            # that there is a gap that needs filling after the final such
            # range to the end of the input range
            if ($r->[$j-1]->end < $end) {
                    push @gap_list, Range->new(main::max($start,
                                                         $r->[$j-1]->end + 1),
                                               $end,
                                               Type => $type);
                    trace "gap after $r->[$j-1], will add $gap_list[-1]" if main::DEBUG && $to_trace;
            }

            # Call recursively to fill in all the gaps.
            foreach my $gap (@gap_list) {
                $self->_add_delete($operation,
                                   $gap->start,
                                   $gap->end,
                                   $value,
                                   Type => $type);
            }

            return;
        }

        # Here, we have taken care of the case where $replace is $NO.
        # Remember that here, r[$i-1]->end < $start <= r[$i]->end
        # If inserting a multiple record, this is where it goes, before the
        # first (if any) existing one if inserting LIFO.  (If this is to go
        # afterwards, FIFO, we below move the pointer to there.)  These imply
        # an insertion, and no change to any existing ranges.  Note that $i
        # can be -1 if this new range doesn't actually duplicate any existing,
        # and comes at the beginning of the list.
        if ($replace == $MULTIPLE_BEFORE || $replace == $MULTIPLE_AFTER) {

            if ($start != $end) {
                Carp::my_carp_bug("$owner_name_of{$addr}Can't cope with adding a multiple record when the range ($start..$end) contains more than one code point.  No action taken.");
                return;
            }

            # If the new code point is within a current range ...
            if ($end >= $r->[$i]->start) {

                # Don't add an exact duplicate, as it isn't really a multiple
                my $existing_value = $r->[$i]->value;
                my $existing_type = $r->[$i]->type;
                return if $value eq $existing_value && $type eq $existing_type;

                # If the multiple value is part of an existing range, we want
                # to split up that range, so that only the single code point
                # is affected.  To do this, we first call ourselves
                # recursively to delete that code point from the table, having
                # preserved its current data above.  Then we call ourselves
                # recursively again to add the new multiple, which we know by
                # the test just above is different than the current code
                # point's value, so it will become a range containing a single
                # code point: just itself.  Finally, we add back in the
                # pre-existing code point, which will again be a single code
                # point range.  Because 'i' likely will have changed as a
                # result of these operations, we can't just continue on, but
                # do this operation recursively as well.  If we are inserting
                # LIFO, the pre-existing code point needs to go after the new
                # one, so use MULTIPLE_AFTER; and vice versa.
                if ($r->[$i]->start != $r->[$i]->end) {
                    $self->_add_delete('-', $start, $end, "");
                    $self->_add_delete('+', $start, $end, $value, Type => $type);
                    return $self->_add_delete('+',
                            $start, $end,
                            $existing_value,
                            Type => $existing_type,
                            Replace => ($replace == $MULTIPLE_BEFORE)
                                       ? $MULTIPLE_AFTER
                                       : $MULTIPLE_BEFORE);
                }
            }

            # If to place this new record after, move to beyond all existing
            # ones; but don't add this one if identical to any of them, as it
            # isn't really a multiple.  This leaves the original order, so
            # that the current request is ignored.  The reasoning is that the
            # previous request that wanted this record to have high priority
            # should have precedence.
            if ($replace == $MULTIPLE_AFTER) {
                while ($i < @$r && $r->[$i]->start == $start) {
                    return if $value eq $r->[$i]->value
                              && $type eq $r->[$i]->type;
                    $i++;
                }
            }
            else {
                # If instead we are to place this new record before any
                # existing ones, remove any identical ones that come after it.
                # This changes the existing order so that the new one is
                # first, as is being requested.
                for (my $j = $i + 1;
                     $j < @$r && $r->[$j]->start == $start;
                     $j++)
                {
                    if ($value eq $r->[$j]->value && $type eq $r->[$j]->type) {
                        splice @$r, $j, 1;
                        last;   # There should only be one instance, so no
                                # need to keep looking
                    }
                }
            }

            trace "Adding multiple record at $i with $start..$end, $value" if main::DEBUG && $to_trace;
            my @return = splice @$r,
                                $i,
                                0,
                                Range->new($start,
                                           $end,
                                           Value => $value,
                                           Type => $type);
            if (main::DEBUG && $to_trace) {
                trace "After splice:";
                trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
                trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
                trace "i  =[", $i, "]", $r->[$i] if $i >= 0;
                trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
                trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
                trace 'i+3=[', $i+3, ']', $r->[$i+3] if $i < @$r - 3;
            }
            return @return;
        }

        # Here, we have taken care of $NO and $MULTIPLE_foo replaces.  This
        # leaves delete, insert, and replace either unconditionally or if not
        # equivalent.  $i still points to the first potential affected range.
        # Now find the highest range affected, which will determine the length
        # parameter to splice.  (The input range can span multiple existing
        # ones.)  If this isn't a deletion, while we are looking through the
        # range list, see also if this is a replacement rather than a clean
        # insertion; that is if it will change the values of at least one
        # existing range.  Start off assuming it is an insert, until find it
        # isn't.
        my $clean_insert = $operation eq '+';
        my $j;        # This will point to the highest affected range

        # For non-zero types, the standard form is the value itself;
        my $standard_form = ($type) ? $value : main::standardize($value);

        for ($j = $i; $j < $range_list_size; $j++) {
            trace "Looking for highest affected range; the one at $j is ", $r->[$j] if main::DEBUG && $to_trace;

            # If find a range that it doesn't overlap into, we can stop
            # searching
            last if $end < $r->[$j]->start;

            # Here, overlaps the range at $j.  If the values don't match,
            # and so far we think this is a clean insertion, it becomes a
            # non-clean insertion, i.e., a 'change' or 'replace' instead.
            if ($clean_insert) {
                if ($r->[$j]->standard_form ne $standard_form) {
                    $clean_insert = 0;
                    if ($replace == $CROAK) {
                        main::croak("The range to add "
                        . sprintf("%04X", $start)
                        . '-'
                        . sprintf("%04X", $end)
                        . " with value '$value' overlaps an existing range $r->[$j]");
                    }
                }
                else {

                    # Here, the two values are essentially the same.  If the
                    # two are actually identical, replacing wouldn't change
                    # anything so skip it.
                    my $pre_existing = $r->[$j]->value;
                    if ($pre_existing ne $value) {

                        # Here the new and old standardized values are the
                        # same, but the non-standardized values aren't.  If
                        # replacing unconditionally, then replace
                        if( $replace == $UNCONDITIONALLY) {
                            $clean_insert = 0;
                        }
                        else {

                            # Here, are replacing conditionally.  Decide to
                            # replace or not based on which appears to look
                            # the "nicest".  If one is mixed case and the
                            # other isn't, choose the mixed case one.
                            my $new_mixed = $value =~ /[A-Z]/
                                            && $value =~ /[a-z]/;
                            my $old_mixed = $pre_existing =~ /[A-Z]/
                                            && $pre_existing =~ /[a-z]/;

                            if ($old_mixed != $new_mixed) {
                                $clean_insert = 0 if $new_mixed;
                                if (main::DEBUG && $to_trace) {
                                    if ($clean_insert) {
                                        trace "Retaining $pre_existing over $value";
                                    }
                                    else {
                                        trace "Replacing $pre_existing with $value";
                                    }
                                }
                            }
                            else {

                                # Here casing wasn't different between the two.
                                # If one has hyphens or underscores and the
                                # other doesn't, choose the one with the
                                # punctuation.
                                my $new_punct = $value =~ /[-_]/;
                                my $old_punct = $pre_existing =~ /[-_]/;

                                if ($old_punct != $new_punct) {
                                    $clean_insert = 0 if $new_punct;
                                    if (main::DEBUG && $to_trace) {
                                        if ($clean_insert) {
                                            trace "Retaining $pre_existing over $value";
                                        }
                                        else {
                                            trace "Replacing $pre_existing with $value";
                                        }
                                    }
                                }   # else existing one is just as "good";
                                    # retain it to save cycles.
                            }
                        }
                    }
                }
            }
        } # End of loop looking for highest affected range.

        # Here, $j points to one beyond the highest range that this insertion
        # affects (hence to beyond the range list if that range is the final
        # one in the range list).

        # The splice length is all the affected ranges.  Get it before
        # subtracting, for efficiency, so we don't have to later add 1.
        my $length = $j - $i;

        $j--;        # $j now points to the highest affected range.
        trace "Final affected range is $j: $r->[$j]" if main::DEBUG && $to_trace;

        # Here, have taken care of $NO and $MULTIPLE_foo replaces.
        # $j points to the highest affected range.  But it can be < $i or even
        # -1.  These happen only if the insertion is entirely in the gap
        # between r[$i-1] and r[$i].  Here's why: j < i means that the j loop
        # above exited first time through with $end < $r->[$i]->start.  (And
        # then we subtracted one from j)  This implies also that $start <
        # $r->[$i]->start, but we know from above that $r->[$i-1]->end <
        # $start, so the entire input range is in the gap.
        if ($j < $i) {

            # Here the entire input range is in the gap before $i.

            if (main::DEBUG && $to_trace) {
                if ($i) {
                    trace "Entire range is between $r->[$i-1] and $r->[$i]";
                }
                else {
                    trace "Entire range is before $r->[$i]";
                }
            }
            return if $operation ne '+'; # Deletion of a non-existent range is
                                         # a no-op
        }
        else {

            # Here part of the input range is not in the gap before $i.  Thus,
            # there is at least one affected one, and $j points to the highest
            # such one.

            # At this point, here is the situation:
            # This is not an insertion of a multiple, nor of tentative ($NO)
            # data.
            #   $i  points to the first element in the current range list that
            #            may be affected by this operation.  In fact, we know
            #            that the range at $i is affected because we are in
            #            the else branch of this 'if'
            #   $j  points to the highest affected range.
            # In other words,
            #   r[$i-1]->end < $start <= r[$i]->end
            # And:
            #   r[$i-1]->end < $start <= $end < r[$j+1]->start
            #
            # Also:
            #   $clean_insert is a boolean which is set true if and only if
            #        this is a "clean insertion", i.e., not a change nor a
            #        deletion (multiple was handled above).

            # We now have enough information to decide if this call is a no-op
            # or not.  It is a no-op if this is an insertion of already
            # existing data.  To be so, it must be contained entirely in one
            # range.

            if (main::DEBUG && $to_trace && $clean_insert
                                         && $start >= $r->[$i]->start
                                         && $end   <= $r->[$i]->end)
            {
                    trace "no-op";
            }
            return if $clean_insert
                      && $start >= $r->[$i]->start
                      && $end   <= $r->[$i]->end;
        }

        # Here, we know that some action will have to be taken.  We have
        # calculated the offset and length (though adjustments may be needed)
        # for the splice.  Now start constructing the replacement list.
        my @replacement;
        my $splice_start = $i;

        my $extends_below;
        my $extends_above;

        # See if should extend any adjacent ranges.
        if ($operation eq '-') { # Don't extend deletions
            $extends_below = $extends_above = 0;
        }
        else {  # Here, should extend any adjacent ranges.  See if there are
                # any.
            $extends_below = ($i > 0
                            # can't extend unless adjacent
                            && $r->[$i-1]->end == $start -1
                            # can't extend unless are same standard value
                            && $r->[$i-1]->standard_form eq $standard_form
                            # can't extend unless share type
                            && $r->[$i-1]->type == $type);
            $extends_above = ($j+1 < $range_list_size
                            && $r->[$j+1]->start == $end +1
                            && $r->[$j+1]->standard_form eq $standard_form
                            && $r->[$j+1]->type == $type);
        }
        if ($extends_below && $extends_above) { # Adds to both
            $splice_start--;     # start replace at element below
            $length += 2;        # will replace on both sides
            trace "Extends both below and above ranges" if main::DEBUG && $to_trace;

            # The result will fill in any gap, replacing both sides, and
            # create one large range.
            @replacement = Range->new($r->[$i-1]->start,
                                      $r->[$j+1]->end,
                                      Value => $value,
                                      Type => $type);
        }
        else {

            # Here we know that the result won't just be the conglomeration of
            # a new range with both its adjacent neighbors.  But it could
            # extend one of them.

            if ($extends_below) {

                # Here the new element adds to the one below, but not to the
                # one above.  If inserting, and only to that one range,  can
                # just change its ending to include the new one.
                if ($length == 0 && $clean_insert) {
                    $r->[$i-1]->set_end($end);
                    trace "inserted range extends range to below so it is now $r->[$i-1]" if main::DEBUG && $to_trace;
                    return;
                }
                else {
                    trace "Changing inserted range to start at ", sprintf("%04X",  $r->[$i-1]->start), " instead of ", sprintf("%04X", $start) if main::DEBUG && $to_trace;
                    $splice_start--;        # start replace at element below
                    $length++;              # will replace the element below
                    $start = $r->[$i-1]->start;
                }
            }
            elsif ($extends_above) {

                # Here the new element adds to the one above, but not below.
                # Mirror the code above
                if ($length == 0 && $clean_insert) {
                    $r->[$j+1]->set_start($start);
                    trace "inserted range extends range to above so it is now $r->[$j+1]" if main::DEBUG && $to_trace;
                    return;
                }
                else {
                    trace "Changing inserted range to end at ", sprintf("%04X",  $r->[$j+1]->end), " instead of ", sprintf("%04X", $end) if main::DEBUG && $to_trace;
                    $length++;        # will replace the element above
                    $end = $r->[$j+1]->end;
                }
            }

            trace "Range at $i is $r->[$i]" if main::DEBUG && $to_trace;

            # Finally, here we know there will have to be a splice.
            # If the change or delete affects only the highest portion of the
            # first affected range, the range will have to be split.  The
            # splice will remove the whole range, but will replace it by a new
            # range containing just the unaffected part.  So, in this case,
            # add to the replacement list just this unaffected portion.
            if (! $extends_below
                && $start > $r->[$i]->start && $start <= $r->[$i]->end)
            {
                push @replacement,
                    Range->new($r->[$i]->start,
                               $start - 1,
                               Value => $r->[$i]->value,
                               Type => $r->[$i]->type);
            }

            # In the case of an insert or change, but not a delete, we have to
            # put in the new stuff;  this comes next.
            if ($operation eq '+') {
                push @replacement, Range->new($start,
                                              $end,
                                              Value => $value,
                                              Type => $type);
            }

            trace "Range at $j is $r->[$j]" if main::DEBUG && $to_trace && $j != $i;
            #trace "$end >=", $r->[$j]->start, " && $end <", $r->[$j]->end if main::DEBUG && $to_trace;

            # And finally, if we're changing or deleting only a portion of the
            # highest affected range, it must be split, as the lowest one was.
            if (! $extends_above
                && $j >= 0  # Remember that j can be -1 if before first
                            # current element
                && $end >= $r->[$j]->start
                && $end < $r->[$j]->end)
            {
                push @replacement,
                    Range->new($end + 1,
                               $r->[$j]->end,
                               Value => $r->[$j]->value,
                               Type => $r->[$j]->type);
            }
        }

        # And do the splice, as calculated above
        if (main::DEBUG && $to_trace) {
            trace "replacing $length element(s) at $i with ";
            foreach my $replacement (@replacement) {
                trace "    $replacement";
            }
            trace "Before splice:";
            trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
            trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
            trace "i  =[", $i, "]", $r->[$i];
            trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
            trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
        }

        my @return = splice @$r, $splice_start, $length, @replacement;

        if (main::DEBUG && $to_trace) {
            trace "After splice:";
            trace 'i-2=[', $i-2, ']', $r->[$i-2] if $i >= 2;
            trace 'i-1=[', $i-1, ']', $r->[$i-1] if $i >= 1;
            trace "i  =[", $i, "]", $r->[$i];
            trace 'i+1=[', $i+1, ']', $r->[$i+1] if $i < @$r - 1;
            trace 'i+2=[', $i+2, ']', $r->[$i+2] if $i < @$r - 2;
            trace "removed ", @return if @return;
        }

        # An actual deletion could have changed the maximum in the list.
        # There was no deletion if the splice didn't return something, but
        # otherwise recalculate it.  This is done too rarely to worry about
        # performance.
        if ($operation eq '-' && @return) {
            if (@$r) {
                $max{$addr} = $r->[-1]->end;
            }
            else {  # Now empty
                $max{$addr} = $max_init;
            }
        }
        return @return;
    }

    sub reset_each_range($self) {  # reset the iterator for each_range();
        undef $each_range_iterator{pack 'J', refaddr $self};
        return;
    }

    sub each_range($self) {
        # Iterate over each range in a range list.  Results are undefined if
        # the range list is changed during the iteration.
        my $addr = pack 'J', refaddr $self;

        return if $self->is_empty;

        $each_range_iterator{$addr} = -1
                                if ! defined $each_range_iterator{$addr};
        $each_range_iterator{$addr}++;
        return $ranges{$addr}->[$each_range_iterator{$addr}]
                        if $each_range_iterator{$addr} < @{$ranges{$addr}};
        undef $each_range_iterator{$addr};
        return;
    }

    sub count($self) {        # Returns count of code points in range list
        my $addr = pack 'J', refaddr $self;

        my $count = 0;
        foreach my $range (@{$ranges{$addr}}) {
            $count += $range->end - $range->start + 1;
        }
        return $count;
    }

    sub delete_range($self, $start, $end) {    # Delete a range
        return $self->_add_delete('-', $start, $end, "");
    }

    sub is_empty($self) { # Returns boolean as to if a range list is empty
        return scalar @{$ranges{pack 'J', refaddr $self}} == 0;
    }

    sub hash($self) {
        # Quickly returns a scalar suitable for separating tables into
        # buckets, i.e. it is a hash function of the contents of a table, so
        # there are relatively few conflicts.
        my $addr = pack 'J', refaddr $self;

        # These are quickly computable.  Return looks like 'min..max;count'
        return $self->min . "..$max{$addr};" . scalar @{$ranges{$addr}};
    }
} # End closure for _Range_List_Base

package Range_List;
use parent '-norequire', '_Range_List_Base';

# A Range_List is a range list for match tables; i.e. the range values are
# not significant.  Thus a number of operations can be safely added to it,
# such as inversion, intersection.  Note that union is also an unsafe
# operation when range values are cared about, and that method is in the base
# class, not here.  But things are set up so that that method is callable only
# during initialization.  Only in this derived class, is there an operation
# that combines two tables.  A Range_Map can thus be used to initialize a
# Range_List, and its mappings will be in the list, but are not significant to
# this class.

sub trace { return main::trace(@_); }

{ # Closure

    use overload
        fallback => 0,
        '+' => sub { my $self = shift;
                    my $other = shift;

                    return $self->_union($other)
                },
        '+=' => sub { my $self = shift;
                    my $other = shift;
                    my $reversed = shift;

                    if ($reversed) {
                        Carp::my_carp_bug("Bad news.  Can't cope with '"
                        . ref($other)
                        . ' += '
                        . ref($self)
                        . "'.  undef returned.");
                        return;
                    }

                    return $self->_union($other)
                },
        '&' => sub { my $self = shift;
                    my $other = shift;

                    return $self->_intersect($other, 0);
                },
        '&=' => sub { my $self = shift;
                    my $other = shift;
                    my $reversed = shift;

                    if ($reversed) {
                        Carp::my_carp_bug("Bad news.  Can't cope with '"
                        . ref($other)
                        . ' &= '
                        . ref($self)
                        . "'.  undef returned.");
                        return;
                    }

                    return $self->_intersect($other, 0);
                },
        '~' => "_invert",
        '-' => "_subtract",
    ;

    sub _invert($self, @) {
        # Returns a new Range_List that gives all code points not in $self.
        my $new = Range_List->new;

        # Go through each range in the table, finding the gaps between them
        my $max = -1;   # Set so no gap before range beginning at 0
        for my $range ($self->ranges) {
            my $start = $range->start;
            my $end   = $range->end;

            # If there is a gap before this range, the inverse will contain
            # that gap.
            if ($start > $max + 1) {
                $new->add_range($max + 1, $start - 1);
            }
            $max = $end;
        }

        # And finally, add the gap from the end of the table to the max
        # possible code point
        if ($max < $MAX_WORKING_CODEPOINT) {
            $new->add_range($max + 1, $MAX_WORKING_CODEPOINT);
        }
        return $new;
    }

    sub _subtract($self, $other, $reversed=0) {
        # Returns a new Range_List with the argument deleted from it.  The
        # argument can be a single code point, a range, or something that has
        # a range, with the _range_list() method on it returning them

        if ($reversed) {
            Carp::my_carp_bug("Bad news.  Can't cope with '"
            . ref($other)
            . ' - '
            . ref($self)
            . "'.  undef returned.");
            return;
        }

        my $new = Range_List->new(Initialize => $self);

        if (! ref $other) { # Single code point
            $new->delete_range($other, $other);
        }
        elsif ($other->isa('Range')) {
            $new->delete_range($other->start, $other->end);
        }
        elsif ($other->can('_range_list')) {
            foreach my $range ($other->_range_list->ranges) {
                $new->delete_range($range->start, $range->end);
            }
        }
        else {
            Carp::my_carp_bug("Can't cope with a "
                        . ref($other)
                        . " argument to '-'.  Subtraction ignored."
                        );
            return $self;
        }

        return $new;
    }

    sub _intersect($a_object, $b_object, $check_if_overlapping=0) {
        # Returns either a boolean giving whether the two inputs' range lists
        # intersect (overlap), or a new Range_List containing the intersection
        # of the two lists.  The optional final parameter being true indicates
        # to do the check instead of the intersection.

        if (! defined $b_object) {
            my $message = "";
            $message .= $a_object->_owner_name_of if defined $a_object;
            Carp::my_carp_bug($message .= "Called with undefined value.  Intersection not done.");
            return;
        }

        # a & b = !(!a | !b), or in our terminology = ~ ( ~a + -b )
        # Thus the intersection could be much more simply be written:
        #   return ~(~$a_object + ~$b_object);
        # But, this is slower, and when taking the inverse of a large
        # range_size_1 table, back when such tables were always stored that
        # way, it became prohibitively slow, hence the code was changed to the
        # below

        if ($b_object->isa('Range')) {
            $b_object = Range_List->new(Initialize => $b_object,
                                        Owner => $a_object->_owner_name_of);
        }
        $b_object = $b_object->_range_list if $b_object->can('_range_list');

        my @a_ranges = $a_object->ranges;
        my @b_ranges = $b_object->ranges;

        #local $to_trace = 1 if main::DEBUG;
        trace "intersecting $a_object with ", scalar @a_ranges, "ranges and $b_object with", scalar @b_ranges, " ranges" if main::DEBUG && $to_trace;

        # Start with the first range in each list
        my $a_i = 0;
        my $range_a = $a_ranges[$a_i];
        my $b_i = 0;
        my $range_b = $b_ranges[$b_i];

        my $new = __PACKAGE__->new(Owner => $a_object->_owner_name_of)
                                                if ! $check_if_overlapping;

        # If either list is empty, there is no intersection and no overlap
        if (! defined $range_a || ! defined $range_b) {
            return $check_if_overlapping ? 0 : $new;
        }
        trace "range_a[$a_i]=$range_a; range_b[$b_i]=$range_b" if main::DEBUG && $to_trace;

        # Otherwise, must calculate the intersection/overlap.  Start with the
        # very first code point in each list
        my $a = $range_a->start;
        my $b = $range_b->start;

        # Loop through all the ranges of each list; in each iteration, $a and
        # $b are the current code points in their respective lists
        while (1) {

            # If $a and $b are the same code point, ...
            if ($a == $b) {

                # it means the lists overlap.  If just checking for overlap
                # know the answer now,
                return 1 if $check_if_overlapping;

                # The intersection includes this code point plus anything else
                # common to both current ranges.
                my $start = $a;
                my $end = main::min($range_a->end, $range_b->end);
                if (! $check_if_overlapping) {
                    trace "adding intersection range ", sprintf("%04X", $start) . ".." . sprintf("%04X", $end) if main::DEBUG && $to_trace;
                    $new->add_range($start, $end);
                }

                # Skip ahead to the end of the current intersect
                $a = $b = $end;

                # If the current intersect ends at the end of either range (as
                # it must for at least one of them), the next possible one
                # will be the beginning code point in it's list's next range.
                if ($a == $range_a->end) {
                    $range_a = $a_ranges[++$a_i];
                    last unless defined $range_a;
                    $a = $range_a->start;
                }
                if ($b == $range_b->end) {
                    $range_b = $b_ranges[++$b_i];
                    last unless defined $range_b;
                    $b = $range_b->start;
                }

                trace "range_a[$a_i]=$range_a; range_b[$b_i]=$range_b" if main::DEBUG && $to_trace;
            }
            elsif ($a < $b) {

                # Not equal, but if the range containing $a encompasses $b,
                # change $a to be the middle of the range where it does equal
                # $b, so the next iteration will get the intersection
                if ($range_a->end >= $b) {
                    $a = $b;
                }
                else {

                    # Here, the current range containing $a is entirely below
                    # $b.  Go try to find a range that could contain $b.
                    $a_i = $a_object->_search_ranges($b);

                    # If no range found, quit.
                    last unless defined $a_i;

                    # The search returns $a_i, such that
                    #   range_a[$a_i-1]->end < $b <= range_a[$a_i]->end
                    # Set $a to the beginning of this new range, and repeat.
                    $range_a = $a_ranges[$a_i];
                    $a = $range_a->start;
                }
            }
            else { # Here, $b < $a.

                # Mirror image code to the leg just above
                if ($range_b->end >= $a) {
                    $b = $a;
                }
                else {
                    $b_i = $b_object->_search_ranges($a);
                    last unless defined $b_i;
                    $range_b = $b_ranges[$b_i];
                    $b = $range_b->start;
                }
            }
        } # End of looping through ranges.

        # Intersection fully computed, or now know that there is no overlap
        return $check_if_overlapping ? 0 : $new;
    }

    sub overlaps($self, $other) {
        # Returns boolean giving whether the two arguments overlap somewhere
        return $self->_intersect($other, 1);
    }

    sub add_range($self, $start, $end) {
        # Add a range to the list.
        return $self->_add_delete('+', $start, $end, "");
    }

    sub matches_identically_to($self, $other) {
        # Return a boolean as to whether or not two Range_Lists match identical
        # sets of code points.
        # These are ordered in increasing real time to figure out (at least
        # until a patch changes that and doesn't change this)
        return 0 if $self->max != $other->max;
        return 0 if $self->min != $other->min;
        return 0 if $self->range_count != $other->range_count;
        return 0 if $self->count != $other->count;

        # Here they could be identical because all the tests above passed.
        # The loop below is somewhat simpler since we know they have the same
        # number of elements.  Compare range by range, until reach the end or
        # find something that differs.
        my @a_ranges = $self->ranges;
        my @b_ranges = $other->ranges;
        for my $i (0 .. @a_ranges - 1) {
            my $a = $a_ranges[$i];
            my $b = $b_ranges[$i];
            trace "self $a; other $b" if main::DEBUG && $to_trace;
            return 0 if ! defined $b
                        || $a->start != $b->start
                        || $a->end != $b->end;
        }
        return 1;
    }

    sub is_code_point_usable($code, $try_hard) {
        # This used only for making the test script.  See if the input
        # proposed trial code point is one that Perl will handle.  If second
        # parameter is 0, it won't select some code points for various
        # reasons, noted below.
        return 0 if $code < 0;                # Never use a negative

        # shun null.  I'm (khw) not sure why this was done, but NULL would be
        # the character very frequently used.
        return $try_hard if $code == 0x0000;

        # shun non-character code points.
        return $try_hard if $code >= 0xFDD0 && $code <= 0xFDEF;
        return $try_hard if ($code & 0xFFFE) == 0xFFFE; # includes FFFF

        return $try_hard if $code > $MAX_UNICODE_CODEPOINT;   # keep in range
        return $try_hard if $code >= 0xD800 && $code <= 0xDFFF; # no surrogate

        return 1;
    }

    sub get_valid_code_point($self) {
        # Return a code point that's part of the range list.  Returns nothing
        # if the table is empty or we can't find a suitable code point.  This
        # used only for making the test script.

        # On first pass, don't choose less desirable code points; if no good
        # one is found, repeat, allowing a less desirable one to be selected.
        for my $try_hard (0, 1) {

            # Look through all the ranges for a usable code point.
            for my $set (reverse $self->ranges) {

                # Try the edge cases first, starting with the end point of the
                # range.
                my $end = $set->end;
                return $end if is_code_point_usable($end, $try_hard);
                $end = $MAX_UNICODE_CODEPOINT + 1 if $end > $MAX_UNICODE_CODEPOINT;

                # End point didn't, work.  Start at the beginning and try
                # every one until find one that does work.
                for my $trial ($set->start .. $end - 1) {
                    return $trial if is_code_point_usable($trial, $try_hard);
                }
            }
        }
        return ();  # If none found, give up.
    }

    sub get_invalid_code_point($self) {
        # Return a code point that's not part of the table.  Returns nothing
        # if the table covers all code points or a suitable code point can't
        # be found.  This used only for making the test script.

        # Just find a valid code point of the inverse, if any.
        return Range_List->new(Initialize => ~ $self)->get_valid_code_point;
    }
} # end closure for Range_List

package Range_Map;
use parent '-norequire', '_Range_List_Base';

# A Range_Map is a range list in which the range values (called maps) are
# significant, and hence shouldn't be manipulated by our other code, which
# could be ambiguous or lose things.  For example, in taking the union of two
# lists, which share code points, but which have differing values, which one
# has precedence in the union?
# It turns out that these operations aren't really necessary for map tables,
# and so this class was created to make sure they aren't accidentally
# applied to them.

{ # Closure

    sub add_map($self, @add) {
        # Add a range containing a mapping value to the list
        return $self->_add_delete('+', @add);
    }

    sub replace_map($self, @list) {
        # Replace a range
        return $self->_add_delete('+', @list, Replace => $UNCONDITIONALLY);
    }

    sub add_duplicate {
        # Adds entry to a range list which can duplicate an existing entry

        my $self = shift;
        my $code_point = shift;
        my $value = shift;
        my %args = @_;
        my $replace = delete $args{'Replace'} // $MULTIPLE_BEFORE;
        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        return $self->add_map($code_point, $code_point,
                                $value, Replace => $replace);
    }
} # End of closure for package Range_Map

package _Base_Table;

# A table is the basic data structure that gets written out into a file for
# use by the Perl core.  This is the abstract base class implementing the
# common elements from the derived ones.  A list of the methods to be
# furnished by an implementing class is just after the constructor.

sub standardize { return main::standardize($_[0]); }
sub trace { return main::trace(@_); }

{ # Closure

    main::setup_package();

    my %range_list;
    # Object containing the ranges of the table.
    main::set_access('range_list', \%range_list, 'p_r', 'p_s');

    my %full_name;
    # The full table name.
    main::set_access('full_name', \%full_name, 'r');

    my %name;
    # The table name, almost always shorter
    main::set_access('name', \%name, 'r');

    my %short_name;
    # The shortest of all the aliases for this table, with underscores removed
    main::set_access('short_name', \%short_name);

    my %nominal_short_name_length;
    # The length of short_name before removing underscores
    main::set_access('nominal_short_name_length',
                    \%nominal_short_name_length);

    my %complete_name;
    # The complete name, including property.
    main::set_access('complete_name', \%complete_name, 'r');

    my %property;
    # Parent property this table is attached to.
    main::set_access('property', \%property, 'r');

    my %aliases;
    # Ordered list of alias objects of the table's name.  The first ones in
    # the list are output first in comments
    main::set_access('aliases', \%aliases, 'readable_array');

    my %comment;
    # A comment associated with the table for human readers of the files
    main::set_access('comment', \%comment, 's');

    my %description;
    # A comment giving a short description of the table's meaning for human
    # readers of the files.
    main::set_access('description', \%description, 'readable_array');

    my %note;
    # A comment giving a short note about the table for human readers of the
    # files.
    main::set_access('note', \%note, 'readable_array');

    my %fate;
    # Enum; there are a number of possibilities for what happens to this
    # table: it could be normal, or suppressed, or not for external use.  See
    # values at definition for $SUPPRESSED.
    main::set_access('fate', \%fate, 'r');

    my %find_table_from_alias;
    # The parent property passes this pointer to a hash which this class adds
    # all its aliases to, so that the parent can quickly take an alias and
    # find this table.
    main::set_access('find_table_from_alias', \%find_table_from_alias, 'p_r');

    my %locked;
    # After this table is made equivalent to another one; we shouldn't go
    # changing the contents because that could mean it's no longer equivalent
    main::set_access('locked', \%locked, 'r');

    my %file_path;
    # This gives the final path to the file containing the table.  Each
    # directory in the path is an element in the array
    main::set_access('file_path', \%file_path, 'readable_array');

    my %status;
    # What is the table's status, normal, $OBSOLETE, etc.  Enum
    main::set_access('status', \%status, 'r');

    my %status_info;
    # A comment about its being obsolete, or whatever non normal status it has
    main::set_access('status_info', \%status_info, 'r');

    my %caseless_equivalent;
    # The table this is equivalent to under /i matching, if any.
    main::set_access('caseless_equivalent', \%caseless_equivalent, 'r', 's');

    my %range_size_1;
    # Is the table to be output with each range only a single code point?
    # This is done to avoid breaking existing code that may have come to rely
    # on this behavior in previous versions of this program.)
    main::set_access('range_size_1', \%range_size_1, 'r', 's');

    my %perl_extension;
    # A boolean set iff this table is a Perl extension to the Unicode
    # standard.
    main::set_access('perl_extension', \%perl_extension, 'r');

    my %output_range_counts;
    # A boolean set iff this table is to have comments written in the
    # output file that contain the number of code points in the range.
    # The constructor can override the global flag of the same name.
    main::set_access('output_range_counts', \%output_range_counts, 'r');

    my %write_as_invlist;
    # A boolean set iff the output file for this table is to be in the form of
    # an inversion list/map.
    main::set_access('write_as_invlist', \%write_as_invlist, 'r');

    my %format;
    # The format of the entries of the table.  This is calculated from the
    # data in the table (or passed in the constructor).  This is an enum e.g.,
    # $STRING_FORMAT.  It is marked protected as it should not be generally
    # used to override calculations.
    main::set_access('format', \%format, 'r', 'p_s');

    my %has_dependency;
    # A boolean that gives whether some other table in this property is
    # defined as the complement of this table.  This is a crude, but currently
    # sufficient, mechanism to make this table not get destroyed before what
    # is dependent on it is.  Other dependencies could be added, so the name
    # was chosen to reflect a more general situation than actually is
    # currently the case.
    main::set_access('has_dependency', \%has_dependency, 'r', 's');

    sub new {
        # All arguments are key => value pairs, which you can see below, most
        # of which match fields documented above.  Otherwise: Re_Pod_Entry,
        # OK_as_Filename, and Fuzzy apply to the names of the table, and are
        # documented in the Alias package

        return Carp::carp_too_few_args(\@_, 2) if main::DEBUG && @_ < 2;

        my $class = shift;

        my $self = bless \do { my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        my %args = @_;

        $name{$addr} = delete $args{'Name'};
        $find_table_from_alias{$addr} = delete $args{'_Alias_Hash'};
        $full_name{$addr} = delete $args{'Full_Name'};
        my $complete_name = $complete_name{$addr}
                          = delete $args{'Complete_Name'};
        $format{$addr} = delete $args{'Format'};
        $output_range_counts{$addr} = delete $args{'Output_Range_Counts'};
        $property{$addr} = delete $args{'_Property'};
        $range_list{$addr} = delete $args{'_Range_List'};
        $status{$addr} = delete $args{'Status'} || $NORMAL;
        $status_info{$addr} = delete $args{'_Status_Info'} || "";
        $range_size_1{$addr} = delete $args{'Range_Size_1'} || 0;
        $caseless_equivalent{$addr} = delete $args{'Caseless_Equivalent'} || 0;
        $fate{$addr} = delete $args{'Fate'} || $ORDINARY;
        $write_as_invlist{$addr} = delete $args{'Write_As_Invlist'};# No default
        my $ucd = delete $args{'UCD'};

        my $description = delete $args{'Description'};
        my $ok_as_filename = delete $args{'OK_as_Filename'};
        my $loose_match = delete $args{'Fuzzy'};
        my $note = delete $args{'Note'};
        my $make_re_pod_entry = delete $args{'Re_Pod_Entry'};
        my $perl_extension = delete $args{'Perl_Extension'};
        my $suppression_reason = delete $args{'Suppression_Reason'};

        # Shouldn't have any left over
        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        # Can't use || above because conceivably the name could be 0, and
        # can't use // operator in case this program gets used in Perl 5.8
        $full_name{$addr} = $name{$addr} if ! defined $full_name{$addr};
        $output_range_counts{$addr} = $output_range_counts if
                                        ! defined $output_range_counts{$addr};

        $aliases{$addr} = [ ];
        $comment{$addr} = [ ];
        $description{$addr} = [ ];
        $note{$addr} = [ ];
        $file_path{$addr} = [ ];
        $locked{$addr} = "";
        $has_dependency{$addr} = 0;

        push @{$description{$addr}}, $description if $description;
        push @{$note{$addr}}, $note if $note;

        if ($fate{$addr} == $PLACEHOLDER) {

            # A placeholder table doesn't get documented, is a perl extension,
            # and quite likely will be empty
            $make_re_pod_entry = 0 if ! defined $make_re_pod_entry;
            $perl_extension = 1 if ! defined $perl_extension;
            $ucd = 0 if ! defined $ucd;
            push @tables_that_may_be_empty, $complete_name{$addr};
            $self->add_comment(<<END);
This is a placeholder because it is not in Version $string_version of Unicode,
but is needed by the Perl core to work gracefully.  Because it is not in this
version of Unicode, it will not be listed in $pod_file.pod
END
        }
        elsif (exists $why_suppressed{$complete_name}
                # Don't suppress if overridden
                && ! grep { $_ eq $complete_name{$addr} }
                                                    @output_mapped_properties)
        {
            $fate{$addr} = $SUPPRESSED;
        }
        elsif ($fate{$addr} == $SUPPRESSED) {
            Carp::my_carp_bug("Need reason for suppressing") unless $suppression_reason;
            # Though currently unused
        }
        elsif ($suppression_reason) {
            Carp::my_carp_bug("A reason was given for suppressing, but not suppressed");
        }

        # If hasn't set its status already, see if it is on one of the
        # lists of properties or tables that have particular statuses; if
        # not, is normal.  The lists are prioritized so the most serious
        # ones are checked first
        if (! $status{$addr}) {
            if (exists $why_deprecated{$complete_name}) {
                $status{$addr} = $DEPRECATED;
            }
            elsif (exists $why_stabilized{$complete_name}) {
                $status{$addr} = $STABILIZED;
            }
            elsif (exists $why_obsolete{$complete_name}) {
                $status{$addr} = $OBSOLETE;
            }

            # Existence above doesn't necessarily mean there is a message
            # associated with it.  Use the most serious message.
            if ($status{$addr}) {
                if ($why_deprecated{$complete_name}) {
                    $status_info{$addr}
                                = $why_deprecated{$complete_name};
                }
                elsif ($why_stabilized{$complete_name}) {
                    $status_info{$addr}
                                = $why_stabilized{$complete_name};
                }
                elsif ($why_obsolete{$complete_name}) {
                    $status_info{$addr}
                                = $why_obsolete{$complete_name};
                }
            }
        }

        $perl_extension{$addr} = $perl_extension || 0;

        # Don't list a property by default that is internal only
        if ($fate{$addr} > $MAP_PROXIED) {
            $make_re_pod_entry = 0 if ! defined $make_re_pod_entry;
            $ucd = 0 if ! defined $ucd;
        }
        else {
            $ucd = 1 if ! defined $ucd;
        }

        # By convention what typically gets printed only or first is what's
        # first in the list, so put the full name there for good output
        # clarity.  Other routines rely on the full name being first on the
        # list
        $self->add_alias($full_name{$addr},
                            OK_as_Filename => $ok_as_filename,
                            Fuzzy => $loose_match,
                            Re_Pod_Entry => $make_re_pod_entry,
                            Status => $status{$addr},
                            UCD => $ucd,
                            );

        # Then comes the other name, if meaningfully different.
        if (standardize($full_name{$addr}) ne standardize($name{$addr})) {
            $self->add_alias($name{$addr},
                            OK_as_Filename => $ok_as_filename,
                            Fuzzy => $loose_match,
                            Re_Pod_Entry => $make_re_pod_entry,
                            Status => $status{$addr},
                            UCD => $ucd,
                            );
        }

        return $self;
    }

    # Here are the methods that are required to be defined by any derived
    # class
    for my $sub (qw(
                    handle_special_range
                    append_to_body
                    pre_body
                ))
                # write() knows how to write out normal ranges, but it calls
                # handle_special_range() when it encounters a non-normal one.
                # append_to_body() is called by it after it has handled all
                # ranges to add anything after the main portion of the table.
                # And finally, pre_body() is called after all this to build up
                # anything that should appear before the main portion of the
                # table.  Doing it this way allows things in the middle to
                # affect what should appear before the main portion of the
                # table.
    {
        no strict "refs";
        *$sub = sub {
            Carp::my_carp_bug( __LINE__
                              . ": Must create method '$sub()' for "
                              . ref shift);
            return;
        }
    }

    use overload
        fallback => 0,
        "." => \&main::_operator_dot,
        ".=" => \&main::_operator_dot_equal,
        '!=' => \&main::_operator_not_equal,
        '==' => \&main::_operator_equal,
    ;

    sub ranges {
        # Returns the array of ranges associated with this table.

        return $range_list{pack 'J', refaddr shift}->ranges;
    }

    sub add_alias {
        # Add a synonym for this table.

        return Carp::carp_too_few_args(\@_, 3) if main::DEBUG && @_ < 3;

        my $self = shift;
        my $name = shift;       # The name to add.
        my $pointer = shift;    # What the alias hash should point to.  For
                                # map tables, this is the parent property;
                                # for match tables, it is the table itself.

        my %args = @_;
        my $loose_match = delete $args{'Fuzzy'};

        my $ok_as_filename = delete $args{'OK_as_Filename'};
        $ok_as_filename = 1 unless defined $ok_as_filename;

        # An internal name does not get documented, unless overridden by the
        # input; same for making tests for it.
        my $status = delete $args{'Status'} || (($name =~ /^_/)
                                                ? $INTERNAL_ALIAS
                                                : $NORMAL);
        my $make_re_pod_entry = delete $args{'Re_Pod_Entry'}
                                            // (($status ne $INTERNAL_ALIAS)
                                               ? (($name =~ /^_/) ? $NO : $YES)
                                               : $NO);
        my $ucd = delete $args{'UCD'} // (($name =~ /^_/) ? 0 : 1);

        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        # Capitalize the first letter of the alias unless it is one of the CJK
        # ones which specifically begins with a lower 'k'.  Do this because
        # Unicode has varied whether they capitalize first letters or not, and
        # have later changed their minds and capitalized them, but not the
        # other way around.  So do it always and avoid changes from release to
        # release
        $name = ucfirst($name) unless $name =~ /^k[A-Z]/;

        my $addr = pack 'J', refaddr $self;

        # Figure out if should be loosely matched if not already specified.
        if (! defined $loose_match) {

            # Is a loose_match if isn't null, and doesn't begin with an
            # underscore and isn't just a number
            if ($name ne ""
                && substr($name, 0, 1) ne '_'
                && $name !~ qr{^[0-9_.+-/]+$})
            {
                $loose_match = 1;
            }
            else {
                $loose_match = 0;
            }
        }

        # If this alias has already been defined, do nothing.
        return if defined $find_table_from_alias{$addr}->{$name};

        # That includes if it is standardly equivalent to an existing alias,
        # in which case, add this name to the list, so won't have to search
        # for it again.
        my $standard_name = main::standardize($name);
        if (defined $find_table_from_alias{$addr}->{$standard_name}) {
            $find_table_from_alias{$addr}->{$name}
                        = $find_table_from_alias{$addr}->{$standard_name};
            return;
        }

        # Set the index hash for this alias for future quick reference.
        $find_table_from_alias{$addr}->{$name} = $pointer;
        $find_table_from_alias{$addr}->{$standard_name} = $pointer;
        local $to_trace = 0 if main::DEBUG;
        trace "adding alias $name to $pointer" if main::DEBUG && $to_trace;
        trace "adding alias $standard_name to $pointer" if main::DEBUG && $to_trace;


        # Put the new alias at the end of the list of aliases unless the final
        # element begins with an underscore (meaning it is for internal perl
        # use) or is all numeric, in which case, put the new one before that
        # one.  This floats any all-numeric or underscore-beginning aliases to
        # the end.  This is done so that they are listed last in output lists,
        # to encourage the user to use a better name (either more descriptive
        # or not an internal-only one) instead.  This ordering is relied on
        # implicitly elsewhere in this program, like in short_name()
        my $list = $aliases{$addr};
        my $insert_position = (@$list == 0
                                || (substr($list->[-1]->name, 0, 1) ne '_'
                                    && $list->[-1]->name =~ /\D/))
                            ? @$list
                            : @$list - 1;
        splice @$list,
                $insert_position,
                0,
                Alias->new($name, $loose_match, $make_re_pod_entry,
                           $ok_as_filename, $status, $ucd);

        # This name may be shorter than any existing ones, so clear the cache
        # of the shortest, so will have to be recalculated.
        undef $short_name{pack 'J', refaddr $self};
        return;
    }

    sub short_name($self, $nominal_length_ptr=undef) {
        # Returns a name suitable for use as the base part of a file name.
        # That is, shorter wins.  It can return undef if there is no suitable
        # name.  The name has all non-essential underscores removed.

        # The optional second parameter is a reference to a scalar in which
        # this routine will store the length the returned name had before the
        # underscores were removed, or undef if the return is undef.

        # The shortest name can change if new aliases are added.  So using
        # this should be deferred until after all these are added.  The code
        # that does that should clear this one's cache.
        # Any name with alphabetics is preferred over an all numeric one, even
        # if longer.

        my $addr = pack 'J', refaddr $self;

        # For efficiency, don't recalculate, but this means that adding new
        # aliases could change what the shortest is, so the code that does
        # that needs to undef this.
        if (defined $short_name{$addr}) {
            if ($nominal_length_ptr) {
                $$nominal_length_ptr = $nominal_short_name_length{$addr};
            }
            return $short_name{$addr};
        }

        # Look at each alias
        my $is_last_resort = 0;
        my $deprecated_or_discouraged
                                = qr/ ^ (?: $DEPRECATED | $DISCOURAGED ) $/x;
        foreach my $alias ($self->aliases()) {

            # Don't use an alias that isn't ok to use for an external name.
            next if ! $alias->ok_as_filename;

            my $name = main::Standardize($alias->name);
            trace $self, $name if main::DEBUG && $to_trace;

            # Take the first one, or any non-deprecated non-discouraged one
            # over one that is, or a shorter one that isn't numeric.  This
            # relies on numeric aliases always being last in the array
            # returned by aliases().  Any alpha one will have precedence.
            if (   ! defined $short_name{$addr}
                || (   $is_last_resort
                    && $alias->status !~ $deprecated_or_discouraged)
                || ($name =~ /\D/
                    && length($name) < length($short_name{$addr})))
            {
                # Remove interior underscores.
                ($short_name{$addr} = $name) =~ s/ (?<= . ) _ (?= . ) //xg;

                $nominal_short_name_length{$addr} = length $name;
                $is_last_resort = $alias->status =~ $deprecated_or_discouraged;
            }
        }

        # If the short name isn't a nice one, perhaps an equivalent table has
        # a better one.
        if (   $self->can('children')
            && (   ! defined $short_name{$addr}
                || $short_name{$addr} eq ""
                || $short_name{$addr} eq "_"))
        {
            my $return;
            foreach my $follower ($self->children) {    # All equivalents
                my $follower_name = $follower->short_name;
                next unless defined $follower_name;

                # Anything (except undefined) is better than underscore or
                # empty
                if (! defined $return || $return eq "_") {
                    $return = $follower_name;
                    next;
                }

                # If the new follower name isn't "_" and is shorter than the
                # current best one, prefer the new one.
                next if $follower_name eq "_";
                next if length $follower_name > length $return;
                $return = $follower_name;
            }
            $short_name{$addr} = $return if defined $return;
        }

        # If no suitable external name return undef
        if (! defined $short_name{$addr}) {
            $$nominal_length_ptr = undef if $nominal_length_ptr;
            return;
        }

        # Don't allow a null short name.
        if ($short_name{$addr} eq "") {
            $short_name{$addr} = '_';
            $nominal_short_name_length{$addr} = 1;
        }

        trace $self, $short_name{$addr} if main::DEBUG && $to_trace;

        if ($nominal_length_ptr) {
            $$nominal_length_ptr = $nominal_short_name_length{$addr};
        }
        return $short_name{$addr};
    }

    sub external_name($self) {
        # Returns the external name that this table should be known by.  This
        # is usually the short_name, but not if the short_name is undefined,
        # in which case the external_name is arbitrarily set to the
        # underscore.

        my $short = $self->short_name;
        return $short if defined $short;

        return '_';
    }

    sub add_description($self, $description) { # Adds the parameter as a short description.
        push @{$description{pack 'J', refaddr $self}}, $description;

        return;
    }

    sub add_note($self, $note) { # Adds the parameter as a short note.
        push @{$note{pack 'J', refaddr $self}}, $note;

        return;
    }

    sub add_comment($self, $comment) { # Adds the parameter as a comment.

        return unless $debugging_build;

        chomp $comment;

        push @{$comment{pack 'J', refaddr $self}}, $comment;

        return;
    }

    sub comment($self) {
        # Return the current comment for this table.  If called in list
        # context, returns the array of comments.  In scalar, returns a string
        # of each element joined together with a period ending each.

        my $addr = pack 'J', refaddr $self;
        my @list = @{$comment{$addr}};
        return @list if wantarray;
        my $return = "";
        foreach my $sentence (@list) {
            $return .= '.  ' if $return;
            $return .= $sentence;
            $return =~ s/\.$//;
        }
        $return .= '.' if $return;
        return $return;
    }

    sub initialize($self, $initialization) {
        # Initialize the table with the argument which is any valid
        # initialization for range lists.

        my $addr = pack 'J', refaddr $self;

        # Replace the current range list with a new one of the same exact
        # type.
        my $class = ref $range_list{$addr};
        $range_list{$addr} = $class->new(Owner => $self,
                                        Initialize => $initialization);
        return;

    }

    sub header($self) {
        # The header that is output for the table in the file it is written
        # in.
        my $return = "";
        $return .= $DEVELOPMENT_ONLY if $compare_versions;
        $return .= $HEADER;
        return $return;
    }

    sub merge_single_annotation_line ($output, $annotation, $annotation_column) {

        # This appends an annotation comment, $annotation, to $output,
        # starting in or after column $annotation_column, removing any
        # pre-existing comment from $output.

        $annotation =~ s/^ \s* \# \  //x;
        $output =~ s/ \s* ( \# \N* )? \n //x;
        $output = Text::Tabs::expand($output);

        my $spaces = $annotation_column - length $output;
        $spaces = 2 if $spaces < 0;  # Have 2 blanks before the comment

        $output = sprintf "%s%*s# %s",
                            $output,
                            $spaces,
                            " ",
                            $annotation;
        return Text::Tabs::unexpand $output;
    }

    sub write($self, $use_adjustments=0, $suppress_value=0) {
        # Write a representation of the table to its file.  It calls several
        # functions furnished by sub-classes of this abstract base class to
        # handle non-normal ranges, to add stuff before the table, and at its
        # end.  If the table is to be written so that adjustments are
        # required, this does that conversion.


        # $use_adjustments ? output in adjusted format or not
        # $suppress_value Optional, if the value associated with
        # a range equals this one, don't write
        # the range

        my $addr = pack 'J', refaddr $self;
        my $write_as_invlist = $write_as_invlist{$addr};

        # Start with the header
        my @HEADER = $self->header;

        # Then the comments
        push @HEADER, "\n", main::simple_fold($comment{$addr}, '# '), "\n"
                                                        if $comment{$addr};

        # Things discovered processing the main body of the document may
        # affect what gets output before it, therefore pre_body() isn't called
        # until after all other processing of the table is done.

        # The main body looks like a 'here' document.  If there are comments,
        # get rid of them when processing it.
        my @OUT;
        if ($annotate || $output_range_counts) {
            # Use the line below in Perls that don't have /r
            #push @OUT, 'return join "\n",  map { s/\s*#.*//mg; $_ } split "\n", <<\'END\';' . "\n";
            push @OUT, "return <<'END' =~ s/\\s*#.*//mgr;\n";
        } else {
            push @OUT, "return <<'END';\n";
        }

        if ($range_list{$addr}->is_empty) {

            # This is a kludge for empty tables to silence a warning in
            # utf8.c, which can't really deal with empty tables, but it can
            # deal with a table that matches nothing, as the inverse of 'All'
            # does.
            push @OUT, "!Unicode::UCD::All\n";
        }
        elsif ($self->name eq 'N'

               # To save disk space and table cache space, avoid putting out
               # binary N tables, but instead create a file which just inverts
               # the Y table.  Since the file will still exist and occupy a
               # certain number of blocks, might as well output the whole
               # thing if it all will fit in one block.   The number of
               # ranges below is an approximate number for that.
               && ($self->property->type == $BINARY
                   || $self->property->type == $FORCED_BINARY)
               # && $self->property->tables == 2  Can't do this because the
               #        non-binary properties, like NFDQC aren't specifiable
               #        by the notation
               && $range_list{$addr}->ranges > 15
               && ! $annotate)  # Under --annotate, want to see everything
        {
            push @OUT, "!Unicode::UCD::" . $self->property->name . "\n";
        }
        else {
            my $range_size_1 = $range_size_1{$addr};

            # To make it more readable, use a minimum indentation
            my $comment_indent;

            # These are used only in $annotate option
            my $format;         # e.g. $HEX_ADJUST_FORMAT
            my $include_name;   # ? Include the character's name in the
                                # annotation?
            my $include_cp;     # ? Include its code point

            if (! $annotate) {
                $comment_indent = ($self->isa('Map_Table'))
                                  ? 24
                                  : ($write_as_invlist)
                                    ? 8
                                    : 16;
            }
            else {
                $format = $self->format;

                # The name of the character is output only for tables that
                # don't already include the name in the output.
                my $property = $self->property;
                $include_name =
                    !  ($property == $perl_charname
                        || $property == main::property_ref('Unicode_1_Name')
                        || $property == main::property_ref('Name')
                        || $property == main::property_ref('Name_Alias')
                       );

                # Don't include the code point in the annotation where all
                # lines are a single code point, so it can be easily found in
                # the first column
                $include_cp = ! $range_size_1;

                if (! $self->isa('Map_Table')) {
                    $comment_indent = ($write_as_invlist) ? 8 : 16;
                }
                else {
                    $comment_indent = 16;

                    # There are just a few short ranges in this table, so no
                    # need to include the code point in the annotation.
                    $include_cp = 0 if $format eq $DECOMP_STRING_FORMAT;

                    # We're trying to get this to look good, as the whole
                    # point is to make human-readable tables.  It is easier to
                    # read if almost all the annotation comments begin in the
                    # same column.  Map tables have varying width maps, so can
                    # create a jagged comment appearance.  This code does a
                    # preliminary pass through these tables looking for the
                    # maximum width map in each, and causing the comments to
                    # begin just to the right of that.  However, if the
                    # comments begin too far to the right of most lines, it's
                    # hard to line them up horizontally with their real data.
                    # Therefore we ignore the longest outliers
                    my $ignore_longest_X_percent = 2;  # Discard longest X%

                    # Each key in this hash is a width of at least one of the
                    # maps in the table.  Its value is how many lines have
                    # that width.
                    my %widths;

                    # We won't space things further left than one tab stop
                    # after the rest of the line; initializing it to that
                    # number saves some work.
                    my $max_map_width = 8;

                    # Fill in the %widths hash
                    my $total = 0;
                    for my $set ($range_list{$addr}->ranges) {
                        my $value = $set->value;

                        # These range types don't appear in the main table
                        next if $set->type == 0
                                && defined $suppress_value
                                && $value eq $suppress_value;
                        next if $set->type == $MULTI_CP
                                || $set->type == $NULL;

                        # Include 2 spaces before the beginning of the
                        # comment
                        my $this_width = length($value) + 2;

                        # Ranges of the remaining non-zero types usually
                        # occupy just one line (maybe occasionally two, but
                        # this doesn't have to be dead accurate).  This is
                        # because these ranges are like "unassigned code
                        # points"
                        my $count = ($set->type != 0)
                                    ? 1
                                    : $set->end - $set->start + 1;
                        $widths{$this_width} += $count;
                        $total += $count;
                        $max_map_width = $this_width
                                            if $max_map_width < $this_width;
                    }

                    # If the widest map gives us less than two tab stops
                    # worth, just take it as-is.
                    if ($max_map_width > 16) {

                        # Otherwise go through %widths until we have included
                        # the desired percentage of lines in the whole table.
                        my $running_total = 0;
                        foreach my $width (sort { $a <=> $b } keys %widths)
                        {
                            $running_total += $widths{$width};
                            use integer;
                            if ($running_total * 100 / $total
                                            >= 100 - $ignore_longest_X_percent)
                            {
                                $max_map_width = $width;
                                last;
                            }
                        }
                    }
                    $comment_indent += $max_map_width;
                }
            }

            # Values for previous time through the loop.  Initialize to
            # something that won't be adjacent to the first iteration;
            # only $previous_end matters for that.
            my $previous_start;
            my $previous_end = -2;
            my $previous_value;

            # Values for next time through the portion of the loop that splits
            # the range.  0 in $next_start means there is no remaining portion
            # to deal with.
            my $next_start = 0;
            my $next_end;
            my $next_value;
            my $offset = 0;
            my $invlist_count = 0;

            my $output_value_in_hex = $self->isa('Map_Table')
                                && ($self->format eq $HEX_ADJUST_FORMAT
                                    || $self->to_output_map == $EXTERNAL_MAP);
            # Use leading zeroes just for files whose format should not be
            # changed from what it has been.  Otherwise, they just take up
            # space and time to process.
            my $hex_format = ($self->isa('Map_Table')
                              && $self->to_output_map == $EXTERNAL_MAP)
                             ? "%04X"
                             : "%X";

            # The values for some of these tables are stored in mktables as
            # hex strings.  Normally, these are just output as strings without
            # change, but when we are doing adjustments, we have to operate on
            # these numerically, so we convert those to decimal to do that,
            # and back to hex for output
            my $convert_map_to_from_hex = 0;
            my $output_map_in_hex = 0;
            if ($self->isa('Map_Table')) {
                $convert_map_to_from_hex
                   = ($use_adjustments && $self->format eq $HEX_ADJUST_FORMAT)
                      || ($annotate && $self->format eq $HEX_FORMAT);
                $output_map_in_hex = $convert_map_to_from_hex
                                 || $self->format eq $HEX_FORMAT;
            }

            # To store any annotations about the characters.
            my @annotation;

            # Output each range as part of the here document.
            RANGE:
            for my $set ($range_list{$addr}->ranges) {
                if ($set->type != 0) {
                    $self->handle_special_range($set);
                    next RANGE;
                }
                my $start = $set->start;
                my $end   = $set->end;
                my $value  = $set->value;

                # Don't output ranges whose value is the one to suppress
                next RANGE if defined $suppress_value
                              && $value eq $suppress_value;

                $value = CORE::hex $value if $convert_map_to_from_hex;


                {   # This bare block encloses the scope where we may need to
                    # 'redo' to.  Consider a table that is to be written out
                    # using single item ranges.  This is given in the
                    # $range_size_1 boolean.  To accomplish this, we split the
                    # range each time through the loop into two portions, the
                    # first item, and the rest.  We handle that first item
                    # this time in the loop, and 'redo' to repeat the process
                    # for the rest of the range.
                    #
                    # We may also have to do it, with other special handling,
                    # if the table has adjustments.  Consider the table that
                    # contains the lowercasing maps.  mktables stores the
                    # ASCII range ones as 26 ranges:
                    #       ord('A') => ord('a'), .. ord('Z') => ord('z')
                    # For compactness, the table that gets written has this as
                    # just one range
                    #       ( ord('A') .. ord('Z') ) => ord('a')
                    # and the software that reads the tables is smart enough
                    # to "connect the dots".  This change is accomplished in
                    # this loop by looking to see if the current iteration
                    # fits the paradigm of the previous iteration, and if so,
                    # we merge them by replacing the final output item with
                    # the merged data.  Repeated 25 times, this gets A-Z.  But
                    # we also have to make sure we don't screw up cases where
                    # we have internally stored
                    #       ( 0x1C4 .. 0x1C6 ) => 0x1C5
                    # This single internal range has to be output as 3 ranges,
                    # which is done by splitting, like we do for $range_size_1
                    # tables.  (There are very few of such ranges that need to
                    # be split, so the gain of doing the combining of other
                    # ranges far outweighs the splitting of these.)  The
                    # values to use for the redo at the end of this block are
                    # set up just below in the scalars whose names begin with
                    # '$next_'.

                    if (($use_adjustments || $range_size_1) && $end != $start)
                    {
                        $next_start = $start + 1;
                        $next_end = $end;
                        $next_value = $value;
                        $end = $start;
                    }

                    if ($use_adjustments && ! $range_size_1) {

                        # If this range is adjacent to the previous one, and
                        # the values in each are integers that are also
                        # adjacent (differ by 1), then this range really
                        # extends the previous one that is already in element
                        # $OUT[-1].  So we pop that element, and pretend that
                        # the range starts with whatever it started with.
                        # $offset is incremented by 1 each time so that it
                        # gives the current offset from the first element in
                        # the accumulating range, and we keep in $value the
                        # value of that first element.
                        if ($start == $previous_end + 1
                            && $value =~ /^ -? \d+ $/xa
                            && $previous_value =~ /^ -? \d+ $/xa
                            && ($value == ($previous_value + ++$offset)))
                        {
                            pop @OUT;
                            $start = $previous_start;
                            $value = $previous_value;
                        }
                        else {
                            $offset = 0;
                            if (@annotation == 1) {
                                $OUT[-1] = merge_single_annotation_line(
                                    $OUT[-1], $annotation[0], $comment_indent);
                            }
                            else {
                                push @OUT, @annotation;
                            }
                        }
                        undef @annotation;

                        # Save the current values for the next time through
                        # the loop.
                        $previous_start = $start;
                        $previous_end = $end;
                        $previous_value = $value;
                    }

                    if ($write_as_invlist) {
                        if (   $previous_end > 0
                            && $output_range_counts{$addr})
                        {
                            my $complement_count = $start - $previous_end - 1;
                            if ($complement_count > 1) {
                                $OUT[-1] = merge_single_annotation_line(
                                    $OUT[-1],
                                       "#"
                                     . (" " x 17)
                                     . "["
                                     .  main::clarify_code_point_count(
                                                            $complement_count)
                                      . "] in complement\n",
                                    $comment_indent);
                            }
                        }

                        # Inversion list format has a single number per line,
                        # the starting code point of a range that matches the
                        # property
                        push @OUT, $start, "\n";
                        $invlist_count++;

                        # Add a comment with the size of the range, if
                        # requested.
                        if ($output_range_counts{$addr}) {
                            $OUT[-1] = merge_single_annotation_line(
                                    $OUT[-1],
                                    "# ["
                                      . main::clarify_code_point_count($end - $start + 1)
                                      . "]\n",
                                    $comment_indent);
                        }
                    }
                    elsif ($start != $end) { # If there is a range
                        if ($end == $MAX_WORKING_CODEPOINT) {
                            push @OUT, sprintf "$hex_format\t$hex_format",
                                                $start,
                                                $MAX_PLATFORM_CODEPOINT;
                        }
                        else {
                            push @OUT, sprintf "$hex_format\t$hex_format",
                                                $start,       $end;
                        }
                        if (length $value) {
                            if ($convert_map_to_from_hex) {
                                $OUT[-1] .= sprintf "\t$hex_format\n", $value;
                            }
                            else {
                                $OUT[-1] .= "\t$value\n";
                            }
                        }

                        # Add a comment with the size of the range, if
                        # requested.
                        if ($output_range_counts{$addr}) {
                            $OUT[-1] = merge_single_annotation_line(
                                    $OUT[-1],
                                    "# ["
                                      . main::clarify_code_point_count($end - $start + 1)
                                      . "]\n",
                                    $comment_indent);
                        }
                    }
                    else { # Here to output a single code point per line.

                        # Use any passed in subroutine to output.
                        if (ref $range_size_1 eq 'CODE') {
                            for my $i ($start .. $end) {
                                push @OUT, &{$range_size_1}($i, $value);
                            }
                        }
                        else {

                            # Here, caller is ok with default output.
                            for (my $i = $start; $i <= $end; $i++) {
                                if ($convert_map_to_from_hex) {
                                    push @OUT,
                                        sprintf "$hex_format\t\t$hex_format\n",
                                                 $i,            $value;
                                }
                                else {
                                    push @OUT, sprintf $hex_format, $i;
                                    $OUT[-1] .= "\t\t$value" if $value ne "";
                                    $OUT[-1] .= "\n";
                                }
                            }
                        }
                    }

                    if ($annotate) {
                        for (my $i = $start; $i <= $end; $i++) {
                            my $annotation = "";

                            # Get character information if don't have it already
                            main::populate_char_info($i)
                                                     if ! defined $viacode[$i];
                            my $type = $annotate_char_type[$i];

                            # Figure out if should output the next code points
                            # as part of a range or not.  If this is not in an
                            # annotation range, then won't output as a range,
                            # so returns $i.  Otherwise use the end of the
                            # annotation range, but no further than the
                            # maximum possible end point of the loop.
                            my $range_end =
                                        $range_size_1
                                        ? $start
                                        : main::min(
                                          $annotate_ranges->value_of($i) || $i,
                                          $end);

                            # Use a range if it is a range, and either is one
                            # of the special annotation ranges, or the range
                            # is at most 3 long.  This last case causes the
                            # algorithmically named code points to be output
                            # individually in spans of at most 3, as they are
                            # the ones whose $type is > 0.
                            if ($range_end != $i
                                && ( $type < 0 || $range_end - $i > 2))
                            {
                                # Here is to output a range.  We don't allow a
                                # caller-specified output format--just use the
                                # standard one.
                                my $range_name = $viacode[$i];

                                # For the code points which end in their hex
                                # value, we eliminate that from the output
                                # annotation, and capitalize only the first
                                # letter of each word.
                                if ($type == $CP_IN_NAME) {
                                    my $hex = sprintf $hex_format, $i;
                                    $range_name =~ s/-$hex$//;
                                    my @words = split " ", $range_name;
                                    for my $word (@words) {
                                        $word =
                                          ucfirst(lc($word)) if $word ne 'CJK';
                                    }
                                    $range_name = join " ", @words;
                                }
                                elsif ($type == $HANGUL_SYLLABLE) {
                                    $range_name = "Hangul Syllable";
                                }

                                # If the annotation would just repeat what's
                                # already being output as the range, skip it.
                                # (When an inversion list is being written, it
                                # isn't a repeat, as that always is in
                                # decimal)
                                if (   $write_as_invlist
                                    || $i != $start
                                    || $range_end < $end)
                                {
                                    if ($range_end < $MAX_WORKING_CODEPOINT)
                                    {
                                        $annotation = sprintf "%04X..%04X",
                                                              $i,   $range_end;
                                    }
                                    else {
                                        $annotation = sprintf "%04X..INFINITY",
                                                               $i;
                                    }
                                }
                                else { # Indent if not displaying code points
                                    $annotation = " " x 4;
                                }

                                if ($range_name) {
                                    $annotation .= " $age[$i]" if $age[$i];
                                    $annotation .= " $range_name";
                                }

                                # Include the number of code points in the
                                # range
                                my $count =
                                    main::clarify_code_point_count($range_end - $i + 1);
                                $annotation .= " [$count]\n";

                                # Skip to the end of the range
                                $i = $range_end;
                            }
                            else { # Not in a range.
                                my $comment = "";

                                # When outputting the names of each character,
                                # use the character itself if printable
                                $comment .= "'" . main::display_chr($i) . "' "
                                                            if $printable[$i];

                                my $output_value = $value;

                                # Determine the annotation
                                if ($format eq $DECOMP_STRING_FORMAT) {

                                    # This is very specialized, with the type
                                    # of decomposition beginning the line
                                    # enclosed in <...>, and the code points
                                    # that the code point decomposes to
                                    # separated by blanks.  Create two
                                    # strings, one of the printable
                                    # characters, and one of their official
                                    # names.
                                    (my $map = $output_value)
                                                    =~ s/ \ * < .*? > \ +//x;
                                    my $tostr = "";
                                    my $to_name = "";
                                    my $to_chr = "";
                                    foreach my $to (split " ", $map) {
                                        $to = CORE::hex $to;
                                        $to_name .= " + " if $to_name;
                                        $to_chr .= main::display_chr($to);
                                        main::populate_char_info($to)
                                                    if ! defined $viacode[$to];
                                        $to_name .=  $viacode[$to];
                                    }

                                    $comment .=
                                    "=> '$to_chr'; $viacode[$i] => $to_name";
                                }
                                else {
                                    $output_value += $i - $start
                                                   if $use_adjustments
                                                      # Don't try to adjust a
                                                      # non-integer
                                                   && $output_value !~ /[-\D]/;

                                    if ($output_map_in_hex) {
                                        main::populate_char_info($output_value)
                                          if ! defined $viacode[$output_value];
                                        $comment .= " => '"
                                        . main::display_chr($output_value)
                                        . "'; " if $printable[$output_value];
                                    }
                                    if ($include_name && $viacode[$i]) {
                                        $comment .= " " if $comment;
                                        $comment .= $viacode[$i];
                                    }
                                    if ($output_map_in_hex) {
                                        $comment .=
                                                " => $viacode[$output_value]"
                                                    if $viacode[$output_value];
                                        $output_value = sprintf($hex_format,
                                                                $output_value);
                                    }
                                }

                                if ($include_cp) {
                                    $annotation = sprintf "%04X %s", $i, $age[$i];
                                    if ($use_adjustments) {
                                        $annotation .= " => $output_value";
                                    }
                                }

                                if ($comment ne "") {
                                    $annotation .= " " if $annotation ne "";
                                    $annotation .= $comment;
                                }
                                $annotation .= "\n" if $annotation ne "";
                            }

                            if ($annotation ne "") {
                                push @annotation, (" " x $comment_indent)
                                                  .  "# $annotation";
                            }
                        }

                        # If not adjusting, we don't have to go through the
                        # loop again to know that the annotation comes next
                        # in the output.
                        if (! $use_adjustments) {
                            if (@annotation == 1) {
                                $OUT[-1] = merge_single_annotation_line(
                                    $OUT[-1], $annotation[0], $comment_indent);
                            }
                            else {
                                push @OUT, map { Text::Tabs::unexpand $_ }
                                               @annotation;
                            }
                            undef @annotation;
                        }
                    }

                    # Add the beginning of the range that doesn't match the
                    # property, except if the just added match range extends
                    # to infinity.  We do this after any annotations for the
                    # match range.
                    if ($write_as_invlist && $end < $MAX_WORKING_CODEPOINT) {
                        push @OUT, $end + 1, "\n";
                        $invlist_count++;
                    }

                    # If we split the range, set up so the next time through
                    # we get the remainder, and redo.
                    if ($next_start) {
                        $start = $next_start;
                        $end = $next_end;
                        $value = $next_value;
                        $next_start = 0;
                        redo;
                    }
                } # End of redo block
            } # End of loop through all the table's ranges

            push @OUT, @annotation; # Add orphaned annotation, if any

            splice @OUT, 1, 0, "V$invlist_count\n" if $invlist_count;
        }

        # Add anything that goes after the main body, but within the here
        # document,
        my $append_to_body = $self->append_to_body;
        push @OUT, $append_to_body if $append_to_body;

        # And finish the here document.
        push @OUT, "END\n";

        # Done with the main portion of the body.  Can now figure out what
        # should appear before it in the file.
        my $pre_body = $self->pre_body;
        push @HEADER, $pre_body, "\n" if $pre_body;

        # All these files should have a .pl suffix added to them.
        my @file_with_pl = @{$file_path{$addr}};
        $file_with_pl[-1] .= '.pl';

        main::write(\@file_with_pl,
                    $annotate,      # utf8 iff annotating
                    \@HEADER,
                    \@OUT);
        return;
    }

    sub set_status($self, $status, $info) {    # Set the table's status
        # status The status enum value
        # info Any message associated with it.
        my $addr = pack 'J', refaddr $self;

        $status{$addr} = $status;
        $status_info{$addr} = $info;
        return;
    }

    sub set_fate($self, $fate, $reason=undef) {  # Set the fate of a table
        my $addr = pack 'J', refaddr $self;

        return if $fate{$addr} == $fate;    # If no-op

        # Can only change the ordinary fate, except if going to $MAP_PROXIED
        return if $fate{$addr} != $ORDINARY && $fate != $MAP_PROXIED;

        $fate{$addr} = $fate;

        # Don't document anything to do with a non-normal fated table
        if ($fate != $ORDINARY) {
            my $put_in_pod = ($fate == $MAP_PROXIED) ? 1 : 0;
            foreach my $alias ($self->aliases) {
                $alias->set_ucd($put_in_pod);

                # MAP_PROXIED doesn't affect the match tables
                next if $fate == $MAP_PROXIED;
                $alias->set_make_re_pod_entry($put_in_pod);
            }
        }

        # Save the reason for suppression for output
        if ($fate >= $SUPPRESSED) {
            $reason = "" unless defined $reason;
            $why_suppressed{$complete_name{$addr}} = $reason;
        }

        return;
    }

    sub lock($self) {
        # Don't allow changes to the table from now on.  This stores a stack
        # trace of where it was called, so that later attempts to modify it
        # can immediately show where it got locked.
        my $addr = pack 'J', refaddr $self;

        $locked{$addr} = "";

        my $line = (caller(0))[2];
        my $i = 1;

        # Accumulate the stack trace
        while (1) {
            my ($pkg, $file, $caller_line, $caller) = caller $i++;

            last unless defined $caller;

            $locked{$addr} .= "    called from $caller() at line $line\n";
            $line = $caller_line;
        }
        $locked{$addr} .= "    called from main at line $line\n";

        return;
    }

    sub carp_if_locked($self) {
        # Return whether a table is locked or not, and, by the way, complain
        # if is locked
        my $addr = pack 'J', refaddr $self;

        return 0 if ! $locked{$addr};
        Carp::my_carp_bug("Can't modify a locked table. Stack trace of locking:\n$locked{$addr}\n\n");
        return 1;
    }

    sub set_file_path($self, @path) { # Set the final directory path for this table
        @{$file_path{pack 'J', refaddr $self}} = @path;
        return
    }

    # Accessors for the range list stored in this table.  First for
    # unconditional
    for my $sub (qw(
                    containing_range
                    contains
                    count
                    each_range
                    hash
                    is_empty
                    matches_identically_to
                    max
                    min
                    range_count
                    reset_each_range
                    type_of
                    value_of
                ))
    {
        no strict "refs";
        *$sub = sub {
            use strict "refs";
            my $self = shift;
            return $self->_range_list->$sub(@_);
        }
    }

    # Then for ones that should fail if locked
    for my $sub (qw(
                    delete_range
                ))
    {
        no strict "refs";
        *$sub = sub {
            use strict "refs";
            my $self = shift;

            return if $self->carp_if_locked;
            no overloading;
            return $self->_range_list->$sub(@_);
        }
    }

} # End closure

package Map_Table;
use parent '-norequire', '_Base_Table';

# A Map Table is a table that contains the mappings from code points to
# values.  There are two weird cases:
# 1) Anomalous entries are ones that aren't maps of ranges of code points, but
#    are written in the table's file at the end of the table nonetheless.  It
#    requires specially constructed code to handle these; utf8.c can not read
#    these in, so they should not go in $map_directory.  As of this writing,
#    the only case that these happen is for named sequences used in
#    charnames.pm.   But this code doesn't enforce any syntax on these, so
#    something else could come along that uses it.
# 2) Specials are anything that doesn't fit syntactically into the body of the
#    table.  The ranges for these have a map type of non-zero.  The code below
#    knows about and handles each possible type.   In most cases, these are
#    written as part of the header.
#
# A map table deliberately can't be manipulated at will unlike match tables.
# This is because of the ambiguities having to do with what to do with
# overlapping code points.  And there just isn't a need for those things;
# what one wants to do is just query, add, replace, or delete mappings, plus
# write the final result.
# However, there is a method to get the list of possible ranges that aren't in
# this table to use for defaulting missing code point mappings.  And,
# map_add_or_replace_non_nulls() does allow one to add another table to this
# one, but it is clearly very specialized, and defined that the other's
# non-null values replace this one's if there is any overlap.

sub trace { return main::trace(@_); }

{ # Closure

    main::setup_package();

    my %default_map;
    # Many input files omit some entries; this gives what the mapping for the
    # missing entries should be
    main::set_access('default_map', \%default_map, 'r');

    my %anomalous_entries;
    # Things that go in the body of the table which don't fit the normal
    # scheme of things, like having a range.  Not much can be done with these
    # once there except to output them.  This was created to handle named
    # sequences.
    main::set_access('anomalous_entry', \%anomalous_entries, 'a');
    main::set_access('anomalous_entries',       # Append singular, read plural
                    \%anomalous_entries,
                    'readable_array');
    my %to_output_map;
    # Enum as to whether or not to write out this map table, and how:
    #   0               don't output
    #   $EXTERNAL_MAP   means its existence is noted in the documentation, and
    #                   it should not be removed nor its format changed.  This
    #                   is done for those files that have traditionally been
    #                   output.
    #   $INTERNAL_MAP   means Perl reserves the right to do anything it wants
    #                   with this file
    #   $OUTPUT_ADJUSTED means that it is an $INTERNAL_MAP, and instead of
    #                   outputting the actual mappings as-is, we adjust things
    #                   to create a much more compact table. Only those few
    #                   tables where the mapping is convertible at least to an
    #                   integer and compacting makes a big difference should
    #                   have this.  Hence, the default is to not do this
    #                   unless the table's default mapping is to $CODE_POINT,
    #                   and the range size is not 1.
    main::set_access('to_output_map', \%to_output_map, 's');

    sub new {
        my $class = shift;
        my $name = shift;

        my %args = @_;

        # Optional initialization data for the table.
        my $initialize = delete $args{'Initialize'};

        my $default_map = delete $args{'Default_Map'};
        my $property = delete $args{'_Property'};
        my $full_name = delete $args{'Full_Name'};
        my $to_output_map = delete $args{'To_Output_Map'};

        # Rest of parameters passed on

        my $range_list = Range_Map->new(Owner => $property);

        my $self = $class->SUPER::new(
                                    Name => $name,
                                    Complete_Name =>  $full_name,
                                    Full_Name => $full_name,
                                    _Property => $property,
                                    _Range_List => $range_list,
                                    Write_As_Invlist => 0,
                                    %args);

        my $addr = pack 'J', refaddr $self;

        $anomalous_entries{$addr} = [];
        $default_map{$addr} = $default_map;
        $to_output_map{$addr} = $to_output_map;

        $self->initialize($initialize) if defined $initialize;

        return $self;
    }

    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {

        my $name = $self->property->full_name;
        $name = '""' if $name eq "";
        return "Map table for Property '$name'";
    }

    sub add_alias {
        # Add a synonym for this table (which means the property itself)
        my $self = shift;
        my $name = shift;
        # Rest of parameters passed on.

        $self->SUPER::add_alias($name, $self->property, @_);
        return;
    }

    sub add_map {
        # Add a range of code points to the list of specially-handled code
        # points.  0 is assumed if the type of special is not passed
        # in.

        my $self = shift;
        my $lower = shift;
        my $upper = shift;
        my $string = shift;
        my %args = @_;

        my $type = delete $args{'Type'} || 0;
        # Rest of parameters passed on

        # Can't change the table if locked.
        return if $self->carp_if_locked;

        $self->_range_list->add_map($lower, $upper,
                                    $string,
                                    @_,
                                    Type => $type);
        return;
    }

    sub append_to_body($self) {
        # Adds to the written HERE document of the table's body any anomalous
        # entries in the table..
        my $addr = pack 'J', refaddr $self;

        return "" unless @{$anomalous_entries{$addr}};
        return join("\n", @{$anomalous_entries{$addr}}) . "\n";
    }

    sub map_add_or_replace_non_nulls($self, $other) {
        # This adds the mappings in the table $other to $self.  Non-null
        # mappings from $other override those in $self.  It essentially merges
        # the two tables, with the second having priority except for null
        # mappings.
        return if $self->carp_if_locked;

        if (! $other->isa(__PACKAGE__)) {
            Carp::my_carp_bug("$other should be a "
                        . __PACKAGE__
                        . ".  Not a '"
                        . ref($other)
                        . "'.  Not added;");
            return;
        }

        local $to_trace = 0 if main::DEBUG;

        my $self_range_list = $self->_range_list;
        my $other_range_list = $other->_range_list;
        foreach my $range ($other_range_list->ranges) {
            my $value = $range->value;
            next if $value eq "";
            $self_range_list->_add_delete('+',
                                          $range->start,
                                          $range->end,
                                          $value,
                                          Type => $range->type,
                                          Replace => $UNCONDITIONALLY);
        }

        return;
    }

    sub set_default_map($self, $map, $use_full_name=0) {
        # Define what code points that are missing from the input files should
        # map to.  The optional second parameter 'full_name' indicates to
        # force using the full name of the map instead of its standard name.
        if ($use_full_name && $use_full_name ne 'full_name') {
            Carp::my_carp_bug("Second parameter to set_default_map() if"
                            . " present, must be 'full_name'");
        }

        my $addr = pack 'J', refaddr $self;

        # Convert the input to the standard equivalent, if any (won't have any
        # for $STRING properties)
        my $standard = $self->property->table($map);
        if (defined $standard) {
            $map = ($use_full_name)
                   ? $standard->full_name
                   : $standard->name;
        }

        # Warn if there already is a non-equivalent default map for this
        # property.  Note that a default map can be a ref, which means that
        # what it actually means is delayed until later in the program, and it
        # IS permissible to override it here without a message.
        my $default_map = $default_map{$addr};
        if (defined $default_map
            && ! ref($default_map)
            && $default_map ne $map
            && main::Standardize($map) ne $default_map)
        {
            my $property = $self->property;
            my $map_table = $property->table($map);
            my $default_table = $property->table($default_map);
            if (defined $map_table
                && defined $default_table
                && $map_table != $default_table)
            {
                Carp::my_carp("Changing the default mapping for "
                            . $property
                            . " from $default_map to $map'");
            }
        }

        $default_map{$addr} = $map;

        # Don't also create any missing table for this map at this point,
        # because if we did, it could get done before the main table add is
        # done for PropValueAliases.txt; instead the caller will have to make
        # sure it exists, if desired.
        return;
    }

    sub to_output_map($self) {
        # Returns boolean: should we write this map table?
        my $addr = pack 'J', refaddr $self;

        # If overridden, use that
        return $to_output_map{$addr} if defined $to_output_map{$addr};

        my $full_name = $self->full_name;
        return $global_to_output_map{$full_name}
                                if defined $global_to_output_map{$full_name};

        # If table says to output, do so; if says to suppress it, do so.
        my $fate = $self->fate;
        return $INTERNAL_MAP if $fate == $INTERNAL_ONLY;
        return $EXTERNAL_MAP if grep { $_ eq $full_name } @output_mapped_properties;
        return 0 if $fate == $SUPPRESSED || $fate == $MAP_PROXIED;

        my $type = $self->property->type;

        # Don't want to output binary map tables even for debugging.
        return 0 if $type == $BINARY;

        # But do want to output string ones.  All the ones that remain to
        # be dealt with (i.e. which haven't explicitly been set to external)
        # are for internal Perl use only.  The default for those that map to
        # $CODE_POINT and haven't been restricted to a single element range
        # is to use the adjusted form.
        if ($type == $STRING) {
            return $INTERNAL_MAP if $self->range_size_1
                                    || $default_map{$addr} ne $CODE_POINT;
            return $OUTPUT_ADJUSTED;
        }

        # Otherwise is an $ENUM, do output it, for Perl's purposes
        return $INTERNAL_MAP;
    }

    sub inverse_list($self) {
        # Returns a Range_List that is gaps of the current table.  That is,
        # the inversion
        my $current = Range_List->new(Initialize => $self->_range_list,
                                Owner => $self->property);
        return ~ $current;
    }

    sub header($self) {
        my $return = $self->SUPER::header();

        if ($self->to_output_map >= $INTERNAL_MAP) {
            $return .= $INTERNAL_ONLY_HEADER;
        }
        else {
            # Other properties have fixed formats.
            my $property_name = $self->property->full_name;

            $return .= <<END;

# !!!!!!!   IT IS DEPRECATED TO USE THIS FILE   !!!!!!!

# This file is for internal use by core Perl only.  It is retained for
# backwards compatibility with applications that may have come to rely on it,
# but its format and even its name or existence are subject to change without
# notice in a future Perl version.  Don't use it directly.  Instead, its
# contents are now retrievable through a stable API in the Unicode::UCD
# module: Unicode::UCD::prop_invmap('$property_name') (Values for individual
# code points can be retrieved via Unicode::UCD::charprop());
END
        }
        return $return;
    }

    sub set_final_comment($self) {
        # Just before output, create the comment that heads the file
        # containing this table.

        return unless $debugging_build;

        # No sense generating a comment if aren't going to write it out.
        return if ! $self->to_output_map;

        my $addr = pack 'J', refaddr $self;

        my $property = $self->property;

        # Get all the possible names for this property.  Don't use any that
        # aren't ok for use in a file name, etc.  This is perhaps causing that
        # flag to do double duty, and may have to be changed in the future to
        # have our own flag for just this purpose; but it works now to exclude
        # Perl generated synonyms from the lists for properties, where the
        # name is always the proper Unicode one.
        my @property_aliases = grep { $_->ok_as_filename } $self->aliases;

        my $count = $self->count;
        my $default_map = $default_map{$addr};

        # The ranges that map to the default aren't output, so subtract that
        # to get those actually output.  A property with matching tables
        # already has the information calculated.
        if ($property->type != $STRING && $property->type != $FORCED_BINARY) {
            $count -= $property->table($default_map)->count;
        }
        elsif (defined $default_map) {

            # But for $STRING properties, must calculate now.  Subtract the
            # count from each range that maps to the default.
            foreach my $range ($self->_range_list->ranges) {
                if ($range->value eq $default_map) {
                    $count -= $range->end +1 - $range->start;
                }
            }

        }

        # Get a  string version of $count with underscores in large numbers,
        # for clarity.
        my $string_count = main::clarify_code_point_count($count);

        my $code_points = ($count == 1)
                        ? 'single code point'
                        : "$string_count code points";

        my $mapping;
        my $these_mappings;
        my $are;
        if (@property_aliases <= 1) {
            $mapping = 'mapping';
            $these_mappings = 'this mapping';
            $are = 'is'
        }
        else {
            $mapping = 'synonymous mappings';
            $these_mappings = 'these mappings';
            $are = 'are'
        }
        my $cp;
        if ($count >= $MAX_UNICODE_CODEPOINTS) {
            $cp = "any code point in Unicode Version $string_version";
        }
        else {
            my $map_to;
            if ($default_map eq "") {
                $map_to = 'the empty string';
            }
            elsif ($default_map eq $CODE_POINT) {
                $map_to = "itself";
            }
            else {
                $map_to = "'$default_map'";
            }
            if ($count == 1) {
                $cp = "the single code point";
            }
            else {
                $cp = "one of the $code_points";
            }
            $cp .= " in Unicode Version $unicode_version for which the mapping is not to $map_to";
        }

        my $comment = "";

        my $status = $self->status;
        if ($status ne $NORMAL) {
            my $warn = uc $status_past_participles{$status};
            $comment .= <<END;

!!!!!!!   $warn !!!!!!!!!!!!!!!!!!!
 All property or property=value combinations contained in this file are $warn.
 See $unicode_reference_url for what this means.

END
        }
        $comment .= "This file returns the $mapping:\n";

        my $ucd_accessible_name = "";
        my $has_underscore_name = 0;
        my $full_name = $self->property->full_name;
        for my $i (0 .. @property_aliases - 1) {
            my $name = $property_aliases[$i]->name;
            $has_underscore_name = 1 if $name =~ /^_/;
            $comment .= sprintf("%-8s%s\n", " ", $name . '(cp)');
            if ($property_aliases[$i]->ucd) {
                if ($name eq $full_name) {
                    $ucd_accessible_name = $full_name;
                }
                elsif (! $ucd_accessible_name) {
                    $ucd_accessible_name = $name;
                }
            }
        }
        $comment .= "\nwhere 'cp' is $cp.";
        if ($ucd_accessible_name) {
            $comment .= "  Note that $these_mappings";
            if ($has_underscore_name) {
                $comment .= " (except for the one(s) that begin with an underscore)";
            }
            $comment .= " $are accessible via the functions prop_invmap('$full_name') or charprop() in Unicode::UCD";

        }

        # And append any commentary already set from the actual property.
        $comment .= "\n\n" . $self->comment if $self->comment;
        if ($self->description) {
            $comment .= "\n\n" . join " ", $self->description;
        }
        if ($self->note) {
            $comment .= "\n\n" . join " ", $self->note;
        }
        $comment .= "\n";

        if (! $self->perl_extension) {
            $comment .= <<END;

For information about what this property really means, see:
$unicode_reference_url
END
        }

        if ($count) {        # Format differs for empty table
                $comment.= "\nThe format of the ";
            if ($self->range_size_1) {
                $comment.= <<END;
main body of lines of this file is: CODE_POINT\\t\\tMAPPING where CODE_POINT
is in hex; MAPPING is what CODE_POINT maps to.
END
            }
            else {

                # There are tables which end up only having one element per
                # range, but it is not worth keeping track of for making just
                # this comment a little better.
                $comment .= <<END;
non-comment portions of the main body of lines of this file is:
START\\tSTOP\\tMAPPING where START is the starting code point of the
range, in hex; STOP is the ending point, or if omitted, the range has just one
code point; MAPPING is what each code point between START and STOP maps to.
END
                if ($self->output_range_counts) {
                    $comment .= <<END;
Numbers in comments in [brackets] indicate how many code points are in the
range (omitted when the range is a single code point or if the mapping is to
the null string).
END
                }
            }
        }
        $self->set_comment(main::join_lines($comment));
        return;
    }

    my %swash_keys; # Makes sure don't duplicate swash names.

    # The remaining variables are temporaries used while writing each table,
    # to output special ranges.
    my @multi_code_point_maps;  # Map is to more than one code point.

    sub handle_special_range($self, $range) {
        # Called in the middle of write when it finds a range it doesn't know
        # how to handle.

        my $addr = pack 'J', refaddr $self;

        my $type = $range->type;

        my $low = $range->start;
        my $high = $range->end;
        my $map = $range->value;

        # No need to output the range if it maps to the default.
        return if $map eq $default_map{$addr};

        my $property = $self->property;

        # Switch based on the map type...
        if ($type == $HANGUL_SYLLABLE) {

            # These are entirely algorithmically determinable based on
            # some constants furnished by Unicode; for now, just set a
            # flag to indicate that have them.  After everything is figured
            # out, we will output the code that does the algorithm.  (Don't
            # output them if not needed because we are suppressing this
            # property.)
            $has_hangul_syllables = 1 if $property->to_output_map;
        }
        elsif ($type == $CP_IN_NAME) {

            # Code points whose name ends in their code point are also
            # algorithmically determinable, but need information about the map
            # to do so.  Both the map and its inverse are stored in data
            # structures output in the file.  They are stored in the mean time
            # in global lists The lists will be written out later into Name.pm,
            # which is created only if needed.  In order to prevent duplicates
            # in the list, only add to them for one property, should multiple
            # ones need them.
            if ($needing_code_points_ending_in_code_point == 0) {
                $needing_code_points_ending_in_code_point = $property;
            }
            if ($property == $needing_code_points_ending_in_code_point) {
                push @{$names_ending_in_code_point{$map}->{'low'}}, $low;
                push @{$names_ending_in_code_point{$map}->{'high'}}, $high;

                my $squeezed = $map =~ s/[-\s]+//gr;
                push @{$loose_names_ending_in_code_point{$squeezed}->{'low'}},
                                                                          $low;
                push @{$loose_names_ending_in_code_point{$squeezed}->{'high'}},
                                                                         $high;

                # Calculate the set of legal characters in names of this
                # series.  It includes every character in the name prefix.
                my %legal;
                $legal{$_} = 1 for split //, $map;

                # Plus the hex code point chars, blank, and minus.  Also \n
                # can show up as being required due to anchoring
                for my $i ('0' .. '9', 'A' .. 'F', '-', ' ', "\n") {
                    $legal{$i} = 1;
                }
                my $legal = join "", sort { $a cmp $b } keys %legal;

                # The legal chars can be used in match optimizations
                push @code_points_ending_in_code_point, { low => $low,
                                                        high => $high,
                                                        name => $map,
                                                        legal => $legal,
                                                        };
            }
        }
        elsif ($range->type == $MULTI_CP || $range->type == $NULL) {

            # Multi-code point maps and null string maps have an entry
            # for each code point in the range.  They use the same
            # output format.
            for my $code_point ($low .. $high) {

                # The pack() below can't cope with surrogates.  XXX This may
                # no longer be true
                if ($code_point >= 0xD800 && $code_point <= 0xDFFF) {
                    Carp::my_carp("Surrogate code point '$code_point' in mapping to '$map' in $self.  No map created");
                    next;
                }

                # Generate the hash entries for these in the form that
                # utf8.c understands.
                my $tostr = "";
                my $to_name = "";
                my $to_chr = "";
                foreach my $to (split " ", $map) {
                    if ($to !~ /^$code_point_re$/) {
                        Carp::my_carp("Illegal code point '$to' in mapping '$map' from $code_point in $self.  No map created");
                        next;
                    }
                    $tostr .= sprintf "\\x{%s}", $to;
                    $to = CORE::hex $to;
                    if ($annotate) {
                        $to_name .= " + " if $to_name;
                        $to_chr .= main::display_chr($to);
                        main::populate_char_info($to)
                                            if ! defined $viacode[$to];
                        $to_name .=  $viacode[$to];
                    }
                }

                # The unpack yields a list of the bytes that comprise the
                # UTF-8 of $code_point, which are each placed in \xZZ format
                # and output in the %s to map to $tostr, so the result looks
                # like:
                # "\xC4\xB0" => "\x{0069}\x{0307}",
                my $utf8 = sprintf(qq["%s" => "$tostr",],
                        join("", map { sprintf "\\x%02X", $_ }
                            unpack("U0C*", chr $code_point)));

                # Add a comment so that a human reader can more easily
                # see what's going on.
                push @multi_code_point_maps,
                        sprintf("%-45s # U+%04X", $utf8, $code_point);
                if (! $annotate) {
                    $multi_code_point_maps[-1] .= " => $map";
                }
                else {
                    main::populate_char_info($code_point)
                                    if ! defined $viacode[$code_point];
                    $multi_code_point_maps[-1] .= " '"
                        . main::display_chr($code_point)
                        . "' => '$to_chr'; $viacode[$code_point] => $to_name";
                }
            }
        }
        else {
            Carp::my_carp("Unrecognized map type '$range->type' in '$range' in $self.  Not written");
        }

        return;
    }

    sub pre_body($self) {
        # Returns the string that should be output in the file before the main
        # body of this table.  It isn't called until the main body is
        # calculated, saving a pass.  The string includes some hash entries
        # identifying the format of the body, and what the single value should
        # be for all ranges missing from it.  It also includes any code points
        # which have map_types that don't go in the main table.

        my $addr = pack 'J', refaddr $self;

        my $name = $self->property->swash_name;

        # Currently there is nothing in the pre_body unless a swash is being
        # generated.
        return unless defined $name;

        if (defined $swash_keys{$name}) {
            Carp::my_carp(main::join_lines(<<END
Already created a swash name '$name' for $swash_keys{$name}.  This means that
the same name desired for $self shouldn't be used.  Bad News.  This must be
fixed before production use, but proceeding anyway
END
            ));
        }
        $swash_keys{$name} = "$self";

        my $pre_body = "";

        # Here we assume we were called after have gone through the whole
        # file.  If we actually generated anything for each map type, add its
        # respective header and trailer
        my $specials_name = "";
        if (@multi_code_point_maps) {
            $specials_name = "Unicode::UCD::ToSpec$name";
            $pre_body .= <<END;

# Some code points require special handling because their mappings are each to
# multiple code points.  These do not appear in the main body, but are defined
# in the hash below.

# Each key is the string of N bytes that together make up the UTF-8 encoding
# for the code point.  (i.e. the same as looking at the code point's UTF-8
# under "use bytes").  Each value is the UTF-8 of the translation, for speed.
\%$specials_name = (
END
            $pre_body .= join("\n", @multi_code_point_maps) . "\n);\n";
        }

        my $format = $self->format;

        my $return = "";

        my $output_adjusted = ($self->to_output_map == $OUTPUT_ADJUSTED);
        if ($output_adjusted) {
            if ($specials_name) {
                $return .= <<END;
# The mappings in the non-hash portion of this file must be modified to get the
# correct values by adding the code point ordinal number to each one that is
# numeric.
END
            }
            else {
                $return .= <<END;
# The mappings must be modified to get the correct values by adding the code
# point ordinal number to each one that is numeric.
END
            }
        }

        $return .= <<END;

# The name this table is to be known by, with the format of the mappings in
# the main body of the table, and what all code points missing from this file
# map to.
\$Unicode::UCD::SwashInfo{'To$name'}{'format'} = '$format'; # $map_table_formats{$format}
END
        if ($specials_name) {
            $return .= <<END;
\$Unicode::UCD::SwashInfo{'To$name'}{'specials_name'} = '$specials_name'; # Name of hash of special mappings
END
        }
        my $default_map = $default_map{$addr};

        # For $CODE_POINT default maps and using adjustments, instead the default
        # becomes zero.
        $return .= "\$Unicode::UCD::SwashInfo{'To$name'}{'missing'} = '"
                .  (($output_adjusted && $default_map eq $CODE_POINT)
                   ? "0"
                   : $default_map)
                . "';";

        if ($default_map eq $CODE_POINT) {
            $return .= ' # code point maps to itself';
        }
        elsif ($default_map eq "") {
            $return .= ' # code point maps to the empty string';
        }
        $return .= "\n";

        $return .= $pre_body;

        return $return;
    }

    sub write($self) {
        # Write the table to the file.

        my $addr = pack 'J', refaddr $self;

        # Clear the temporaries
        undef @multi_code_point_maps;

        # Calculate the format of the table if not already done.
        my $format = $self->format;
        my $type = $self->property->type;
        my $default_map = $self->default_map;
        if (! defined $format) {
            if ($type == $BINARY) {

                # Don't bother checking the values, because we elsewhere
                # verify that a binary table has only 2 values.
                $format = $BINARY_FORMAT;
            }
            else {
                my @ranges = $self->_range_list->ranges;

                # default an empty table based on its type and default map
                if (! @ranges) {

                    # But it turns out that the only one we can say is a
                    # non-string (besides binary, handled above) is when the
                    # table is a string and the default map is to a code point
                    if ($type == $STRING && $default_map eq $CODE_POINT) {
                        $format = $HEX_FORMAT;
                    }
                    else {
                        $format = $STRING_FORMAT;
                    }
                }
                else {

                    # Start with the most restrictive format, and as we find
                    # something that doesn't fit with that, change to the next
                    # most restrictive, and so on.
                    $format = $DECIMAL_FORMAT;
                    foreach my $range (@ranges) {
                        next if $range->type != 0;  # Non-normal ranges don't
                                                    # affect the main body
                        my $map = $range->value;
                        if ($map ne $default_map) {
                            last if $format eq $STRING_FORMAT;  # already at
                                                                # least
                                                                # restrictive
                            $format = $INTEGER_FORMAT
                                                if $format eq $DECIMAL_FORMAT
                                                    && $map !~ / ^ [0-9] $ /x;
                            $format = $FLOAT_FORMAT
                                            if $format eq $INTEGER_FORMAT
                                                && $map !~ / ^ -? [0-9]+ $ /x;
                            $format = $RATIONAL_FORMAT
                                if $format eq $FLOAT_FORMAT
                                    && $map !~ / ^ -? [0-9]+ \. [0-9]* $ /x;
                            $format = $HEX_FORMAT
                                if ($format eq $RATIONAL_FORMAT
                                       && $map !~
                                           m/ ^ -? [0-9]+ ( \/ [0-9]+ )? $ /x)
                                        # Assume a leading zero means hex,
                                        # even if all digits are 0-9
                                    || ($format eq $INTEGER_FORMAT
                                        && $map =~ /^0[0-9A-F]/);
                            $format = $STRING_FORMAT if $format eq $HEX_FORMAT
                                                       && $map =~ /[^0-9A-F]/;
                        }
                    }
                }
            }
        } # end of calculating format

        if ($default_map eq $CODE_POINT
            && $format ne $HEX_FORMAT
            && ! defined $self->format)    # manual settings are always
                                           # considered ok
        {
            Carp::my_carp_bug("Expecting hex format for mapping table for $self, instead got '$format'")
        }

        # If the output is to be adjusted, the format of the table that gets
        # output is actually 'a' or 'ax' instead of whatever it is stored
        # internally as.
        my $output_adjusted = ($self->to_output_map == $OUTPUT_ADJUSTED);
        if ($output_adjusted) {
            if ($default_map eq $CODE_POINT) {
                $format = $HEX_ADJUST_FORMAT;
            }
            else {
                $format = $ADJUST_FORMAT;
            }
        }

        $self->_set_format($format);

        return $self->SUPER::write(
            $output_adjusted,
            $default_map);   # don't write defaulteds
    }

    # Accessors for the underlying list that should fail if locked.
    for my $sub (qw(
                    add_duplicate
                    replace_map
                ))
    {
        no strict "refs";
        *$sub = sub {
            use strict "refs";
            my $self = shift;

            return if $self->carp_if_locked;
            return $self->_range_list->$sub(@_);
        }
    }
} # End closure for Map_Table

package Match_Table;
use parent '-norequire', '_Base_Table';

# A Match table is one which is a list of all the code points that have
# the same property and property value, for use in \p{property=value}
# constructs in regular expressions.  It adds very little data to the base
# structure, but many methods, as these lists can be combined in many ways to
# form new ones.
# There are only a few concepts added:
# 1) Equivalents and Relatedness.
#    Two tables can match the identical code points, but have different names.
#    This always happens when there is a perl single form extension
#    \p{IsProperty} for the Unicode compound form \P{Property=True}.  The two
#    tables are set to be related, with the Perl extension being a child, and
#    the Unicode property being the parent.
#
#    It may be that two tables match the identical code points and we don't
#    know if they are related or not.  This happens most frequently when the
#    Block and Script properties have the exact range.  But note that a
#    revision to Unicode could add new code points to the script, which would
#    now have to be in a different block (as the block was filled, or there
#    would have been 'Unknown' script code points in it and they wouldn't have
#    been identical).  So we can't rely on any two properties from Unicode
#    always matching the same code points from release to release, and thus
#    these tables are considered coincidentally equivalent--not related.  When
#    two tables are unrelated but equivalent, one is arbitrarily chosen as the
#    'leader', and the others are 'equivalents'.  This concept is useful
#    to minimize the number of tables written out.  Only one file is used for
#    any identical set of code points, with entries in UCD.pl mapping all
#    the involved tables to it.
#
#    Related tables will always be identical; we set them up to be so.  Thus
#    if the Unicode one is deprecated, the Perl one will be too.  Not so for
#    unrelated tables.  Relatedness makes generating the documentation easier.
#
# 2) Complement.
#    Like equivalents, two tables may be the inverses of each other, the
#    intersection between them is null, and the union is every Unicode code
#    point.  The two tables that occupy a binary property are necessarily like
#    this.  By specifying one table as the complement of another, we can avoid
#    storing it on disk (using the other table and performing a fast
#    transform), and some memory and calculations.
#
# 3) Conflicting.  It may be that there will eventually be name clashes, with
#    the same name meaning different things.  For a while, there actually were
#    conflicts, but they have so far been resolved by changing Perl's or
#    Unicode's definitions to match the other, but when this code was written,
#    it wasn't clear that that was what was going to happen.  (Unicode changed
#    because of protests during their beta period.)  Name clashes are warned
#    about during compilation, and the documentation.  The generated tables
#    are sane, free of name clashes, because the code suppresses the Perl
#    version.  But manual intervention to decide what the actual behavior
#    should be may be required should this happen.  The introductory comments
#    have more to say about this.
#
# 4) Definition.  This is a string for human consumption that specifies the
#    code points that this table matches.  This is used only for the generated
#    pod file.  It may be specified explicitly, or automatically computed.
#    Only the first portion of complicated definitions is computed and
#    displayed.

sub standardize { return main::standardize($_[0]); }
sub trace { return main::trace(@_); }


{ # Closure

    main::setup_package();

    my %leader;
    # The leader table of this one; initially $self.
    main::set_access('leader', \%leader, 'r');

    my %equivalents;
    # An array of any tables that have this one as their leader
    main::set_access('equivalents', \%equivalents, 'readable_array');

    my %parent;
    # The parent table to this one, initially $self.  This allows us to
    # distinguish between equivalent tables that are related (for which this
    # is set to), and those which may not be, but share the same output file
    # because they match the exact same set of code points in the current
    # Unicode release.
    main::set_access('parent', \%parent, 'r');

    my %children;
    # An array of any tables that have this one as their parent
    main::set_access('children', \%children, 'readable_array');

    my %conflicting;
    # Array of any tables that would have the same name as this one with
    # a different meaning.  This is used for the generated documentation.
    main::set_access('conflicting', \%conflicting, 'readable_array');

    my %matches_all;
    # Set in the constructor for tables that are expected to match all code
    # points.
    main::set_access('matches_all', \%matches_all, 'r');

    my %complement;
    # Points to the complement that this table is expressed in terms of; 0 if
    # none.
    main::set_access('complement', \%complement, 'r');

    my %definition;
    # Human readable string of the first few ranges of code points matched by
    # this table
    main::set_access('definition', \%definition, 'r', 's');

    sub new {
        my $class = shift;

        my %args = @_;

        # The property for which this table is a listing of property values.
        my $property = delete $args{'_Property'};

        my $name = delete $args{'Name'};
        my $full_name = delete $args{'Full_Name'};
        $full_name = $name if ! defined $full_name;

        # Optional
        my $initialize = delete $args{'Initialize'};
        my $matches_all = delete $args{'Matches_All'} || 0;
        my $format = delete $args{'Format'};
        my $definition = delete $args{'Definition'} // "";
        # Rest of parameters passed on.

        my $range_list = Range_List->new(Initialize => $initialize,
                                         Owner => $property);

        my $complete = $full_name;
        $complete = '""' if $complete eq "";  # A null name shouldn't happen,
                                              # but this helps debug if it
                                              # does
        # The complete name for a match table includes it's property in a
        # compound form 'property=table', except if the property is the
        # pseudo-property, perl, in which case it is just the single form,
        # 'table' (If you change the '=' must also change the ':' in lots of
        # places in this program that assume an equal sign)
        $complete = $property->full_name . "=$complete" if $property != $perl;

        my $self = $class->SUPER::new(%args,
                                      Name => $name,
                                      Complete_Name => $complete,
                                      Full_Name => $full_name,
                                      _Property => $property,
                                      _Range_List => $range_list,
                                      Format => $EMPTY_FORMAT,
                                      Write_As_Invlist => 1,
                                      );
        my $addr = pack 'J', refaddr $self;

        $conflicting{$addr} = [ ];
        $equivalents{$addr} = [ ];
        $children{$addr} = [ ];
        $matches_all{$addr} = $matches_all;
        $leader{$addr} = $self;
        $parent{$addr} = $self;
        $complement{$addr} = 0;
        $definition{$addr} = $definition;

        if (defined $format && $format ne $EMPTY_FORMAT) {
            Carp::my_carp_bug("'Format' must be '$EMPTY_FORMAT' in a match table instead of '$format'.  Using '$EMPTY_FORMAT'");
        }

        return $self;
    }

    # See this program's beginning comment block about overloading these.
    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
        '=' => sub {
                    my $self = shift;

                    return if $self->carp_if_locked;
                    return $self;
                },

        '+' => sub {
                        my $self = shift;
                        my $other = shift;

                        return $self->_range_list + $other;
                    },
        '&' => sub {
                        my $self = shift;
                        my $other = shift;

                        return $self->_range_list & $other;
                    },
        '+=' => sub {
                        my $self = shift;
                        my $other = shift;
                        my $reversed = shift;

                        if ($reversed) {
                            Carp::my_carp_bug("Bad news.  Can't cope with '"
                            . ref($other)
                            . ' += '
                            . ref($self)
                            . "'.  undef returned.");
                            return;
                        }

                        return if $self->carp_if_locked;

                        if (ref $other) {

                            # Change the range list of this table to be the
                            # union of the two.
                            $self->_set_range_list($self->_range_list
                                                    + $other);
                        }
                        else {    # $other is just a simple value
                            $self->add_range($other, $other);
                        }
                        return $self;
                    },
        '&=' => sub {
                        my $self = shift;
                        my $other = shift;
                        my $reversed = shift;

                        if ($reversed) {
                            Carp::my_carp_bug("Bad news.  Can't cope with '"
                            . ref($other)
                            . ' &= '
                            . ref($self)
                            . "'.  undef returned.");
                            return;
                        }

                        return if $self->carp_if_locked;
                        $self->_set_range_list($self->_range_list & $other);
                        return $self;
                    },
        '-' => sub { my $self = shift;
                    my $other = shift;
                    my $reversed = shift;
                    if ($reversed) {
                        Carp::my_carp_bug("Bad news.  Can't cope with '"
                        . ref($other)
                        . ' - '
                        . ref($self)
                        . "'.  undef returned.");
                        return;
                    }

                    return $self->_range_list - $other;
                },
        '~' => sub { my $self = shift;
                    return ~ $self->_range_list;
                },
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {

        my $name = $self->complete_name;
        return "Table '$name'";
    }

    sub _range_list {
        # Returns the range list associated with this table, which will be the
        # complement's if it has one.

        my $self = shift;
        my $complement = $self->complement;

        # In order to avoid re-complementing on each access, only do the
        # complement the first time, and store the result in this table's
        # range list to use henceforth.  However, this wouldn't work if the
        # controlling (complement) table changed after we do this, so lock it.
        # Currently, the value of the complement isn't needed until after it
        # is fully constructed, so this works.  If this were to change, the
        # each_range iteration functionality would no longer work on this
        # complement.
        if ($complement != 0 && $self->SUPER::_range_list->count == 0) {
            $self->_set_range_list($self->SUPER::_range_list
                                + ~ $complement->_range_list);
            $complement->lock;
        }

        return $self->SUPER::_range_list;
    }

    sub add_alias {
        # Add a synonym for this table.  See the comments in the base class

        my $self = shift;
        my $name = shift;
        # Rest of parameters passed on.

        $self->SUPER::add_alias($name, $self, @_);
        return;
    }

    sub add_conflicting {
        # Add the name of some other object to the list of ones that name
        # clash with this match table.

        my $self = shift;
        my $conflicting_name = shift;   # The name of the conflicting object
        my $p = shift || 'p';           # Optional, is this a \p{} or \P{} ?
        my $conflicting_object = shift; # Optional, the conflicting object
                                        # itself.  This is used to
                                        # disambiguate the text if the input
                                        # name is identical to any of the
                                        # aliases $self is known by.
                                        # Sometimes the conflicting object is
                                        # merely hypothetical, so this has to
                                        # be an optional parameter.
        Carp::carp_extra_args(\@_) if main::DEBUG && @_;

        my $addr = pack 'J', refaddr $self;

        # Check if the conflicting name is exactly the same as any existing
        # alias in this table (as long as there is a real object there to
        # disambiguate with).
        if (defined $conflicting_object) {
            foreach my $alias ($self->aliases) {
                if (standardize($alias->name) eq standardize($conflicting_name)) {

                    # Here, there is an exact match.  This results in
                    # ambiguous comments, so disambiguate by changing the
                    # conflicting name to its object's complete equivalent.
                    $conflicting_name = $conflicting_object->complete_name;
                    last;
                }
            }
        }

        # Convert to the \p{...} final name
        $conflicting_name = "\\$p" . "{$conflicting_name}";

        # Only add once
        return if grep { $conflicting_name eq $_ } @{$conflicting{$addr}};

        push @{$conflicting{$addr}}, $conflicting_name;

        return;
    }

    sub is_set_equivalent_to($self, $other=undef) {
        # Return boolean of whether or not the other object is a table of this
        # type and has been marked equivalent to this one.

        return 0 if ! defined $other; # Can happen for incomplete early
                                      # releases
        unless ($other->isa(__PACKAGE__)) {
            my $ref_other = ref $other;
            my $ref_self = ref $self;
            Carp::my_carp_bug("Argument to 'is_set_equivalent_to' must be another $ref_self, not a '$ref_other'.  $other not set equivalent to $self.");
            return 0;
        }

        # Two tables are equivalent if they have the same leader.
        return $leader{pack 'J', refaddr $self} == $leader{pack 'J', refaddr $other};
        return;
    }

    sub set_equivalent_to {
        # Set $self equivalent to the parameter table.
        # The required Related => 'x' parameter is a boolean indicating
        # whether these tables are related or not.  If related, $other becomes
        # the 'parent' of $self; if unrelated it becomes the 'leader'
        #
        # Related tables share all characteristics except names; equivalents
        # not quite so many.
        # If they are related, one must be a perl extension.  This is because
        # we can't guarantee that Unicode won't change one or the other in a
        # later release even if they are identical now.

        my $self = shift;
        my $other = shift;

        my %args = @_;
        my $related = delete $args{'Related'};

        Carp::carp_extra_args(\%args) if main::DEBUG && %args;

        return if ! defined $other;     # Keep on going; happens in some early
                                        # Unicode releases.

        if (! defined $related) {
            Carp::my_carp_bug("set_equivalent_to must have 'Related => [01] parameter.  Assuming $self is not related to $other");
            $related = 0;
        }

        # If already are equivalent, no need to re-do it;  if subroutine
        # returns null, it found an error, also do nothing
        my $are_equivalent = $self->is_set_equivalent_to($other);
        return if ! defined $are_equivalent || $are_equivalent;

        my $addr = pack 'J', refaddr $self;
        my $current_leader = ($related) ? $parent{$addr} : $leader{$addr};

        if ($related) {
            if ($current_leader->perl_extension) {
                if ($other->perl_extension) {
                    Carp::my_carp_bug("Use add_alias() to set two Perl tables '$self' and '$other', equivalent.");
                    return;
                }
            } elsif ($self->property != $other->property    # Depending on
                                                            # situation, might
                                                            # be better to use
                                                            # add_alias()
                                                            # instead for same
                                                            # property
                     && ! $other->perl_extension

                         # We allow the sc and scx properties to be marked as
                         # related.  They are in fact related, and this allows
                         # the pod to show that better.  This test isn't valid
                         # if this is an early Unicode release without the scx
                         # property (having that also implies the sc property
                         # exists, so don't have to test for no 'sc')
                     && (   ! defined $scx
                         && ! (   (   $self->property == $script
                                   || $self->property == $scx)
                               && (   $self->property == $script
                                   || $self->property == $scx))))
            {
                Carp::my_carp_bug("set_equivalent_to should have 'Related => 0 for equivalencing two Unicode properties.  Assuming $self is not related to $other");
                $related = 0;
            }
        }

        if (! $self->is_empty && ! $self->matches_identically_to($other)) {
            Carp::my_carp_bug("$self should be empty or match identically to $other.  Not setting equivalent");
            return;
        }

        my $leader = pack 'J', refaddr $current_leader;
        my $other_addr = pack 'J', refaddr $other;

        # Any tables that are equivalent to or children of this table must now
        # instead be equivalent to or (children) to the new leader (parent),
        # still equivalent.  The equivalency includes their matches_all info,
        # and for related tables, their fate and status.
        # All related tables are of necessity equivalent, but the converse
        # isn't necessarily true
        my $status = $other->status;
        my $status_info = $other->status_info;
        my $fate = $other->fate;
        my $matches_all = $matches_all{other_addr};
        my $caseless_equivalent = $other->caseless_equivalent;
        foreach my $table ($current_leader, @{$equivalents{$leader}}) {
            next if $table == $other;
            trace "setting $other to be the leader of $table, status=$status" if main::DEBUG && $to_trace;

            my $table_addr = pack 'J', refaddr $table;
            $leader{$table_addr} = $other;
            $matches_all{$table_addr} = $matches_all;
            $self->_set_range_list($other->_range_list);
            push @{$equivalents{$other_addr}}, $table;
            if ($related) {
                $parent{$table_addr} = $other;
                push @{$children{$other_addr}}, $table;
                $table->set_status($status, $status_info);

                # This reason currently doesn't get exposed outside; otherwise
                # would have to look up the parent's reason and use it instead.
                $table->set_fate($fate, "Parent's fate");

                $self->set_caseless_equivalent($caseless_equivalent);
            }
        }

        # Now that we've declared these to be equivalent, any changes to one
        # of the tables would invalidate that equivalency.
        $self->lock;
        $other->lock;
        return;
    }

    sub set_complement($self, $other) {
        # Set $self to be the complement of the parameter table.  $self is
        # locked, as what it contains should all come from the other table.

        if ($other->complement != 0) {
            Carp::my_carp_bug("Can't set $self to be the complement of $other, which itself is the complement of " . $other->complement);
            return;
        }
        $complement{pack 'J', refaddr $self} = $other;

        # Be sure the other property knows we are depending on them; or the
        # other table if it is one in the current property.
        if ($self->property != $other->property) {
            $other->property->set_has_dependency(1);
        }
        else {
            $other->set_has_dependency(1);
        }
        $self->lock;
        return;
    }

    sub add_range($self, @range) { # Add a range to the list for this table.
        # Rest of parameters passed on

        return if $self->carp_if_locked;
        return $self->_range_list->add_range(@range);
    }

    sub header($self) {
        # All match tables are to be used only by the Perl core.
        return $self->SUPER::header() . $INTERNAL_ONLY_HEADER;
    }

    sub pre_body {  # Does nothing for match tables.
        return
    }

    sub append_to_body {  # Does nothing for match tables.
        return
    }

    sub set_fate($self, $fate, $reason=undef) {
        $self->SUPER::set_fate($fate, $reason);

        # All children share this fate
        foreach my $child ($self->children) {
            $child->set_fate($fate, $reason);
        }
        return;
    }

    sub calculate_table_definition
    {
        # Returns a human-readable string showing some or all of the code
        # points matched by this table.  The string will include a
        # bracketed-character class for all characters matched in the 00-FF
        # range, and the first few ranges matched beyond that.
        my $max_ranges = 6;

        my $self = shift;
        my $definition = $self->definition || "";

        # Skip this if already have a definition.
        return $definition if $definition;

        my $lows_string = "";   # The string representation of the 0-FF
                                # characters
        my $string_range = "";  # The string rep. of the above FF ranges
        my $range_count = 0;    # How many ranges in $string_rage

        my @lows_invlist;       # The inversion list of the 0-FF code points
        my $first_non_control = ord(" ");   # Everything below this is a
                                            # control, on ASCII or EBCDIC
        my $max_table_code_point = $self->max;

        # On ASCII platforms, the range 80-FF contains no printables.
        my $highest_printable = ((main::NON_ASCII_PLATFORM) ? 255 : 126);


        # Look through the first few ranges matched by this table.
        $self->reset_each_range;    # Defensive programming
        while (defined (my $range = $self->each_range())) {
            my $start = $range->start;
            my $end = $range->end;

            # Accumulate an inversion list of the 00-FF code points
            if ($start < 256 && ($start > 0 || $end < 256)) {
                push @lows_invlist, $start;
                push @lows_invlist, 1 + (($end < 256) ? $end : 255);

                # Get next range if there are more ranges below 256
                next if $end < 256 && $end < $max_table_code_point;

                # If the range straddles the 255/256 boundary, we split it
                # there.  We already added above the low portion to the
                # inversion list
                $start = 256 if $end > 256;
            }

            # Here, @lows_invlist contains the code points below 256, and
            # there is no other range, or the current one starts at or above
            # 256.  Generate the [char class] for the 0-255 ones.
            while (@lows_invlist) {

                # If this range (necessarily the first one, by the way) starts
                # at 0 ...
                if ($lows_invlist[0] == 0) {

                    # If it ends within the block of controls, that means that
                    # some controls are in it and some aren't.  Since Unicode
                    # properties pretty much only know about a few of the
                    # controls, like \n, \t, this means that its one of them
                    # that isn't in the range.  Complement the inversion list
                    # which will likely cause these to be output using their
                    # mnemonics, hence being clearer.
                    if ($lows_invlist[1] < $first_non_control) {
                        $lows_string .= '^';
                        shift @lows_invlist;
                        push @lows_invlist, 256;
                    }
                    elsif ($lows_invlist[1] <= $highest_printable) {

                        # Here, it extends into the printables block.  Split
                        # into two ranges so that the controls are separate.
                        $lows_string .= sprintf "\\x00-\\x%02x",
                                                    $first_non_control - 1;
                        $lows_invlist[0] = $first_non_control;
                    }
                }

                # If the range completely contains the printables, don't
                # individually spell out the printables.
                if (    $lows_invlist[0] <= $first_non_control
                    && $lows_invlist[1] > $highest_printable)
                {
                    $lows_string .= sprintf "\\x%02x-\\x%02x",
                                        $lows_invlist[0], $lows_invlist[1] - 1;
                    shift @lows_invlist;
                    shift @lows_invlist;
                    next;
                }

                # Here, the range may include some but not all printables.
                # Look at each one individually
                foreach my $ord (shift @lows_invlist .. shift(@lows_invlist) - 1) {
                    my $char = chr $ord;

                    # If there is already something in the list, an
                    # alphanumeric char could be the next in sequence.  If so,
                    # we start or extend a range.  That is, we could have so
                    # far something like 'a-c', and the next char is a 'd', so
                    # we change it to 'a-d'.  We use native_to_unicode()
                    # because a-z on EBCDIC means 26 chars, and excludes the
                    # gap ones.
                    if ($lows_string ne "" && $char =~ /[[:alnum:]]/) {
                        my $prev = substr($lows_string, -1);
                        if (   $prev !~ /[[:alnum:]]/
                            ||   utf8::native_to_unicode(ord $prev) + 1
                              != utf8::native_to_unicode(ord $char))
                        {
                            # Not extending the range
                            $lows_string .= $char;
                        }
                        elsif (   length $lows_string > 1
                               && substr($lows_string, -2, 1) eq '-')
                        {
                            # We had a sequence like '-c' and the current
                            # character is 'd'.  Extend the range.
                            substr($lows_string, -1, 1) = $char;
                        }
                        else {
                            # We had something like 'd' and this is 'e'.
                            # Start a range.
                            $lows_string .= "-$char";
                        }
                    }
                    elsif ($char =~ /[[:graph:]]/) {

                        # We output a graphic char as-is, preceded by a
                        # backslash if it is a metacharacter
                        $lows_string .= '\\'
                                if $char =~ /[\\\^\$\@\%\|()\[\]\{\}\-\/"']/;
                        $lows_string .= $char;
                    } # Otherwise use mnemonic for any that have them
                    elsif ($char =~ /[\a]/) {
                        $lows_string .= '\a';
                    }
                    elsif ($char =~ /[\b]/) {
                        $lows_string .= '\b';
                    }
                    elsif ($char eq "\e") {
                        $lows_string .= '\e';
                    }
                    elsif ($char eq "\f") {
                        $lows_string .= '\f';
                    }
                    elsif ($char eq "\cK") {
                        $lows_string .= '\cK';
                    }
                    elsif ($char eq "\n") {
                        $lows_string .= '\n';
                    }
                    elsif ($char eq "\r") {
                        $lows_string .= '\r';
                    }
                    elsif ($char eq "\t") {
                        $lows_string .= '\t';
                    }
                    else {

                        # Here is a non-graphic without a mnemonic.  We use \x
                        # notation.  But if the ordinal of this is one above
                        # the previous, create or extend the range
                        my $hex_representation = sprintf("%02x", ord $char);
                        if (   length $lows_string >= 4
                            && substr($lows_string, -4, 2) eq '\\x'
                            && hex(substr($lows_string, -2)) + 1 == ord $char)
                        {
                            if (       length $lows_string >= 5
                                &&     substr($lows_string, -5, 1) eq '-'
                                && (   length $lows_string == 5
                                    || substr($lows_string, -6, 1) ne '\\'))
                            {
                                substr($lows_string, -2) = $hex_representation;
                            }
                            else {
                                $lows_string .= '-\\x' . $hex_representation;
                            }
                        }
                        else {
                            $lows_string .= '\\x' . $hex_representation;
                        }
                    }
                }
            }

            # Done with assembling the string of all lows.  If there are only
            # lows in the property, are completely done.
            if ($max_table_code_point < 256) {
                $self->reset_each_range;
                last;
            }

            # Otherwise, quit if reached max number of non-lows ranges.  If
            # there are lows, count them as one unit towards the maximum.
            $range_count++;
            if ($range_count > (($lows_string eq "") ? $max_ranges : $max_ranges - 1)) {
                $string_range .= " ...";
                $self->reset_each_range;
                last;
            }

            # Otherwise add this range.
            $string_range .= ", " if $string_range ne "";
            if ($start == $end) {
                $string_range .= sprintf("U+%04X", $start);
            }
            elsif ($end >= $MAX_WORKING_CODEPOINT)  {
                $string_range .= sprintf("U+%04X..infinity", $start);
            }
            else  {
                $string_range .= sprintf("U+%04X..%04X",
                                        $start, $end);
            }
        }

        # Done with all the ranges we're going to look at.  Assemble the
        # definition from the lows + non-lows.

        if ($lows_string ne "" || $string_range ne "") {
            if ($lows_string ne "") {
                $definition .= "[$lows_string]";
                $definition .= ", " if $string_range;
            }
            $definition .= $string_range;
        }

        return $definition;
    }

    sub write($self) {
        return $self->SUPER::write(0); # No adjustments
    }

    # $leader - Should only be called on the leader table of an equivalent group
    sub set_final_comment($leader) {
        # This creates a comment for the file that is to hold the match table
        # $self.  It is somewhat convoluted to make the English read nicely,
        # but, heh, it's just a comment.
        # This should be called only with the leader match table of all the
        # ones that share the same file.  It lists all such tables, ordered so
        # that related ones are together.

        return unless $debugging_build;

        my $addr = pack 'J', refaddr $leader;

        if ($leader{$addr} != $leader) {
            Carp::my_carp_bug(<<END
set_final_comment() must be called on a leader table, which $leader is not.
It is equivalent to $leader{$addr}.  No comment created
END
            );
            return;
        }

        # Get the number of code points matched by each of the tables in this
        # file, and add underscores for clarity.
        my $count = $leader->count;
        my $unicode_count;
        my $non_unicode_string;
        if ($count > $MAX_UNICODE_CODEPOINTS) {
            $unicode_count = $count - ($MAX_WORKING_CODEPOINT
                                       - $MAX_UNICODE_CODEPOINT);
            $non_unicode_string = "All above-Unicode code points match as well, and are also returned";
        }
        else {
            $unicode_count = $count;
            $non_unicode_string = "";
        }
        my $string_count = main::clarify_code_point_count($unicode_count);

        my $loose_count = 0;        # how many aliases loosely matched
        my $compound_name = "";     # ? Are any names compound?, and if so, an
                                    # example
        my $properties_with_compound_names = 0;    # count of these


        my %flags;              # The status flags used in the file
        my $total_entries = 0;  # number of entries written in the comment
        my $matches_comment = ""; # The portion of the comment about the
                                  # \p{}'s
        my @global_comments;    # List of all the tables' comments that are
                                # there before this routine was called.
        my $has_ucd_alias = 0;  # If there is an alias that is accessible via
                                # Unicode::UCD.  If not, then don't say it is
                                # in the comment

        # Get list of all the parent tables that are equivalent to this one
        # (including itself).
        my @parents = grep { $parent{main::objaddr $_} == $_ }
                            main::uniques($leader, @{$equivalents{$addr}});
        my $has_unrelated = (@parents >= 2);  # boolean, ? are there unrelated
                                              # tables
        for my $parent (@parents) {

            my $property = $parent->property;

            # Special case 'N' tables in properties with two match tables when
            # the other is a 'Y' one.  These are likely to be binary tables,
            # but not necessarily.  In either case, \P{} will match the
            # complement of \p{}, and so if something is a synonym of \p, the
            # complement of that something will be the synonym of \P.  This
            # would be true of any property with just two match tables, not
            # just those whose values are Y and N; but that would require a
            # little extra work, and there are none such so far in Unicode.
            my $perl_p = 'p';        # which is it?  \p{} or \P{}
            my @yes_perl_synonyms;   # list of any synonyms for the 'Y' table

            if (scalar $property->tables == 2
                && $parent == $property->table('N')
                && defined (my $yes = $property->table('Y')))
            {
                my $yes_addr = pack 'J', refaddr $yes;
                @yes_perl_synonyms
                    = grep { $_->property == $perl }
                                    main::uniques($yes,
                                                $parent{$yes_addr},
                                                $parent{$yes_addr}->children);

                # But these synonyms are \P{} ,not \p{}
                $perl_p = 'P';
            }

            my @description;        # Will hold the table description
            my @note;               # Will hold the table notes.
            my @conflicting;        # Will hold the table conflicts.

            # Look at the parent, any yes synonyms, and all the children
            my $parent_addr = pack 'J', refaddr $parent;
            for my $table ($parent,
                           @yes_perl_synonyms,
                           @{$children{$parent_addr}})
            {
                my $table_addr = pack 'J', refaddr $table;
                my $table_property = $table->property;

                # Tables are separated by a blank line to create a grouping.
                $matches_comment .= "\n" if $matches_comment;

                # The table is named based on the property and value
                # combination it is for, like script=greek.  But there may be
                # a number of synonyms for each side, like 'sc' for 'script',
                # and 'grek' for 'greek'.  Any combination of these is a valid
                # name for this table.  In this case, there are three more,
                # 'sc=grek', 'sc=greek', and 'script='grek'.  Rather than
                # listing all possible combinations in the comment, we make
                # sure that each synonym occurs at least once, and add
                # commentary that the other combinations are possible.
                # Because regular expressions don't recognize things like
                # \p{jsn=}, only look at non-null right-hand-sides
                my @property_aliases = grep { $_->status ne $INTERNAL_ALIAS } $table_property->aliases;
                my @table_aliases = grep { $_->name ne "" } $table->aliases;

                # The alias lists above are already ordered in the order we
                # want to output them.  To ensure that each synonym is listed,
                # we must use the max of the two numbers.  But if there are no
                # legal synonyms (nothing in @table_aliases), then we don't
                # list anything.
                my $listed_combos = (@table_aliases)
                                    ?  main::max(scalar @table_aliases,
                                                 scalar @property_aliases)
                                    : 0;
                trace "$listed_combos, tables=", scalar @table_aliases, "; property names=", scalar @property_aliases if main::DEBUG;

                my $property_had_compound_name = 0;

                for my $i (0 .. $listed_combos - 1) {
                    $total_entries++;

                    # The current alias for the property is the next one on
                    # the list, or if beyond the end, start over.  Similarly
                    # for the table (\p{prop=table})
                    my $property_alias = $property_aliases
                                            [$i % @property_aliases]->name;
                    my $table_alias_object = $table_aliases
                                                        [$i % @table_aliases];
                    my $table_alias = $table_alias_object->name;
                    my $loose_match = $table_alias_object->loose_match;
                    $has_ucd_alias |= $table_alias_object->ucd;

                    if ($table_alias !~ /\D/) { # Clarify large numbers.
                        $table_alias = main::clarify_number($table_alias)
                    }

                    # Add a comment for this alias combination
                    my $current_match_comment;
                    if ($table_property == $perl) {
                        $current_match_comment = "\\$perl_p"
                                                    . "{$table_alias}";
                    }
                    else {
                        $current_match_comment
                                        = "\\p{$property_alias=$table_alias}";
                        $property_had_compound_name = 1;
                    }

                    # Flag any abnormal status for this table.
                    my $flag = $property->status
                                || $table->status
                                || $table_alias_object->status;
                    if ($flag && $flag ne $PLACEHOLDER) {
                        $flags{$flag} = $status_past_participles{$flag};
                    }

                    $loose_count++;

                    # Pretty up the comment.  Note the \b; it says don't make
                    # this line a continuation.
                    $matches_comment .= sprintf("\b%-1s%-s%s\n",
                                        $flag,
                                        " " x 7,
                                        $current_match_comment);
                } # End of generating the entries for this table.

                # Save these for output after this group of related tables.
                push @description, $table->description;
                push @note, $table->note;
                push @conflicting, $table->conflicting;

                # And this for output after all the tables.
                push @global_comments, $table->comment;

                # Compute an alternate compound name using the final property
                # synonym and the first table synonym with a colon instead of
                # the equal sign used elsewhere.
                if ($property_had_compound_name) {
                    $properties_with_compound_names ++;
                    if (! $compound_name || @property_aliases > 1) {
                        $compound_name = $property_aliases[-1]->name
                                        . ': '
                                        . $table_aliases[0]->name;
                    }
                }
            } # End of looping through all children of this table

            # Here have assembled in $matches_comment all the related tables
            # to the current parent (preceded by the same info for all the
            # previous parents).  Put out information that applies to all of
            # the current family.
            if (@conflicting) {

                # But output the conflicting information now, as it applies to
                # just this table.
                my $conflicting = join ", ", @conflicting;
                if ($conflicting) {
                    $matches_comment .= <<END;

    Note that contrary to what you might expect, the above is NOT the same as
END
                    $matches_comment .= "any of: " if @conflicting > 1;
                    $matches_comment .= "$conflicting\n";
                }
            }
            if (@description) {
                $matches_comment .= "\n    Meaning: "
                                    . join('; ', @description)
                                    . "\n";
            }
            if (@note) {
                $matches_comment .= "\n    Note: "
                                    . join("\n    ", @note)
                                    . "\n";
            }
        } # End of looping through all tables

        $matches_comment .= "\n$non_unicode_string\n" if $non_unicode_string;


        my $code_points;
        my $match;
        my $any_of_these;
        if ($unicode_count == 1) {
            $match = 'matches';
            $code_points = 'single code point';
        }
        else {
            $match = 'match';
            $code_points = "$string_count code points";
        }

        my $synonyms;
        my $entries;
        if ($total_entries == 1) {
            $synonyms = "";
            $entries = 'entry';
            $any_of_these = 'this'
        }
        else {
            $synonyms = " any of the following regular expression constructs";
            $entries = 'entries';
            $any_of_these = 'any of these'
        }

        my $comment = "";
        if ($has_ucd_alias) {
            $comment .= "Use Unicode::UCD::prop_invlist() to access the contents of this file.\n\n";
        }
        if ($has_unrelated) {
            $comment .= <<END;
This file is for tables that are not necessarily related:  To conserve
resources, every table that matches the identical set of code points in this
version of Unicode uses this file.  Each one is listed in a separate group
below.  It could be that the tables will match the same set of code points in
other Unicode releases, or it could be purely coincidence that they happen to
be the same in Unicode $unicode_version, and hence may not in other versions.

END
        }

        if (%flags) {
            foreach my $flag (sort keys %flags) {
                $comment .= <<END;
'$flag' below means that this form is $flags{$flag}.
END
                if ($flag eq $INTERNAL_ALIAS) {
                    $comment .= "DO NOT USE!!!";
                }
                else {
                    $comment .= "Consult $pod_file.pod";
                }
                $comment .= "\n";
            }
            $comment .= "\n";
        }

        if ($total_entries == 0) {
            Carp::my_carp("No regular expression construct can match $leader, as all names for it are the null string.  Creating file anyway.");
            $comment .= <<END;
This file returns the $code_points in Unicode Version
$unicode_version for
$leader, but it is inaccessible through Perl regular expressions, as
"\\p{prop=}" is not recognized.
END

        } else {
            $comment .= <<END;
This file returns the $code_points in Unicode Version
$unicode_version that
$match$synonyms:

$matches_comment
$pod_file.pod should be consulted for the syntax rules for $any_of_these,
including if adding or subtracting white space, underscore, and hyphen
characters matters or doesn't matter, and other permissible syntactic
variants.  Upper/lower case distinctions never matter.
END

        }
        if ($compound_name) {
            $comment .= <<END;

A colon can be substituted for the equals sign, and
END
            if ($properties_with_compound_names > 1) {
                $comment .= <<END;
within each group above,
END
            }
            $compound_name = sprintf("%-8s\\p{%s}", " ", $compound_name);

            # Note the \b below, it says don't make that line a continuation.
            $comment .= <<END;
anything to the left of the equals (or colon) can be combined with anything to
the right.  Thus, for example,
$compound_name
\bis also valid.
END
        }

        # And append any comment(s) from the actual tables.  They are all
        # gathered here, so may not read all that well.
        if (@global_comments) {
            $comment .= "\n" . join("\n\n", @global_comments) . "\n";
        }

        if ($count) {   # The format differs if no code points, and needs no
                        # explanation in that case
            if ($leader->write_as_invlist) {
                $comment.= <<END;

The first data line of this file begins with the letter V to indicate it is in
inversion list format.  The number following the V gives the number of lines
remaining.  Each of those remaining lines is a single number representing the
starting code point of a range which goes up to but not including the number
on the next line; The 0th, 2nd, 4th... ranges are for code points that match
the property; the 1st, 3rd, 5th... are ranges of code points that don't match
the property.  The final line's range extends to the platform's infinity.
END
            }
            else {
                $comment.= <<END;
The format of the lines of this file is:
START\\tSTOP\\twhere START is the starting code point of the range, in hex;
STOP is the ending point, or if omitted, the range has just one code point.
END
            }
            if ($leader->output_range_counts) {
                $comment .= <<END;
Numbers in comments in [brackets] indicate how many code points are in the
range.
END
            }
        }

        $leader->set_comment(main::join_lines($comment));
        return;
    }

    # Accessors for the underlying list
    for my $sub (qw(
                    get_valid_code_point
                    get_invalid_code_point
                ))
    {
        no strict "refs";
        *$sub = sub {
            use strict "refs";
            my $self = shift;

            return $self->_range_list->$sub(@_);
        }
    }
} # End closure for Match_Table

package Property;

# The Property class represents a Unicode property, or the $perl
# pseudo-property.  It contains a map table initialized empty at construction
# time, and for properties accessible through regular expressions, various
# match tables, created through the add_match_table() method, and referenced
# by the table('NAME') or tables() methods, the latter returning a list of all
# of the match tables.  Otherwise table operations implicitly are for the map
# table.
#
# Most of the data in the property is actually about its map table, so it
# mostly just uses that table's accessors for most methods.  The two could
# have been combined into one object, but for clarity because of their
# differing semantics, they have been kept separate.  It could be argued that
# the 'file' and 'directory' fields should be kept with the map table.
#
# Each property has a type.  This can be set in the constructor, or in the
# set_type accessor, but mostly it is figured out by the data.  Every property
# starts with unknown type, overridden by a parameter to the constructor, or
# as match tables are added, or ranges added to the map table, the data is
# inspected, and the type changed.  After the table is mostly or entirely
# filled, compute_type() should be called to finalize the analysis.
#
# There are very few operations defined.  One can safely remove a range from
# the map table, and property_add_or_replace_non_nulls() adds the maps from another
# table to this one, replacing any in the intersection of the two.

sub standardize { return main::standardize($_[0]); }
sub trace { return main::trace(@_) if main::DEBUG && $to_trace }

{   # Closure

    # This hash will contain as keys, all the aliases of all properties, and
    # as values, pointers to their respective property objects.  This allows
    # quick look-up of a property from any of its names.
    my %alias_to_property_of;

    sub dump_alias_to_property_of {
        # For debugging

        print "\n", main::simple_dumper (\%alias_to_property_of), "\n";
        return;
    }

    sub property_ref($name) {
        # This is a package subroutine, not called as a method.
        # If the single parameter is a literal '*' it returns a list of all
        # defined properties.
        # Otherwise, the single parameter is a name, and it returns a pointer
        # to the corresponding property object, or undef if none.
        #
        # Properties can have several different names.  The 'standard' form of
        # each of them is stored in %alias_to_property_of as they are defined.
        # But it's possible that this subroutine will be called with some
        # variant, so if the initial lookup fails, it is repeated with the
        # standardized form of the input name.  If found, besides returning the
        # result, the input name is added to the list so future calls won't
        # have to do the conversion again.

        if (! defined $name) {
            Carp::my_carp_bug("Undefined input property.  No action taken.");
            return;
        }

        return main::uniques(values %alias_to_property_of) if $name eq '*';

        # Return cached result if have it.
        my $result = $alias_to_property_of{$name};
        return $result if defined $result;

        # Convert the input to standard form.
        my $standard_name = standardize($name);

        $result = $alias_to_property_of{$standard_name};
        return unless defined $result;        # Don't cache undefs

        # Cache the result before returning it.
        $alias_to_property_of{$name} = $result;
        return $result;
    }


    main::setup_package();

    my %map;
    # A pointer to the map table object for this property
    main::set_access('map', \%map);

    my %full_name;
    # The property's full name.  This is a duplicate of the copy kept in the
    # map table, but is needed because stringify needs it during
    # construction of the map table, and then would have a chicken before egg
    # problem.
    main::set_access('full_name', \%full_name, 'r');

    my %table_ref;
    # This hash will contain as keys, all the aliases of any match tables
    # attached to this property, and as values, the pointers to their
    # respective tables.  This allows quick look-up of a table from any of its
    # names.
    main::set_access('table_ref', \%table_ref);

    my %type;
    # The type of the property, $ENUM, $BINARY, etc
    main::set_access('type', \%type, 'r');

    my %file;
    # The filename where the map table will go (if actually written).
    # Normally defaulted, but can be overridden.
    main::set_access('file', \%file, 'r', 's');

    my %directory;
    # The directory where the map table will go (if actually written).
    # Normally defaulted, but can be overridden.
    main::set_access('directory', \%directory, 's');

    my %pseudo_map_type;
    # This is used to affect the calculation of the map types for all the
    # ranges in the table.  It should be set to one of the values that signify
    # to alter the calculation.
    main::set_access('pseudo_map_type', \%pseudo_map_type, 'r');

    my %has_only_code_point_maps;
    # A boolean used to help in computing the type of data in the map table.
    main::set_access('has_only_code_point_maps', \%has_only_code_point_maps);

    my %unique_maps;
    # A list of the first few distinct mappings this property has.  This is
    # used to disambiguate between binary and enum property types, so don't
    # have to keep more than three.
    main::set_access('unique_maps', \%unique_maps);

    my %pre_declared_maps;
    # A boolean that gives whether the input data should declare all the
    # tables used, or not.  If the former, unknown ones raise a warning.
    main::set_access('pre_declared_maps',
                                    \%pre_declared_maps, 'r', 's');

    my %match_subdir;
    # For properties whose shortest names are too long for a DOS 8.3
    # filesystem to distinguish between, this is used to manually give short
    # names for the directory name immediately under $match_tables that the
    # match tables for this property should be placed in.
    main::set_access('match_subdir', \%match_subdir, 'r', 's');

    my %has_dependency;
    # A boolean that gives whether some table somewhere is defined as the
    # complement of a table in this property.  This is a crude, but currently
    # sufficient, mechanism to make this property not get destroyed before
    # what is dependent on it is.  Other dependencies could be added, so the
    # name was chosen to reflect a more general situation than actually is
    # currently the case.
    main::set_access('has_dependency', \%has_dependency, 'r', 's');

    sub new {
        # The only required parameter is the positionally first, name.  All
        # other parameters are key => value pairs.  See the documentation just
        # above for the meanings of the ones not passed directly on to the map
        # table constructor.

        my $class = shift;
        my $name = shift || "";

        my $self = property_ref($name);
        if (defined $self) {
            my $options_string = join ", ", @_;
            $options_string = ".  Ignoring options $options_string" if $options_string;
            Carp::my_carp("$self is already in use.  Using existing one$options_string;");
            return $self;
        }

        my %args = @_;

        $self = bless \do { my $anonymous_scalar }, $class;
        my $addr = pack 'J', refaddr $self;

        $directory{$addr} = delete $args{'Directory'};
        $file{$addr} = delete $args{'File'};
        $full_name{$addr} = delete $args{'Full_Name'} || $name;
        $type{$addr} = delete $args{'Type'} || $UNKNOWN;
        $pseudo_map_type{$addr} = delete $args{'Map_Type'};
        $pre_declared_maps{$addr} = delete $args{'Pre_Declared_Maps'}
                                    # Starting in this release, property
                                    # values should be defined for all
                                    # properties, except those overriding this
                                    // $v_version ge v5.1.0;
        $match_subdir{$addr} = delete $args{'Match_SubDir'};

        # Rest of parameters passed on.

        $has_only_code_point_maps{$addr} = 1;
        $table_ref{$addr} = { };
        $unique_maps{$addr} = { };
        $has_dependency{$addr} = 0;

        $map{$addr} = Map_Table->new($name,
                                    Full_Name => $full_name{$addr},
                                    _Alias_Hash => \%alias_to_property_of,
                                    _Property => $self,
                                    %args);
        return $self;
    }

    # See this program's beginning comment block about overloading the copy
    # constructor.  Few operations are defined on properties, but a couple are
    # useful.  It is safe to take the inverse of a property, and to remove a
    # single code point from it.
    use overload
        fallback => 0,
        qw("") => "_operator_stringify",
        "." => \&main::_operator_dot,
        ".=" => \&main::_operator_dot_equal,
        '==' => \&main::_operator_equal,
        '!=' => \&main::_operator_not_equal,
        '=' => sub { return shift },
        '-=' => "_minus_and_equal",
    ;

    sub _operator_stringify($self, $other="", $reversed=0) {
        return "Property '" .  $self->full_name . "'";
    }

    sub _minus_and_equal($self, $other, $reversed=0) {
        # Remove a single code point from the map table of a property.
        if (ref $other) {
            Carp::my_carp_bug("Bad news.  Can't cope with a "
                        . ref($other)
                        . " argument to '-='.  Subtraction ignored.");
            return $self;
        }
        elsif ($reversed) {   # Shouldn't happen in a -=, but just in case
            Carp::my_carp_bug("Bad news.  Can't cope with subtracting a "
            . ref $self
            . " from a non-object.  undef returned.");
            return;
        }
        else {
            $map{pack 'J', refaddr $self}->delete_range($other, $other);
        }
        return $self;
    }

    sub add_match_table {
        # Add a new match table for this property, with name given by the
        # parameter.  It returns a pointer to the table.

        my $self = shift;
        my $name = shift;
        my %args = @_;

        my $addr = pack 'J', refaddr $self;

        my $table = $table_ref{$addr}{$name};
        my $standard_name = main::standardize($name);
        if (defined $table
            || (defined ($table = $table_ref{$addr}{$standard_name})))
        {
            Carp::my_carp("Table '$name' in $self is already in use.  Using existing one");
            $table_ref{$addr}{$name} = $table;
            return $table;
        }
        else {

            # See if this is a perl extension, if not passed in.
            my $perl_extension = delete $args{'Perl_Extension'};
            $perl_extension
                        = $self->perl_extension if ! defined $perl_extension;

            my $fate;
            my $suppression_reason = "";
            if ($self->name =~ /^_/) {
                $fate = $SUPPRESSED;
                $suppression_reason = "Parent property is internal only";
            }
            elsif ($self->fate >= $SUPPRESSED) {
                $fate = $self->fate;
                $suppression_reason = $why_suppressed{$self->complete_name};

            }
            elsif ($name =~ /^_/) {
                $fate = $INTERNAL_ONLY;
            }
            $table = Match_Table->new(
                                Name => $name,
                                Perl_Extension => $perl_extension,
                                _Alias_Hash => $table_ref{$addr},
                                _Property => $self,
                                Fate => $fate,
                                Suppression_Reason => $suppression_reason,
                                Status => $self->status,
                                _Status_Info => $self->status_info,
                                %args);
            return unless defined $table;
        }

        # Save the names for quick look up
        $table_ref{$addr}{$standard_name} = $table;
        $table_ref{$addr}{$name} = $table;

        # Perhaps we can figure out the type of this property based on the
        # fact of adding this match table.  First, string properties don't
        # have match tables; second, a binary property can't have 3 match
        # tables
        if ($type{$addr} == $UNKNOWN) {
            $type{$addr} = $NON_STRING;
        }
        elsif ($type{$addr} == $STRING) {
            Carp::my_carp("$self Added a match table '$name' to a string property '$self'.  Changed it to a non-string property.  Bad News.");
            $type{$addr} = $NON_STRING;
        }
        elsif ($type{$addr} != $ENUM && $type{$addr} != $FORCED_BINARY) {
            if (scalar main::uniques(values %{$table_ref{$addr}}) > 2) {
                if ($type{$addr} == $BINARY) {
                    Carp::my_carp("$self now has more than 2 tables (with the addition of '$name'), and so is no longer binary.  Changing its type to 'enum'.  Bad News.");
                }
                $type{$addr} = $ENUM;
            }
        }

        return $table;
    }

    sub delete_match_table($self, $table_to_remove) {
        # Delete the table referred to by $2 from the property $1.
        my $addr = pack 'J', refaddr $self;

        # Remove all names that refer to it.
        foreach my $key (keys %{$table_ref{$addr}}) {
            delete $table_ref{$addr}{$key}
                                if $table_ref{$addr}{$key} == $table_to_remove;
        }

        $table_to_remove->DESTROY;
        return;
    }

    sub table($self, $name) {
        # Return a pointer to the match table (with name given by the
        # parameter) associated with this property; undef if none.
        my $addr = pack 'J', refaddr $self;

        return $table_ref{$addr}{$name} if defined $table_ref{$addr}{$name};

        # If quick look-up failed, try again using the standard form of the
        # input name.  If that succeeds, cache the result before returning so
        # won't have to standardize this input name again.
        my $standard_name = main::standardize($name);
        return unless defined $table_ref{$addr}{$standard_name};

        $table_ref{$addr}{$name} = $table_ref{$addr}{$standard_name};
        return $table_ref{$addr}{$name};
    }

    sub tables {
        # Return a list of pointers to all the match tables attached to this
        # property

        return main::uniques(values %{$table_ref{pack 'J', refaddr shift}});
    }

    sub directory {
        # Returns the directory the map table for this property should be
        # output in.  If a specific directory has been specified, that has
        # priority;  'undef' is returned if the type isn't defined;
        # or $map_directory for everything else.

        my $addr = pack 'J', refaddr shift;

        return $directory{$addr} if defined $directory{$addr};
        return undef if $type{$addr} == $UNKNOWN;
        return $map_directory;
    }

    sub swash_name($self) {
        # Return the name that is used to both:
        #   1)  Name the file that the map table is written to.
        #   2)  The name of swash related stuff inside that file.
        # The reason for this is that the Perl core historically has used
        # certain names that aren't the same as the Unicode property names.
        # To continue using these, $file is hard-coded in this file for those,
        # but otherwise the standard name is used.  This is different from the
        # external_name, so that the rest of the files, like in lib can use
        # the standard name always, without regard to historical precedent.
        my $addr = pack 'J', refaddr $self;

        # Swash names are used only on either
        # 1) regular or internal-only map tables
        # 2) otherwise there should be no access to the
        #    property map table from other parts of Perl.
        return if $map{$addr}->fate != $ORDINARY
                  && ! ($map{$addr}->name =~ /^_/
                        && $map{$addr}->fate == $INTERNAL_ONLY);

        return $file{$addr} if defined $file{$addr};
        return $map{$addr}->external_name;
    }

    sub to_create_match_tables($self) {
        # Returns a boolean as to whether or not match tables should be
        # created for this property.

        # The whole point of this pseudo property is match tables.
        return 1 if $self == $perl;

        my $addr = pack 'J', refaddr $self;

        # Don't generate tables of code points that match the property values
        # of a string property.  Such a list would most likely have many
        # property values, each with just one or very few code points mapping
        # to it.
        return 0 if $type{$addr} == $STRING;

        # Otherwise, do.
        return 1;
    }

    sub property_add_or_replace_non_nulls($self, $other) {
        # This adds the mappings in the property $other to $self.  Non-null
        # mappings from $other override those in $self.  It essentially merges
        # the two properties, with the second having priority except for null
        # mappings.

        if (! $other->isa(__PACKAGE__)) {
            Carp::my_carp_bug("$other should be a "
                            . __PACKAGE__
                            . ".  Not a '"
                            . ref($other)
                            . "'.  Not added;");
            return;
        }

        return $map{pack 'J', refaddr $self}->map_add_or_replace_non_nulls($map{pack 'J', refaddr $other});
    }

    sub set_proxy_for {
        # Certain tables are not generally written out to files, but
        # Unicode::UCD has the intelligence to know that the file for $self
        # can be used to reconstruct those tables.  This routine just changes
        # things so that UCD pod entries for those suppressed tables are
        # generated, so the fact that a proxy is used is invisible to the
        # user.

        my $self = shift;

        foreach my $property_name (@_) {
            my $ref = property_ref($property_name);
            next if $ref->to_output_map;
            $ref->set_fate($MAP_PROXIED);
        }
    }

    sub set_type($self, $type) {
        # Set the type of the property.  Mostly this is figured out by the
        # data in the table.  But this is used to set it explicitly.  The
        # reason it is not a standard accessor is that when setting a binary
        # property, we need to make sure that all the true/false aliases are
        # present, as they were omitted in early Unicode releases.

        if ($type != $ENUM
            && $type != $BINARY
            && $type != $FORCED_BINARY
            && $type != $STRING)
        {
            Carp::my_carp("Unrecognized type '$type'.  Type not set");
            return;
        }

        $type{pack 'J', refaddr $self} = $type;
        return if $type != $BINARY && $type != $FORCED_BINARY;

        my $yes = $self->table('Y');
        $yes = $self->table('Yes') if ! defined $yes;
        $yes = $self->add_match_table('Y', Full_Name => 'Yes')
                                                            if ! defined $yes;

        # Add aliases in order wanted, duplicates will be ignored.  We use a
        # binary property present in all releases for its ordered lists of
        # true/false aliases.  Note, that could run into problems in
        # outputting things in that we don't distinguish between the name and
        # full name of these.  Hopefully, if the table was already created
        # before this code is executed, it was done with these set properly.
        my $bm = property_ref("Bidi_Mirrored");
        foreach my $alias ($bm->table("Y")->aliases) {
            $yes->add_alias($alias->name);
        }
        my $no = $self->table('N');
        $no = $self->table('No') if ! defined $no;
        $no = $self->add_match_table('N', Full_Name => 'No') if ! defined $no;
        foreach my $alias ($bm->table("N")->aliases) {
            $no->add_alias($alias->name);
        }

        return;
    }

    sub add_map {
        # Add a map to the property's map table.  This also keeps
        # track of the maps so that the property type can be determined from
        # its data.

        my $self = shift;
        my $start = shift;  # First code point in range
        my $end = shift;    # Final code point in range
        my $map = shift;    # What the range maps to.
        # Rest of parameters passed on.

        my $addr = pack 'J', refaddr $self;

        # If haven't the type of the property, gather information to figure it
        # out.
        if ($type{$addr} == $UNKNOWN) {

            # If the map contains an interior blank or dash, or most other
            # nonword characters, it will be a string property.  This
            # heuristic may actually miss some string properties.  If so, they
            # may need to have explicit set_types called for them.  This
            # happens in the Unihan properties.
            if ($map =~ / (?<= . ) [ -] (?= . ) /x
                || $map =~ / [^\w.\/\ -]  /x)
            {
                $self->set_type($STRING);

                # $unique_maps is used for disambiguating between ENUM and
                # BINARY later; since we know the property is not going to be
                # one of those, no point in keeping the data around
                undef $unique_maps{$addr};
            }
            else {

                # Not necessarily a string.  The final decision has to be
                # deferred until all the data are in.  We keep track of if all
                # the values are code points for that eventual decision.
                $has_only_code_point_maps{$addr} &=
                                            $map =~ / ^ $code_point_re $/x;

                # For the purposes of disambiguating between binary and other
                # enumerations at the end, we keep track of the first three
                # distinct property values.  Once we get to three, we know
                # it's not going to be binary, so no need to track more.
                if (scalar keys %{$unique_maps{$addr}} < 3) {
                    $unique_maps{$addr}{main::standardize($map)} = 1;
                }
            }
        }

        # Add the mapping by calling our map table's method
        return $map{$addr}->add_map($start, $end, $map, @_);
    }

    sub compute_type($self) {
        # Compute the type of the property: $ENUM, $STRING, or $BINARY.  This
        # should be called after the property is mostly filled with its maps.
        # We have been keeping track of what the property values have been,
        # and now have the necessary information to figure out the type.

        my $addr = pack 'J', refaddr $self;

        my $type = $type{$addr};

        # If already have figured these out, no need to do so again, but we do
        # a double check on ENUMS to make sure that a string property hasn't
        # improperly been classified as an ENUM, so continue on with those.
        return if $type == $STRING
                  || $type == $BINARY
                  || $type == $FORCED_BINARY;

        # If every map is to a code point, is a string property.
        if ($type == $UNKNOWN
            && ($has_only_code_point_maps{$addr}
                || (defined $map{$addr}->default_map
                    && $map{$addr}->default_map eq "")))
        {
            $self->set_type($STRING);
        }
        else {

            # Otherwise, it is to some sort of enumeration.  (The case where
            # it is a Unicode miscellaneous property, and treated like a
            # string in this program is handled in add_map()).  Distinguish
            # between binary and some other enumeration type.  Of course, if
            # there are more than two values, it's not binary.  But more
            # subtle is the test that the default mapping is defined means it
            # isn't binary.  This in fact may change in the future if Unicode
            # changes the way its data is structured.  But so far, no binary
            # properties ever have @missing lines for them, so the default map
            # isn't defined for them.  The few properties that are two-valued
            # and aren't considered binary have the default map defined
            # starting in Unicode 5.0, when the @missing lines appeared; and
            # this program has special code to put in a default map for them
            # for earlier than 5.0 releases.
            if ($type == $ENUM
                || scalar keys %{$unique_maps{$addr}} > 2
                || defined $self->default_map)
            {
                my $tables = $self->tables;
                my $count = $self->count;
                if ($verbosity && $tables > 500 && $tables/$count > .1) {
                    Carp::my_carp_bug("It appears that $self should be a \$STRING property, not an \$ENUM because it has too many match tables: $tables\n");
                }
                $self->set_type($ENUM);
            }
            else {
                $self->set_type($BINARY);
            }
        }
        undef $unique_maps{$addr};  # Garbage collect
        return;
    }

    # $reaons - Ignored unless suppressing
    sub set_fate($self, $fate, $reason=undef) {
        my $addr = pack 'J', refaddr $self;
        if ($fate >= $SUPPRESSED) {
            $why_suppressed{$self->complete_name} = $reason;
        }

        # Each table shares the property's fate, except that MAP_PROXIED
        # doesn't affect match tables
        $map{$addr}->set_fate($fate, $reason);
        if ($fate != $MAP_PROXIED) {
            foreach my $table ($map{$addr}, $self->tables) {
                $table->set_fate($fate, $reason);
            }
        }
        return;
    }


    # Most of the accessors for a property actually apply to its map table.
    # Setup up accessor functions for those, referring to %map
    for my $sub (qw(
                    add_alias
                    add_anomalous_entry
                    add_comment
                    add_conflicting
                    add_description
                    add_duplicate
                    add_note
                    aliases
                    comment
                    complete_name
                    containing_range
                    count
                    default_map
                    definition
                    delete_range
                    description
                    each_range
                    external_name
                    fate
                    file_path
                    format
                    initialize
                    inverse_list
                    is_empty
                    name
                    note
                    perl_extension
                    property
                    range_count
                    ranges
                    range_size_1
                    replace_map
                    reset_each_range
                    set_comment
                    set_default_map
                    set_file_path
                    set_final_comment
                    _set_format
                    set_range_size_1
                    set_status
                    set_to_output_map
                    short_name
                    status
                    status_info
                    to_output_map
                    type_of
                    value_of
                    write
                ))
                    # 'property' above is for symmetry, so that one can take
                    # the property of a property and get itself, and so don't
                    # have to distinguish between properties and tables in
                    # calling code
    {
        no strict "refs";
        *$sub = sub {
            use strict "refs";
            my $self = shift;
            return $map{pack 'J', refaddr $self}->$sub(@_);
        }
    }


} # End closure

package main;

sub display_chr {
    # Converts an ordinal printable character value to a displayable string,
    # using a dotted circle to hold combining characters.

    my $ord = shift;
    my $chr = chr $ord;
    return $chr if $ccc->table(0)->contains($ord);
    return "\x{25CC}$chr";
}

sub join_lines($input) {
    # Returns lines of the input joined together, so that they can be folded
    # properly.
    # This causes continuation lines to be joined together into one long line
    # for folding.  A continuation line is any line that doesn't begin with a
    # space or "\b" (the latter is stripped from the output).  This is so
    # lines can be in a HERE document so as to fit nicely in the terminal
    # width, but be joined together in one long line, and then folded with
    # indents, '#' prefixes, etc, properly handled.
    # A blank separates the joined lines except if there is a break; an extra
    # blank is inserted after a period ending a line.

    # Initialize the return with the first line.
    my ($return, @lines) = split "\n", $input;

    # If the first line is null, it was an empty line, add the \n back in
    $return = "\n" if $return eq "";

    # Now join the remainder of the physical lines.
    for my $line (@lines) {

        # An empty line means wanted a blank line, so add two \n's to get that
        # effect, and go to the next line.
        if (length $line == 0) {
            $return .= "\n\n";
            next;
        }

        # Look at the last character of what we have so far.
        my $previous_char = substr($return, -1, 1);

        # And at the next char to be output.
        my $next_char = substr($line, 0, 1);

        if ($previous_char ne "\n") {

            # Here didn't end wth a nl.  If the next char a blank or \b, it
            # means that here there is a break anyway.  So add a nl to the
            # output.
            if ($next_char eq " " || $next_char eq "\b") {
                $previous_char = "\n";
                $return .= $previous_char;
            }

            # Add an extra space after periods.
            $return .= " " if $previous_char eq '.';
        }

        # Here $previous_char is still the latest character to be output.  If
        # it isn't a nl, it means that the next line is to be a continuation
        # line, with a blank inserted between them.
        $return .= " " if $previous_char ne "\n";

        # Get rid of any \b
        substr($line, 0, 1) = "" if $next_char eq "\b";

        # And append this next line.
        $return .= $line;
    }

    return $return;
}

sub simple_fold( $line, $prefix="", $hanging_indent=0, $right_margin=0) {
    # Returns a string of the input (string or an array of strings) folded
    # into multiple-lines each of no more than $MAX_LINE_WIDTH characters plus
    # a \n
    # This is tailored for the kind of text written by this program,
    # especially the pod file, which can have very long names with
    # underscores in the middle, or words like AbcDefgHij....  We allow
    # breaking in the middle of such constructs if the line won't fit
    # otherwise.  The break in such cases will come either just after an
    # underscore, or just before one of the Capital letters.

    local $to_trace = 0 if main::DEBUG;

    # $prefix Optional string to prepend to each output line
    # $hanging_indent Optional number of spaces to indent
    # continuation lines
    # $right_margin  Optional number of spaces to narrow the
    # total width by.

    # The space available doesn't include what's automatically prepended
    # to each line, or what's reserved on the right.
    my $max = $MAX_LINE_WIDTH - length($prefix) - $right_margin;
    # XXX Instead of using the 'nofold' perhaps better to look up the stack

    if (DEBUG && $hanging_indent >= $max) {
        Carp::my_carp("Too large a hanging indent ($hanging_indent); must be < $max.  Using 0", 'nofold');
        $hanging_indent = 0;
    }

    # First, split into the current physical lines.
    my @line;
    if (ref $line) {        # Better be an array, because not bothering to
                            # test
        foreach my $line (@{$line}) {
            push @line, split /\n/, $line;
        }
    }
    else {
        @line = split /\n/, $line;
    }

    #local $to_trace = 1 if main::DEBUG;
    trace "", join(" ", @line), "\n" if main::DEBUG && $to_trace;

    # Look at each current physical line.
    for (my $i = 0; $i < @line; $i++) {
        Carp::my_carp("Tabs don't work well.", 'nofold') if $line[$i] =~ /\t/;
        #local $to_trace = 1 if main::DEBUG;
        trace "i=$i: $line[$i]\n" if main::DEBUG && $to_trace;

        # Remove prefix, because will be added back anyway, don't want
        # doubled prefix
        $line[$i] =~ s/^$prefix//;

        # Remove trailing space
        $line[$i] =~ s/\s+\Z//;

        # If the line is too long, fold it.
        if (length $line[$i] > $max) {
            my $remainder;

            # Here needs to fold.  Save the leading space in the line for
            # later.
            $line[$i] =~ /^ ( \s* )/x;
            my $leading_space = $1;
            trace "line length", length $line[$i], "; lead length", length($leading_space) if main::DEBUG && $to_trace;

            # If character at final permissible position is white space,
            # fold there, which will delete that white space
            if (substr($line[$i], $max - 1, 1) =~ /\s/) {
                $remainder = substr($line[$i], $max);
                $line[$i] = substr($line[$i], 0, $max - 1);
            }
            else {

                # Otherwise fold at an acceptable break char closest to
                # the max length.  Look at just the maximal initial
                # segment of the line
                my $segment = substr($line[$i], 0, $max - 1);
                if ($segment =~
                    /^ ( .{$hanging_indent}   # Don't look before the
                                              #  indent.
                        \ *                   # Don't look in leading
                                              #  blanks past the indent
                            [^ ] .*           # Find the right-most
                        (?:                   #  acceptable break:
                            [ \s = ]          # space or equal
                            | - (?! [.0-9] )  # or non-unary minus.
                            | [^\\[(] (?= \\ )# break before single backslash
                                              #  not immediately after opening
                                              #  punctuation
                        )                     # $1 includes the character
                    )/x)
                {
                    # Split into the initial part that fits, and remaining
                    # part of the input
                    $remainder = substr($line[$i], length $1);
                    $line[$i] = $1;
                    trace $line[$i] if DEBUG && $to_trace;
                    trace $remainder if DEBUG && $to_trace;
                }

                # If didn't find a good breaking spot, see if there is a
                # not-so-good breaking spot.  These are just after
                # underscores or where the case changes from lower to
                # upper.  Use \a as a soft hyphen, but give up
                # and don't break the line if there is actually a \a
                # already in the input.  We use an ascii character for the
                # soft-hyphen to avoid any attempt by miniperl to try to
                # access the files that this program is creating.
                elsif ($segment !~ /\a/
                       && ($segment =~ s/_/_\a/g
                       || $segment =~ s/ ( (?!\\) [a-z] ) (?= [A-Z] )/$1\a/xg))
                {
                    # Here were able to find at least one place to insert
                    # our substitute soft hyphen.  Find the right-most one
                    # and replace it by a real hyphen.
                    trace $segment if DEBUG && $to_trace;
                    substr($segment,
                            rindex($segment, "\a"),
                            1) = '-';

                    # Then remove the soft hyphen substitutes.
                    $segment =~ s/\a//g;
                    trace $segment if DEBUG && $to_trace;

                    # And split into the initial part that fits, and
                    # remainder of the line
                    my $pos = rindex($segment, '-');
                    $remainder = substr($line[$i], $pos);
                    trace $remainder if DEBUG && $to_trace;
                    $line[$i] = substr($segment, 0, $pos + 1);
                }
            }

            # Here we know if we can fold or not.  If we can, $remainder
            # is what remains to be processed in the next iteration.
            if (defined $remainder) {
                trace "folded='$line[$i]'" if main::DEBUG && $to_trace;

                # Insert the folded remainder of the line as a new element
                # of the array.  (It may still be too long, but we will
                # deal with that next time through the loop.)  Omit any
                # leading space in the remainder.
                $remainder =~ s/^\s+//;
                trace "remainder='$remainder'" if main::DEBUG && $to_trace;

                # But then indent by whichever is larger of:
                # 1) the leading space on the input line;
                # 2) the hanging indent.
                # This preserves indentation in the original line.
                my $lead = ($leading_space)
                            ? length $leading_space
                            : $hanging_indent;
                $lead = max($lead, $hanging_indent);
                splice @line, $i+1, 0, (" " x $lead) . $remainder;
            }
        }

        # Ready to output the line. Get rid of any trailing space
        # And prefix by the required $prefix passed in.
        $line[$i] =~ s/\s+$//;
        $line[$i] = "$prefix$line[$i]\n";
    } # End of looping through all the lines.

    return join "", @line;
}

sub property_ref {  # Returns a reference to a property object.
    return Property::property_ref(@_);
}

sub force_unlink ($filename) {
    return unless file_exists($filename);
    return if CORE::unlink($filename);

    # We might need write permission
    chmod 0777, $filename;
    CORE::unlink($filename) or Carp::my_carp("Couldn't unlink $filename.  Proceeding anyway: $!");
    return;
}

sub write ($file, $use_utf8, @lines) {
    # Given a filename and references to arrays of lines, write the lines of
    # each array to the file
    # Filename can be given as an arrayref of directory names

    # Get into a single string if an array, and get rid of, in Unix terms, any
    # leading '.'
    $file= File::Spec->join(@$file) if ref $file eq 'ARRAY';
    $file = File::Spec->canonpath($file);

    # If has directories, make sure that they all exist
    (undef, my $directories, undef) = File::Spec->splitpath($file);
    File::Path::mkpath($directories) if $directories && ! -d $directories;

    push @files_actually_output, $file;

    force_unlink ($file);

    my $OUT;
    if (not open $OUT, ">", $file) {
        Carp::my_carp("can't open $file for output.  Skipping this file: $!");
        return;
    }

    binmode $OUT, ":utf8" if $use_utf8;

    foreach my $lines_ref (@lines) {
        unless (@$lines_ref) {
            Carp::my_carp("An array of lines for writing to file '$file' is empty; writing it anyway;");
        }

        print $OUT @$lines_ref or die Carp::my_carp("write to '$file' failed: $!");
    }
    close $OUT or die Carp::my_carp("close '$file' failed: $!");

    print "$file written.\n" if $verbosity >= $VERBOSE;

    return;
}


sub Standardize($name=undef) {
    # This converts the input name string into a standardized equivalent to
    # use internally.

    unless (defined $name) {
      Carp::my_carp_bug("Standardize() called with undef.  Returning undef.");
      return;
    }

    # Remove any leading or trailing white space
    $name =~ s/^\s+//g;
    $name =~ s/\s+$//g;

    # Convert interior white space and hyphens into underscores.
    $name =~ s/ (?<= .) [ -]+ (.) /_$1/xg;

    # Capitalize the letter following an underscore, and convert a sequence of
    # multiple underscores to a single one
    $name =~ s/ (?<= .) _+ (.) /_\u$1/xg;

    # And capitalize the first letter, but not for the special cjk ones.
    $name = ucfirst($name) unless $name =~ /^k[A-Z]/;
    return $name;
}

sub standardize ($str=undef) {
    # Returns a lower-cased standardized name, without underscores.  This form
    # is chosen so that it can distinguish between any real versus superficial
    # Unicode name differences.  It relies on the fact that Unicode doesn't
    # have interior underscores, white space, nor dashes in any
    # stricter-matched name.  It should not be used on Unicode code point
    # names (the Name property), as they mostly, but not always follow these
    # rules.

    my $name = Standardize($str);
    return if !defined $name;

    $name =~ s/ (?<= .) _ (?= . ) //xg;
    return lc $name;
}

sub UCD_name ($table, $alias) {
    # Returns the name that Unicode::UCD will use to find a table.  XXX
    # perhaps this function should be placed somewhere, like UCD.pm so that
    # Unicode::UCD can use it directly without duplicating code that can get
    # out-of sync.

    my $property = $table->property;
    $property = ($property == $perl)
                ? ""                # 'perl' is never explicitly stated
                : standardize($property->name) . '=';
    if ($alias->loose_match) {
        return $property . standardize($alias->name);
    }
    else {
        return lc ($property . $alias->name);
    }

    return;
}

{   # Closure

    my $indent_increment = " " x ( $debugging_build ? 2 : 0);
    %main::already_output = ();

    $main::simple_dumper_nesting = 0;

    sub simple_dumper( $item, $indent = "" ) {
        # Like Simple Data::Dumper. Good enough for our needs. We can't use
        # the real thing as we have to run under miniperl.

        # It is designed so that on input it is at the beginning of a line,
        # and the final thing output in any call is a trailing ",\n".

        $indent = "" if ! $debugging_build;

        # nesting level is localized, so that as the call stack pops, it goes
        # back to the prior value.
        local $main::simple_dumper_nesting = $main::simple_dumper_nesting;
        local %main::already_output = %main::already_output;
        $main::simple_dumper_nesting++;
        #print STDERR __LINE__, ": $main::simple_dumper_nesting: $indent$item\n";

        # Determine the indent for recursive calls.
        my $next_indent = $indent . $indent_increment;

        my $output;
        if (! ref $item) {

            # Dump of scalar: just output it in quotes if not a number.  To do
            # so we must escape certain characters, and therefore need to
            # operate on a copy to avoid changing the original
            my $copy = $item;
            $copy = $UNDEF unless defined $copy;

            # Quote non-integers (integers also have optional leading '-')
            if ($copy eq "" || $copy !~ /^ -? \d+ $/x) {

                # Escape apostrophe and backslash
                $copy =~ s/ ( ['\\] ) /\\$1/xg;
                $copy = "'$copy'";
            }
            $output = "$indent$copy,\n";
        }
        else {

            # Keep track of cycles in the input, and refuse to infinitely loop
            my $addr = pack 'J', refaddr $item;
            if (defined $main::already_output{$addr}) {
                return "${indent}ALREADY OUTPUT: $item\n";
            }
            $main::already_output{$addr} = $item;

            if (ref $item eq 'ARRAY') {
                my $using_brackets;
                $output = $indent;
                if ($main::simple_dumper_nesting > 1) {
                    $output .= '[';
                    $using_brackets = 1;
                }
                else {
                    $using_brackets = 0;
                }

                # If the array is empty, put the closing bracket on the same
                # line.  Otherwise, recursively add each array element
                if (@$item == 0) {
                    $output .= " ";
                }
                else {
                    $output .= "\n";
                    for (my $i = 0; $i < @$item; $i++) {

                        # Indent array elements one level
                        $output .= &simple_dumper($item->[$i], $next_indent);
                        next if ! $debugging_build;
                        $output =~ s/\n$//;      # Remove any trailing nl so
                        $output .= " # [$i]\n";  # as to add a comment giving
                                                 # the array index
                    }
                    $output .= $indent;     # Indent closing ']' to orig level
                }
                $output .= ']' if $using_brackets;
                $output .= ",\n";
            }
            elsif (ref $item eq 'HASH') {
                my $is_first_line;
                my $using_braces;
                my $body_indent;

                # No surrounding braces at top level
                $output .= $indent;
                if ($main::simple_dumper_nesting > 1) {
                    $output .= "{\n";
                    $is_first_line = 0;
                    $body_indent = $next_indent;
                    $next_indent .= $indent_increment;
                    $using_braces = 1;
                }
                else {
                    $is_first_line = 1;
                    $body_indent = $indent;
                    $using_braces = 0;
                }

                # Output hashes sorted alphabetically instead of apparently
                # random.  Use caseless alphabetic sort
                foreach my $key (sort { lc $a cmp lc $b } keys %$item)
                {
                    if ($is_first_line) {
                        $is_first_line = 0;
                    }
                    else {
                        $output .= "$body_indent";
                    }

                    # The key must be a scalar, but this recursive call quotes
                    # it
                    $output .= &simple_dumper($key);

                    # And change the trailing comma and nl to the hash fat
                    # comma for clarity, and so the value can be on the same
                    # line
                    $output =~ s/,\n$/ => /;

                    # Recursively call to get the value's dump.
                    my $next = &simple_dumper($item->{$key}, $next_indent);

                    # If the value is all on one line, remove its indent, so
                    # will follow the => immediately.  If it takes more than
                    # one line, start it on a new line.
                    if ($next !~ /\n.*\n/) {
                        $next =~ s/^ *//;
                    }
                    else {
                        $output .= "\n";
                    }
                    $output .= $next;
                }

                $output .= "$indent},\n" if $using_braces;
            }
            elsif (ref $item eq 'CODE' || ref $item eq 'GLOB') {
                $output = $indent . ref($item) . "\n";
                # XXX see if blessed
            }
            elsif ($item->can('dump')) {

                # By convention in this program, objects furnish a 'dump'
                # method.  Since not doing any output at this level, just pass
                # on the input indent
                $output = $item->dump($indent);
            }
            else {
                Carp::my_carp("Can't cope with dumping a " . ref($item) . ".  Skipping.");
            }
        }
        return $output;
    }
}

sub dump_inside_out( $object, $fields_ref, @args ) {
    # Dump inside-out hashes in an object's state by converting them to a
    # regular hash and then calling simple_dumper on that.

    my $addr = pack 'J', refaddr $object;

    my %hash;
    foreach my $key (keys %$fields_ref) {
        $hash{$key} = $fields_ref->{$key}{$addr};
    }

    return simple_dumper(\%hash, @args);
}

sub _operator_dot($self, $other="", $reversed=0) {
    # Overloaded '.' method that is common to all packages.  It uses the
    # package's stringify method.

    foreach my $which (\$self, \$other) {
        next unless ref $$which;
        if ($$which->can('_operator_stringify')) {
            $$which = $$which->_operator_stringify;
        }
        else {
            my $ref = ref $$which;
            my $addr = pack 'J', refaddr $$which;
            $$which = "$ref ($addr)";
        }
    }
    return ($reversed)
            ? "$other$self"
            : "$self$other";
}

sub _operator_dot_equal($self, $other="", $reversed=0) {
    # Overloaded '.=' method that is common to all packages.

    if ($reversed) {
        return $other .= "$self";
    }
    else {
        return "$self" . "$other";
    }
}

sub _operator_equal($self, $other, @) {
    # Generic overloaded '==' routine.  To be equal, they must be the exact
    # same object

    return 0 unless defined $other;
    return 0 unless ref $other;
    no overloading;
    return $self == $other;
}

sub _operator_not_equal($self, $other, @) {
    return ! _operator_equal($self, $other);
}

sub substitute_PropertyAliases($file_object) {
    # Deal with early releases that don't have the crucial PropertyAliases.txt
    # file.

    $file_object->insert_lines(get_old_property_aliases());

    process_PropertyAliases($file_object);
}


sub process_PropertyAliases($file) {
    # This reads in the PropertyAliases.txt file, which contains almost all
    # the character properties in Unicode and their equivalent aliases:
    # scf       ; Simple_Case_Folding         ; sfc
    #
    # Field 0 is the preferred short name for the property.
    # Field 1 is the full name.
    # Any succeeding ones are other accepted names.

    # Add any cjk properties that may have been defined.
    $file->insert_lines(@cjk_properties);

    while ($file->next_line) {

        my @data = split /\s*;\s*/;

        my $full = $data[1];

        # This line is defective in early Perls.  The property in Unihan.txt
        # is kRSUnicode.
        if ($full eq 'Unicode_Radical_Stroke' && @data < 3) {
            push @data, qw(cjkRSUnicode kRSUnicode);
        }

        my $this = Property->new($data[0], Full_Name => $full);

        $this->set_fate($SUPPRESSED, $why_suppressed{$full})
                                                    if $why_suppressed{$full};

        # Start looking for more aliases after these two.
        for my $i (2 .. @data - 1) {
            $this->add_alias($data[$i]);
        }

    }

    my $scf = property_ref("Simple_Case_Folding");
    $scf->add_alias("scf");
    $scf->add_alias("sfc");

    return;
}

sub finish_property_setup($file) {
    # Finishes setting up after PropertyAliases.

    # This entry was missing from this file in earlier Unicode versions
    if (-e 'Jamo.txt' && ! defined property_ref('JSN')) {
        Property->new('JSN', Full_Name => 'Jamo_Short_Name');
    }

    # These are used so much, that we set globals for them.
    $gc = property_ref('General_Category');
    $block = property_ref('Block');
    $script = property_ref('Script');
    $age = property_ref('Age');

    # Perl adds this alias.
    $gc->add_alias('Category');

    # Unicode::Normalize expects this file with this name and directory.
    $ccc = property_ref('Canonical_Combining_Class');
    if (defined $ccc) {
        $ccc->set_file('CombiningClass');
        $ccc->set_directory(File::Spec->curdir());
    }

    # These two properties aren't actually used in the core, but unfortunately
    # the names just above that are in the core interfere with these, so
    # choose different names.  These aren't a problem unless the map tables
    # for these files get written out.
    my $lowercase = property_ref('Lowercase');
    $lowercase->set_file('IsLower') if defined $lowercase;
    my $uppercase = property_ref('Uppercase');
    $uppercase->set_file('IsUpper') if defined $uppercase;

    # Set up the hard-coded default mappings, but only on properties defined
    # for this release
    foreach my $property (keys %default_mapping) {
        my $property_object = property_ref($property);
        next if ! defined $property_object;
        my $default_map = $default_mapping{$property};
        $property_object->set_default_map($default_map);

        # A map of <code point> implies the property is string.
        if ($property_object->type == $UNKNOWN
            && $default_map eq $CODE_POINT)
        {
            $property_object->set_type($STRING);
        }
    }

    # For backwards compatibility with applications that may read the mapping
    # file directly (it was documented in 5.12 and 5.14 as being thusly
    # usable), keep it from being adjusted.  (range_size_1 is
    # used to force the traditional format.)
    foreach my $property (qw(NFKC_Casefold NFKC_Simple_Casefold)) {
        if (defined (my $cf = property_ref($property))) {
            $cf->set_to_output_map($EXTERNAL_MAP);
            $cf->set_range_size_1(1);
        }
    }

    if (defined (my $bmg = property_ref('Bidi_Mirroring_Glyph'))) {
        $bmg->set_to_output_map($EXTERNAL_MAP);
        $bmg->set_range_size_1(1);
    }

    property_ref('Numeric_Value')->set_to_output_map($OUTPUT_ADJUSTED);

    # These two properties have no short names and the file names for them
    # clash in DOS 8.3.  Work around this by creating shorter file names that
    # work
    my $IDCMStart = property_ref("ID_Compat_Math_Start");
    $IDCMStart->set_match_subdir("IDCMStart") if defined $IDCMStart;
    my $IDCMCont= property_ref("ID_Compat_Math_Continue");
    $IDCMCont->set_match_subdir("IDCMContinue") if defined $IDCMCont;

    # The rest of this sub is for properties that need the Multi_Default class
    # to create objects for defaults.  As of v15.0, this is no longer needed.

    return if $v_version ge v15.0.0;

    # Bidi class has a complicated default, but the derived file takes care of
    # the complications, leaving just 'L'.
    if (file_exists("${EXTRACTED}DBidiClass.txt")) {
        property_ref('Bidi_Class')->set_default_map('L');
    }
    else {
        my $default;

        # The derived file was introduced in 3.1.1.  The values below are
        # taken from table 3-8, TUS 3.0
        my $default_R =
            'my $default = Range_List->new;
             $default->add_range(0x0590, 0x05FF);
             $default->add_range(0xFB1D, 0xFB4F);'
        ;

        # The defaults apply only to unassigned characters
        $default_R .= '$gc->table("Unassigned") & $default;';

        if ($v_version lt v3.0.0) {
            $default = Multi_Default->new(R => $default_R, 'L');
        }
        else {

            # AL apparently not introduced until 3.0:  TUS 2.x references are
            # not on-line to check it out
            my $default_AL =
                'my $default = Range_List->new;
                 $default->add_range(0x0600, 0x07BF);
                 $default->add_range(0xFB50, 0xFDFF);
                 $default->add_range(0xFE70, 0xFEFF);'
            ;

            # Non-character code points introduced in this release; aren't AL
            if ($v_version ge 3.1.0) {
                $default_AL .= '$default->delete_range(0xFDD0, 0xFDEF);';
            }
            $default_AL .= '$gc->table("Unassigned") & $default';
            $default = Multi_Default->new(AL => $default_AL,
                                          R => $default_R,
                                          'L');
        }
        property_ref('Bidi_Class')->set_default_map($default);
    }

    # Joining type has a complicated default, but the derived file takes care
    # of the complications, leaving just 'U' (or Non_Joining), except the file
    # is bad in 3.1.0
    if (file_exists("${EXTRACTED}DJoinType.txt") || -e 'ArabicShaping.txt') {
        if (file_exists("${EXTRACTED}DJoinType.txt") && $v_version ne 3.1.0) {
            property_ref('Joining_Type')->set_default_map('Non_Joining');
        }
        else {

            # Otherwise, there are not one, but two possibilities for the
            # missing defaults: T and U.
            # The missing defaults that evaluate to T are given by:
            # T = Mn + Cf - ZWNJ - ZWJ
            # where Mn and Cf are the general category values. In other words,
            # any non-spacing mark or any format control character, except
            # U+200C ZERO WIDTH NON-JOINER (joining type U) and U+200D ZERO
            # WIDTH JOINER (joining type C).
            my $default = Multi_Default->new(
               'T' => '$gc->table("Mn") + $gc->table("Cf") - 0x200C - 0x200D',
               'Non_Joining');
            property_ref('Joining_Type')->set_default_map($default);
        }
    }

    # Line break has a complicated default in early releases. It is 'Unknown'
    # for non-assigned code points; 'AL' for assigned.
    if (file_exists("${EXTRACTED}DLineBreak.txt") || -e 'LineBreak.txt') {
        my $lb = property_ref('Line_Break');
        if (file_exists("${EXTRACTED}DLineBreak.txt")) {
            $lb->set_default_map('Unknown');
        }
        else {
            my $default = Multi_Default->new('AL' => '~ $gc->table("Cn")',
                                             'Unknown',
                                            );
            $lb->set_default_map($default);
        }
    }

    return;
}

sub get_old_property_aliases() {
    # Returns what would be in PropertyAliases.txt if it existed in very old
    # versions of Unicode.  It was derived from the one in 3.2, and pared
    # down based on the data that was actually in the older releases.
    # An attempt was made to use the existence of files to mean inclusion or
    # not of various aliases, but if this was not sufficient, using version
    # numbers was resorted to.

    my @return;

    # These are to be used in all versions (though some are constructed by
    # this program if missing)
    push @return, split /\n/, <<'END';
bc        ; Bidi_Class
Bidi_M    ; Bidi_Mirrored
cf        ; Case_Folding
ccc       ; Canonical_Combining_Class
dm        ; Decomposition_Mapping
dt        ; Decomposition_Type
gc        ; General_Category
isc       ; ISO_Comment
lc        ; Lowercase_Mapping
na        ; Name
na1       ; Unicode_1_Name
nt        ; Numeric_Type
nv        ; Numeric_Value
scf       ; Simple_Case_Folding
slc       ; Simple_Lowercase_Mapping
stc       ; Simple_Titlecase_Mapping
suc       ; Simple_Uppercase_Mapping
tc        ; Titlecase_Mapping
uc        ; Uppercase_Mapping
END

    if (-e 'Blocks.txt') {
        push @return, "blk       ; Block\n";
    }
    if (-e 'ArabicShaping.txt') {
        push @return, split /\n/, <<'END';
jg        ; Joining_Group
jt        ; Joining_Type
END
    }
    if (-e 'PropList.txt') {

        # This first set is in the original old-style proplist.
        push @return, split /\n/, <<'END';
Bidi_C    ; Bidi_Control
Dash      ; Dash
Dia       ; Diacritic
Ext       ; Extender
Hex       ; Hex_Digit
Hyphen    ; Hyphen
IDC       ; ID_Continue
Ideo      ; Ideographic
Join_C    ; Join_Control
Math      ; Math
QMark     ; Quotation_Mark
Term      ; Terminal_Punctuation
WSpace    ; White_Space
END
        # The next sets were added later
        if ($v_version ge v3.0.0) {
            push @return, split /\n/, <<'END';
Upper     ; Uppercase
Lower     ; Lowercase
END
        }
        if ($v_version ge v3.0.1) {
            push @return, split /\n/, <<'END';
NChar     ; Noncharacter_Code_Point
END
        }
        # The next sets were added in the new-style
        if ($v_version ge v3.1.0) {
            push @return, split /\n/, <<'END';
OAlpha    ; Other_Alphabetic
OLower    ; Other_Lowercase
OMath     ; Other_Math
OUpper    ; Other_Uppercase
END
        }
        if ($v_version ge v3.1.1) {
            push @return, "AHex      ; ASCII_Hex_Digit\n";
        }
    }
    if (-e 'EastAsianWidth.txt') {
        push @return, "ea        ; East_Asian_Width\n";
    }
    if (-e 'CompositionExclusions.txt') {
        push @return, "CE        ; Composition_Exclusion\n";
    }
    if (-e 'LineBreak.txt') {
        push @return, "lb        ; Line_Break\n";
    }
    if (-e 'BidiMirroring.txt') {
        push @return, "bmg       ; Bidi_Mirroring_Glyph\n";
    }
    if (-e 'Scripts.txt') {
        push @return, "sc        ; Script\n";
    }
    if (-e 'DNormalizationProps.txt') {
        push @return, split /\n/, <<'END';
Comp_Ex   ; Full_Composition_Exclusion
FC_NFKC   ; FC_NFKC_Closure
NFC_QC    ; NFC_Quick_Check
NFD_QC    ; NFD_Quick_Check
NFKC_QC   ; NFKC_Quick_Check
NFKD_QC   ; NFKD_Quick_Check
XO_NFC    ; Expands_On_NFC
XO_NFD    ; Expands_On_NFD
XO_NFKC   ; Expands_On_NFKC
XO_NFKD   ; Expands_On_NFKD
END
    }
    if (-e 'DCoreProperties.txt') {
        push @return, split /\n/, <<'END';
Alpha     ; Alphabetic
IDS       ; ID_Start
XIDC      ; XID_Continue
XIDS      ; XID_Start
END
        # These can also appear in some versions of PropList.txt
        push @return, "Lower     ; Lowercase\n"
                                    unless grep { $_ =~ /^Lower\b/} @return;
        push @return, "Upper     ; Uppercase\n"
                                    unless grep { $_ =~ /^Upper\b/} @return;
    }

    # This flag requires the DAge.txt file to be copied into the directory.
    if (DEBUG && $compare_versions) {
        push @return, 'age       ; Age';
    }

    return @return;
}

sub substitute_PropValueAliases($file_object) {
    # Deal with early releases that don't have the crucial
    # PropValueAliases.txt file.

    $file_object->insert_lines(get_old_property_value_aliases());

    process_PropValueAliases($file_object);
}

sub process_PropValueAliases($file) {
    # This file contains values that properties look like:
    # bc ; AL        ; Arabic_Letter
    # blk; n/a       ; Greek_And_Coptic                 ; Greek
    #
    # Field 0 is the property.
    # Field 1 is the short name of a property value or 'n/a' if no
    #                short name exists;
    # Field 2 is the full property value name;
    # Any other fields are more synonyms for the property value.
    # Purely numeric property values are omitted from the file; as are some
    # others, fewer and fewer in later releases

    # Entries for the ccc property have an extra field before the
    # abbreviation:
    # ccc;   0; NR   ; Not_Reordered
    # It is the numeric value that the names are synonyms for.

    # There are comment entries for values missing from this file:
    # # @missing: 0000..10FFFF; ISO_Comment; <none>
    # # @missing: 0000..10FFFF; Lowercase_Mapping; <code point>

    if ($v_version lt 4.0.0) {
        $file->insert_lines(split /\n/, <<'END'
Hangul_Syllable_Type; L                                ; Leading_Jamo
Hangul_Syllable_Type; LV                               ; LV_Syllable
Hangul_Syllable_Type; LVT                              ; LVT_Syllable
Hangul_Syllable_Type; NA                               ; Not_Applicable
Hangul_Syllable_Type; T                                ; Trailing_Jamo
Hangul_Syllable_Type; V                                ; Vowel_Jamo
END
        );
    }
    if ($v_version lt 4.1.0) {
        $file->insert_lines(split /\n/, <<'END'
_Perl_GCB; CN                               ; Control
_Perl_GCB; CR                               ; CR
_Perl_GCB; EX                               ; Extend
_Perl_GCB; L                                ; L
_Perl_GCB; LF                               ; LF
_Perl_GCB; LV                               ; LV
_Perl_GCB; LVT                              ; LVT
_Perl_GCB; T                                ; T
_Perl_GCB; V                                ; V
_Perl_GCB; XX                               ; Other
END
        );
    }

    # Add any explicit cjk values
    $file->insert_lines(@cjk_property_values);

    # This line is used only for testing the code that checks for name
    # conflicts.  There is a script Inherited, and when this line is executed
    # it causes there to be a name conflict with the 'Inherited' that this
    # program generates for this block property value
    #$file->insert_lines('blk; n/a; Herited');

    # Process each line of the file ...
    while ($file->next_line) {

        # Fix typo in input file
        s/CCC133/CCC132/g if $v_version eq v6.1.0;

        my ($property, @data) = split /\s*;\s*/;

        # The ccc property has an extra field at the beginning, which is the
        # numeric value.  Move it to be after the other two, mnemonic, fields,
        # so that those will be used as the property value's names, and the
        # number will be an extra alias.  (Rightmost splice removes field 1-2,
        # returning them in a slice; left splice inserts that before anything,
        # thus shifting the former field 0 to after them.)
        splice (@data, 0, 0, splice(@data, 1, 2)) if $property eq 'ccc';

        if ($v_version le v5.0.0 && $property eq 'blk' && $data[1] =~ /-/) {
            my $new_style = $data[1] =~ s/-/_/gr;
            splice @data, 1, 0, $new_style;
        }

        # Field 0 is a short name unless "n/a"; field 1 is the full name.  If
        # there is no short name, use the full one in element 1
        if ($data[0] eq "n/a") {
            $data[0] = $data[1];
        }
        elsif ($data[0] ne $data[1]
               && standardize($data[0]) eq standardize($data[1])
               && $data[1] !~ /[[:upper:]]/)
        {
            # Also, there is a bug in the file in which "n/a" is omitted, and
            # the two fields are identical except for case, and the full name
            # is all lower case.  Copy the "short" name unto the full one to
            # give it some upper case.

            $data[1] = $data[0];
        }

        # Earlier releases had the pseudo property 'qc' that should expand to
        # the ones that replace it below.
        if ($property eq 'qc') {
            if (lc $data[0] eq 'y') {
                $file->insert_lines('NFC_QC; Y      ; Yes',
                                    'NFD_QC; Y      ; Yes',
                                    'NFKC_QC; Y     ; Yes',
                                    'NFKD_QC; Y     ; Yes',
                                    );
            }
            elsif (lc $data[0] eq 'n') {
                $file->insert_lines('NFC_QC; N      ; No',
                                    'NFD_QC; N      ; No',
                                    'NFKC_QC; N     ; No',
                                    'NFKD_QC; N     ; No',
                                    );
            }
            elsif (lc $data[0] eq 'm') {
                $file->insert_lines('NFC_QC; M      ; Maybe',
                                    'NFKC_QC; M     ; Maybe',
                                    );
            }
            else {
                $file->carp_bad_line("qc followed by unexpected '$data[0]");
            }
            next;
        }

        # The first field is the short name, 2nd is the full one.
        my $property_object = property_ref($property);
        my $table = $property_object->add_match_table($data[0],
                                                Full_Name => $data[1]);

        # Start looking for more aliases after these two.
        for my $i (2 .. @data - 1) {
            $table->add_alias($data[$i]);
        }
    } # End of looping through the file

    # As noted in the comments early in the program, it generates tables for
    # the default values for all releases, even those for which the concept
    # didn't exist at the time.  Here we add those if missing.
    if (defined $age && ! defined $age->table('Unassigned')) {
        $age->add_match_table('Unassigned');
    }
    $block->add_match_table('No_Block') if -e 'Blocks.txt'
                                    && ! defined $block->table('No_Block');


    # Now set the default mappings of the properties from the file.  This is
    # done after the loop because a number of properties have only @missings
    # entries in the file, and may not show up until the end.
    my @defaults = $file->get_missings;
    foreach my $default_ref (@defaults) {
        my $default = $default_ref->{default};
        my $property = property_ref($default_ref->{property});
        $property->set_default_map($default);
    }

    return;
}

sub get_old_property_value_aliases () {
    # Returns what would be in PropValueAliases.txt if it existed in very old
    # versions of Unicode.  It was derived from the one in 3.2, and pared
    # down.  An attempt was made to use the existence of files to mean
    # inclusion or not of various aliases, but if this was not sufficient,
    # using version numbers was resorted to.

    my @return = split /\n/, <<'END';
bc ; AN        ; Arabic_Number
bc ; B         ; Paragraph_Separator
bc ; CS        ; Common_Separator
bc ; EN        ; European_Number
bc ; ES        ; European_Separator
bc ; ET        ; European_Terminator
bc ; L         ; Left_To_Right
bc ; ON        ; Other_Neutral
bc ; R         ; Right_To_Left
bc ; WS        ; White_Space

Bidi_M; N; No; F; False
Bidi_M; Y; Yes; T; True

# The standard combining classes are very much different in v1, so only use
# ones that look right (not checked thoroughly)
ccc;   0; NR   ; Not_Reordered
ccc;   1; OV   ; Overlay
ccc;   7; NK   ; Nukta
ccc;   8; KV   ; Kana_Voicing
ccc;   9; VR   ; Virama
ccc; 202; ATBL ; Attached_Below_Left
ccc; 216; ATAR ; Attached_Above_Right
ccc; 218; BL   ; Below_Left
ccc; 220; B    ; Below
ccc; 222; BR   ; Below_Right
ccc; 224; L    ; Left
ccc; 228; AL   ; Above_Left
ccc; 230; A    ; Above
ccc; 232; AR   ; Above_Right
ccc; 234; DA   ; Double_Above

dt ; can       ; canonical
dt ; enc       ; circle
dt ; fin       ; final
dt ; font      ; font
dt ; fra       ; fraction
dt ; init      ; initial
dt ; iso       ; isolated
dt ; med       ; medial
dt ; n/a       ; none
dt ; nb        ; noBreak
dt ; sqr       ; square
dt ; sub       ; sub
dt ; sup       ; super

gc ; C         ; Other                            # Cc | Cf | Cn | Co | Cs
gc ; Cc        ; Control
gc ; Cn        ; Unassigned
gc ; Co        ; Private_Use
gc ; L         ; Letter                           # Ll | Lm | Lo | Lt | Lu
gc ; LC        ; Cased_Letter                     # Ll | Lt | Lu
gc ; Ll        ; Lowercase_Letter
gc ; Lm        ; Modifier_Letter
gc ; Lo        ; Other_Letter
gc ; Lu        ; Uppercase_Letter
gc ; M         ; Mark                             # Mc | Me | Mn
gc ; Mc        ; Spacing_Mark
gc ; Mn        ; Nonspacing_Mark
gc ; N         ; Number                           # Nd | Nl | No
gc ; Nd        ; Decimal_Number
gc ; No        ; Other_Number
gc ; P         ; Punctuation                      # Pc | Pd | Pe | Pf | Pi | Po | Ps
gc ; Pd        ; Dash_Punctuation
gc ; Pe        ; Close_Punctuation
gc ; Po        ; Other_Punctuation
gc ; Ps        ; Open_Punctuation
gc ; S         ; Symbol                           # Sc | Sk | Sm | So
gc ; Sc        ; Currency_Symbol
gc ; Sm        ; Math_Symbol
gc ; So        ; Other_Symbol
gc ; Z         ; Separator                        # Zl | Zp | Zs
gc ; Zl        ; Line_Separator
gc ; Zp        ; Paragraph_Separator
gc ; Zs        ; Space_Separator

nt ; de        ; Decimal
nt ; di        ; Digit
nt ; n/a       ; None
nt ; nu        ; Numeric
END

    if (-e 'ArabicShaping.txt') {
        push @return, split /\n/, <<'END';
jg ; n/a       ; AIN
jg ; n/a       ; ALEF
jg ; n/a       ; DAL
jg ; n/a       ; GAF
jg ; n/a       ; LAM
jg ; n/a       ; MEEM
jg ; n/a       ; NO_JOINING_GROUP
jg ; n/a       ; NOON
jg ; n/a       ; QAF
jg ; n/a       ; SAD
jg ; n/a       ; SEEN
jg ; n/a       ; TAH
jg ; n/a       ; WAW

jt ; C         ; Join_Causing
jt ; D         ; Dual_Joining
jt ; L         ; Left_Joining
jt ; R         ; Right_Joining
jt ; U         ; Non_Joining
jt ; T         ; Transparent
END
        if ($v_version ge v3.0.0) {
            push @return, split /\n/, <<'END';
jg ; n/a       ; ALAPH
jg ; n/a       ; BEH
jg ; n/a       ; BETH
jg ; n/a       ; DALATH_RISH
jg ; n/a       ; E
jg ; n/a       ; FEH
jg ; n/a       ; FINAL_SEMKATH
jg ; n/a       ; GAMAL
jg ; n/a       ; HAH
jg ; n/a       ; HAMZA_ON_HEH_GOAL
jg ; n/a       ; HE
jg ; n/a       ; HEH
jg ; n/a       ; HEH_GOAL
jg ; n/a       ; HETH
jg ; n/a       ; KAF
jg ; n/a       ; KAPH
jg ; n/a       ; KNOTTED_HEH
jg ; n/a       ; LAMADH
jg ; n/a       ; MIM
jg ; n/a       ; NUN
jg ; n/a       ; PE
jg ; n/a       ; QAPH
jg ; n/a       ; REH
jg ; n/a       ; REVERSED_PE
jg ; n/a       ; SADHE
jg ; n/a       ; SEMKATH
jg ; n/a       ; SHIN
jg ; n/a       ; SWASH_KAF
jg ; n/a       ; TAW
jg ; n/a       ; TEH_MARBUTA
jg ; n/a       ; TETH
jg ; n/a       ; YEH
jg ; n/a       ; YEH_BARREE
jg ; n/a       ; YEH_WITH_TAIL
jg ; n/a       ; YUDH
jg ; n/a       ; YUDH_HE
jg ; n/a       ; ZAIN
END
        }
    }


    if (-e 'EastAsianWidth.txt') {
        push @return, split /\n/, <<'END';
ea ; A         ; Ambiguous
ea ; F         ; Fullwidth
ea ; H         ; Halfwidth
ea ; N         ; Neutral
ea ; Na        ; Narrow
ea ; W         ; Wide
END
    }

    if (-e 'LineBreak.txt' || -e 'LBsubst.txt') {
        my @lb = split /\n/, <<'END';
lb ; AI        ; Ambiguous
lb ; AL        ; Alphabetic
lb ; B2        ; Break_Both
lb ; BA        ; Break_After
lb ; BB        ; Break_Before
lb ; BK        ; Mandatory_Break
lb ; CB        ; Contingent_Break
lb ; CL        ; Close_Punctuation
lb ; CM        ; Combining_Mark
lb ; CR        ; Carriage_Return
lb ; EX        ; Exclamation
lb ; GL        ; Glue
lb ; HY        ; Hyphen
lb ; ID        ; Ideographic
lb ; IN        ; Inseperable
lb ; IS        ; Infix_Numeric
lb ; LF        ; Line_Feed
lb ; NS        ; Nonstarter
lb ; NU        ; Numeric
lb ; OP        ; Open_Punctuation
lb ; PO        ; Postfix_Numeric
lb ; PR        ; Prefix_Numeric
lb ; QU        ; Quotation
lb ; SA        ; Complex_Context
lb ; SG        ; Surrogate
lb ; SP        ; Space
lb ; SY        ; Break_Symbols
lb ; XX        ; Unknown
lb ; ZW        ; ZWSpace
END
        # If this Unicode version predates the lb property, we use our
        # substitute one
        if (-e 'LBsubst.txt') {
            $_ = s/^lb/_Perl_LB/r for @lb;
        }
        push @return, @lb;
    }

    if (-e 'DNormalizationProps.txt') {
        push @return, split /\n/, <<'END';
qc ; M         ; Maybe
qc ; N         ; No
qc ; Y         ; Yes
END
    }

    if (-e 'Scripts.txt') {
        push @return, split /\n/, <<'END';
sc ; Arab      ; Arabic
sc ; Armn      ; Armenian
sc ; Beng      ; Bengali
sc ; Bopo      ; Bopomofo
sc ; Cans      ; Canadian_Aboriginal
sc ; Cher      ; Cherokee
sc ; Cyrl      ; Cyrillic
sc ; Deva      ; Devanagari
sc ; Dsrt      ; Deseret
sc ; Ethi      ; Ethiopic
sc ; Geor      ; Georgian
sc ; Goth      ; Gothic
sc ; Grek      ; Greek
sc ; Gujr      ; Gujarati
sc ; Guru      ; Gurmukhi
sc ; Hang      ; Hangul
sc ; Hani      ; Han
sc ; Hebr      ; Hebrew
sc ; Hira      ; Hiragana
sc ; Ital      ; Old_Italic
sc ; Kana      ; Katakana
sc ; Khmr      ; Khmer
sc ; Knda      ; Kannada
sc ; Laoo      ; Lao
sc ; Latn      ; Latin
sc ; Mlym      ; Malayalam
sc ; Mong      ; Mongolian
sc ; Mymr      ; Myanmar
sc ; Ogam      ; Ogham
sc ; Orya      ; Oriya
sc ; Qaai      ; Inherited
sc ; Runr      ; Runic
sc ; Sinh      ; Sinhala
sc ; Syrc      ; Syriac
sc ; Taml      ; Tamil
sc ; Telu      ; Telugu
sc ; Thaa      ; Thaana
sc ; Thai      ; Thai
sc ; Tibt      ; Tibetan
sc ; Yiii      ; Yi
sc ; Zyyy      ; Common
END
    }

    if ($v_version ge v2.0.0) {
        push @return, split /\n/, <<'END';
dt ; com       ; compat
dt ; nar       ; narrow
dt ; sml       ; small
dt ; vert      ; vertical
dt ; wide      ; wide

gc ; Cf        ; Format
gc ; Cs        ; Surrogate
gc ; Lt        ; Titlecase_Letter
gc ; Me        ; Enclosing_Mark
gc ; Nl        ; Letter_Number
gc ; Pc        ; Connector_Punctuation
gc ; Sk        ; Modifier_Symbol
END
    }
    if ($v_version ge v2.1.2) {
        push @return, "bc ; S         ; Segment_Separator\n";
    }
    if ($v_version ge v2.1.5) {
        push @return, split /\n/, <<'END';
gc ; Pf        ; Final_Punctuation
gc ; Pi        ; Initial_Punctuation
END
    }
    if ($v_version ge v2.1.8) {
        push @return, "ccc; 240; IS   ; Iota_Subscript\n";
    }

    if ($v_version ge v3.0.0) {
        push @return, split /\n/, <<'END';
bc ; AL        ; Arabic_Letter
bc ; BN        ; Boundary_Neutral
bc ; LRE       ; Left_To_Right_Embedding
bc ; LRO       ; Left_To_Right_Override
bc ; NSM       ; Nonspacing_Mark
bc ; PDF       ; Pop_Directional_Format
bc ; RLE       ; Right_To_Left_Embedding
bc ; RLO       ; Right_To_Left_Override

ccc; 233; DB   ; Double_Below
END
    }

    if ($v_version ge v3.1.0) {
        push @return, "ccc; 226; R    ; Right\n";
    }

    return @return;
}

sub process_NormalizationsTest($file) {

    # Each line looks like:
    #      source code point; NFC; NFD; NFKC; NFKD
    # e.g.
    #       1E0A;1E0A;0044 0307;1E0A;0044 0307;

    # Process each line of the file ...
    while ($file->next_line) {

        next if /^@/;

        my ($c1, $c2, $c3, $c4, $c5) = split /\s*;\s*/;

        foreach my $var (\$c1, \$c2, \$c3, \$c4, \$c5) {
            $$var = pack "U0U*", map { hex } split " ", $$var;
            $$var =~ s/(\\)/$1$1/g;
        }

        push @normalization_tests,
                "Test_N(q$c1, q$c2, q$c3, q$c4, q$c5);\n";
    } # End of looping through the file
}

sub output_perl_charnames_line ($code_point, $name) {

    # Output the entries in Perl_charnames specially, using 5 digits instead
    # of four.  This makes the entries a constant length, and simplifies
    # charnames.pm which this table is for.  Unicode can have 6 digit
    # ordinals, but they are all private use or noncharacters which do not
    # have names, so won't be in this table.

    return sprintf "%05X\n%s\n\n", $code_point, $name;
}

{ # Closure

    # These are constants to the $property_info hash in this subroutine, to
    # avoid using a quoted-string which might have a typo.
    my $TYPE  = 'type';
    my $DEFAULT_MAP = 'default_map';
    my $DEFAULT_TABLE = 'default_table';
    my $PSEUDO_MAP_TYPE = 'pseudo_map_type';
    my $MISSINGS = 'missings';

    sub process_generic_property_file($file) {
        # This processes a file containing property mappings and puts them
        # into internal map tables.  It should be used to handle any property
        # files that have mappings from a code point or range thereof to
        # something else.  This means almost all the UCD .txt files.
        # each_line_handlers() should be set to adjust the lines of these
        # files, if necessary, to what this routine understands:
        #
        # 0374          ; NFD_QC; N
        # 003C..003E    ; Math
        #
        # the fields are: "codepoint-range ; property; map"
        #
        # meaning the codepoints in the range all have the value 'map' under
        # 'property'.
        # Beginning and trailing white space in each field are not significant.
        # Note there is not a trailing semi-colon in the above.  A trailing
        # semi-colon means the map is a null-string.  An omitted map, as
        # opposed to a null-string, is assumed to be 'Y', based on Unicode
        # table syntax.  (This could have been hidden from this routine by
        # doing it in the $file object, but that would require parsing of the
        # line there, so would have to parse it twice, or change the interface
        # to pass this an array.  So not done.)
        #
        # The map field may begin with a sequence of commands that apply to
        # this range.  Each such command begins and ends with $CMD_DELIM.
        # These are used to indicate, for example, that the mapping for a
        # range has a non-default type.
        #
        # This loops through the file, calling its next_line() method, and
        # then taking the map and adding it to the property's table.
        # Complications arise because any number of properties can be in the
        # file, in any order, interspersed in any way.  The first time a
        # property is seen, it gets information about that property and
        # caches it for quick retrieval later.  It also normalizes the maps
        # so that only one of many synonyms is stored.  The Unicode input
        # files do use some multiple synonyms.

        my %property_info;               # To keep track of what properties
                                         # have already had entries in the
                                         # current file, and info about each,
                                         # so don't have to recompute.
        my $property_name;               # property currently being worked on
        my $property_type;               # and its type
        my $previous_property_name = ""; # name from last time through loop
        my $property_object;             # pointer to the current property's
                                         # object
        my $property_addr;               # the address of that object
        my $default_map;                 # the string that code points missing
                                         # from the file map to
        my $default_table;               # For non-string properties, a
                                         # reference to the match table that
                                         # will contain the list of code
                                         # points that map to $default_map.

        # Get the next real non-comment line
        LINE:
        while ($file->next_line) {

            # Default replacement type; means that if parts of the range have
            # already been stored in our tables, the new map overrides them if
            # they differ more than cosmetically
            my $replace = $IF_NOT_EQUIVALENT;
            my $map_type;            # Default type for the map of this range

            #local $to_trace = 1 if main::DEBUG;
            trace $_ if main::DEBUG && $to_trace;

            # Split the line into components
            my ($range, $property_name, $map, @remainder)
                = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields

            # If more or less on the line than we are expecting, warn and skip
            # the line
            if (@remainder) {
                $file->carp_bad_line('Extra fields');
                next LINE;
            }
            elsif ( ! defined $property_name) {
                $file->carp_bad_line('Missing property');
                next LINE;
            }

            # Examine the range.
            if ($range !~ /^ ($code_point_re) (?:\.\. ($code_point_re) )? $/x)
            {
                $file->carp_bad_line("Range '$range' not of the form 'CP1' or 'CP1..CP2' (where CP1,2 are code points in hex)");
                next LINE;
            }
            my $low = hex $1;
            my $high = (defined $2) ? hex $2 : $low;

            # If changing to a new property, get the things constant per
            # property
            if ($previous_property_name ne $property_name) {

                $property_object = property_ref($property_name);
                if (! defined $property_object) {
                    $file->carp_bad_line("Unexpected property '$property_name'.  Skipped");
                    next LINE;
                }
                $property_addr = pack 'J', refaddr $property_object;

                # Defer changing names until have a line that is acceptable
                # (the 'next' statement above means is unacceptable)
                $previous_property_name = $property_name;

                # If not the first time for this property, retrieve info about
                # it from the cache
                my $this_property_info = $property_info{$property_addr};
                if (defined ($this_property_info->{$TYPE})) {
                    $property_type = $this_property_info->{$TYPE};
                    $default_map = $this_property_info->{$DEFAULT_MAP};
                    $map_type = $this_property_info->{$PSEUDO_MAP_TYPE};
                    $default_table = $this_property_info->{$DEFAULT_TABLE};
                }
                else {

                    # Here, is the first time for this property.  Set up the
                    # cache.
                    $property_type = $this_property_info->{$TYPE}
                                   = $property_object->type;
                    $map_type
                        = $this_property_info->{$PSEUDO_MAP_TYPE}
                        = $property_object->pseudo_map_type;

                    # The Unicode files are set up so that if the map is not
                    # defined, it is a binary property
                    if (! defined $map && $property_type != $BINARY) {
                        if ($property_type != $UNKNOWN
                            && $property_type != $NON_STRING)
                        {
                            $file->carp_bad_line("No mapping defined on a non-binary property.  Using 'Y' for the map");
                        }
                        else {
                            $property_object->set_type($BINARY);
                            $property_type = $this_property_info->{$TYPE}
                                           = $BINARY;
                        }
                    }

                    # Get any @missings default for this property.  This
                    # should precede the first entry for the property in the
                    # input file, and is located in a comment that has been
                    # stored by the Input_file class until we access it here.
                    # It's possible that there is more than one such line
                    # waiting for us; collect them all, and parse
                    my @missings_list;
                    @missings_list = $file->get_missings
                                            if $file->has_missings_defaults;

                    foreach my $default_ref (@missings_list) {

                        # For now, we are only interested in the fallback
                        # default for the entire property. i.e., an @missing
                        # line that is for the whole Unicode range.
                        next if $default_ref->{start} != 0
                             || $default_ref->{end} != $MAX_UNICODE_CODEPOINT;

                        $default_map = $default_ref->{default};

                        # For string properties, the default is just what the
                        # file says, but non-string properties should already
                        # have set up a table for the default property value;
                        # use the table for these, so can resolve synonyms
                        # later to a single standard one.
                        if ($property_type == $STRING
                            || $property_type == $UNKNOWN)
                        {
                            $this_property_info->{$MISSINGS} = $default_map;
                        }
                        else {
                            $default_map =
                               $property_object->table($default_map)->full_name;
                            $this_property_info->{$MISSINGS} = $default_map;
                            $this_property_info->{$DEFAULT_MAP} = $default_map;
                            if (! defined $property_object->default_map) {
                                $property_object->set_default_map($default_map);
                            }
                        }
                    }

                    # For later Unicode versions, multiple @missing lines for
                    # a single property can appear in the files.  The first
                    # always applies to the entire Unicode range, and was
                    # handled above.  The subsequent ones are for smaller
                    # ranges, and can be read as "But for this range, the
                    # default is ...".  So each overrides all the preceding
                    # ones for the range it applies to.  Typically they apply
                    # to disjoint ranges, but don't have to.  What we do is to
                    # set them up to work in reverse order, so that after the
                    # rest of the table is filled, the highest priority
                    # default range fills in any code points that haven't been
                    # specified; then the next highest priority one is
                    # applied, and so forth.
                    if (@missings_list > 1 && $v_version ge v15.0.0) {
                        if ($property_type != $ENUM) {
                            Carp::my_carp_bug("Multiple \@missings lines only"
                                            . " make sense for ENUM-type"
                                            . " properties.  Changing type to"
                                            . " that");
                            $property_type = $this_property_info->{$TYPE}
                                                                        = $ENUM;
                            $property_object->set_type($ENUM);
                        }

                        my $multi = Multi_Default->new();

                        # The overall default should be first on this list,
                        # and is handled differently than the rest.
                        $default_map = shift @missings_list;
                        Carp::my_carp_bug("\@missings needs to be entire range")
                            if $default_map->{start} != 0
                            || $default_map->{end} != $MAX_UNICODE_CODEPOINT;

                        # We already have looked at this line above.  Use that
                        # result
                        $multi->set_final_default($this_property_info->
                                                                  {$MISSINGS});

                        # Now get the individual range elements, and add them
                        # to Multi_Default object
                        while (@missings_list) {
                            my $this_entry = pop @missings_list;
                            my $subrange_default = $this_entry->{default};

                            # Use the short name as a standard
                            $subrange_default = $property_object->
                                        table($subrange_default)->short_name;
                            $multi->append_default($subrange_default,
                                "Range_List->new(Initialize => Range->new("
                              . "$this_entry->{start}, $this_entry->{end}))");
                        }

                        # Override the property's simple default with this.
                        $property_object->set_default_map($multi);
                    }

                    if (! $default_map || $property_type != $ENUM) {

                        # Finished storing all the @missings defaults in the
                        # input file so far.  Get the one for the current
                        # property.
                        my $missings = $this_property_info->{$MISSINGS};

                        # But we likely have separately stored what the
                        # default should be.  (This is to accommodate versions
                        # of the standard where the @missings lines are absent
                        # or incomplete.)  Hopefully the two will match.  But
                        # check it out.
                        $default_map = $property_object->default_map;

                        # If the map is a ref, it means that the default won't
                        # be processed until later, so undef it, so next few
                        # lines will redefine it to something that nothing
                        # will match
                        undef $default_map if ref $default_map;

                        # Create a $default_map if don't have one; maybe a
                        # dummy that won't match anything.
                        if (! defined $default_map) {

                            # Use any @missings line in the file.
                            if (defined $missings) {
                                if (ref $missings) {
                                    $default_map = $missings->full_name;
                                    $default_table = $missings;
                                }
                                else {
                                    $default_map = $missings;
                                }

                                # And store it with the property for outside
                                # use.
                                $property_object->set_default_map($default_map);
                            }
                            else {

                                # Neither an @missings nor a default map.
                                # Create a dummy one, so won't have to test
                                # definedness in the main loop.
                                $default_map = '_Perl This will never be in a'
                                             . ' file from Unicode';
                            }
                        }

                        # Here, we have $default_map defined, possibly in
                        # terms of $missings, but maybe not, and possibly is a
                        # dummy one.
                        if (defined $missings) {

                            # Make sure there is no conflict between the two.
                            # $missings has priority.
                            if (ref $missings) {
                                $default_table
                                        = $property_object->table($default_map);
                                if ( ! defined $default_table
                                    || $default_table != $missings)
                                {
                                    if (! defined $default_table) {
                                        $default_table = $UNDEF;
                                    }
                                    $file->carp_bad_line(<<END
The \@missings line for $property_name in $file says that missings default to
$missings, but we expect it to be $default_table.  $missings used.
END
                                    );
                                    $default_table = $missings;
                                    $default_map = $missings->full_name;
                                }
                                $this_property_info->{$DEFAULT_TABLE}
                                                            = $default_table;
                            }
                            elsif ($default_map ne $missings) {
                                $file->carp_bad_line(<<END
The \@missings line for $property_name in $file says that missings default to
$missings, but we expect it to be $default_map.  $missings used.
END
                                );
                                $default_map = $missings;
                            }
                        }

                        $this_property_info->{$DEFAULT_MAP} = $default_map;

                        # If haven't done so already, find the table
                        # corresponding to this map for non-string properties.
                        if (! defined $default_table
                            && $property_type != $STRING
                            && $property_type != $UNKNOWN)
                        {
                            $default_table
                                        = $this_property_info->{$DEFAULT_TABLE}
                                        = $property_object->table($default_map);
                        }
                    }
                } # End of is first time for this property
            } # End of switching properties.

            # Ready to process the line.
            # The Unicode files are set up so that if the map is not defined,
            # it is a binary property with value 'Y'
            if (! defined $map) {
                $map = 'Y';
            }
            else {

                # If the map begins with a special command to us (enclosed in
                # delimiters), extract the command(s).
                while ($map =~ s/ ^ $CMD_DELIM (.*?) $CMD_DELIM //x) {
                    my $command = $1;
                    if ($command =~  / ^ $REPLACE_CMD= (.*) /x) {
                        $replace = $1;
                    }
                    elsif ($command =~  / ^ $MAP_TYPE_CMD= (.*) /x) {
                        $map_type = $1;
                    }
                    else {
                        $file->carp_bad_line("Unknown command line: '$1'");
                        next LINE;
                    }
                }
            }

            if (   $default_map eq $CODE_POINT
                && $map =~ / ^ $code_point_re $/x)
            {

                # Here, we have a map to a particular code point, and the
                # default map is to a code point itself.  If the range
                # includes the particular code point, change that portion of
                # the range to the default.  This makes sure that in the final
                # table only the non-defaults are listed.
                my $decimal_map = hex $map;
                if ($low <= $decimal_map && $decimal_map <= $high) {

                    # If the range includes stuff before or after the map
                    # we're changing, split it and process the split-off parts
                    # later.
                    if ($low < $decimal_map) {
                        $file->insert_adjusted_lines(
                                            sprintf("%04X..%04X; %s; %s",
                                                    $low,
                                                    $decimal_map - 1,
                                                    $property_name,
                                                    $map));
                    }
                    if ($high > $decimal_map) {
                        $file->insert_adjusted_lines(
                                            sprintf("%04X..%04X; %s; %s",
                                                    $decimal_map + 1,
                                                    $high,
                                                    $property_name,
                                                    $map));
                    }
                    $low = $high = $decimal_map;
                    $map = $CODE_POINT;
                }
            }

            if ($property_type != $STRING && $property_type != $UNKNOWN) {
                my $table = $property_object->table($map);
                if (defined $table) {

                    # Unicode isn't very consistent about which synonym they
                    # use in their .txt files, even within the same file, or
                    # two files that are for the same property.  For enum
                    # properties, we know already what all the synonyms are
                    # (because we processed PropValueAliases already).
                    # Therefore we can take the input and map it to a uniform
                    # value now, saving us trouble later.
                    #
                    # Only if the map is well-behaved do we try this:
                    # non-empty, all non-blank.
                    if ($property_type == $ENUM && $map =~ / ^ \S+ $ /x) {

                        # Use existing practice as much as easily practicable,
                        # so that code that has assumptions about spelling
                        # doesn't have to change
                        my $short_name = $property_object->short_name;
                        if ($short_name =~ / ^ (BC | EA | GC  |HST | JT |
                                                Lb | BT | BPT | NFCQC |
                                                NFKCQC) $ /ix)
                        {
                            $map = $table->short_name;
                        }
                        elsif ($short_name !~ / ^ ( Ccc | Age | InSC | JG |
                                                    SB) $ /ix)
                        {
                            $map = $table->full_name;
                        }
                    }
                    elsif ($table == $default_table) {

                        # When it isn't an ENUM, we we can still tell if
                        # this is a synonym for the default map.  If so, use
                        # the default one instead.
                        $map = $default_map;
                    }
                }
            }

            # And figure out the map type if not known.
            if (! defined $map_type || $map_type == $COMPUTE_NO_MULTI_CP) {
                if ($map eq "") {   # Nulls are always $NULL map type
                    $map_type = $NULL;
                } # Otherwise, non-strings, and those that don't allow
                  # $MULTI_CP, and those that aren't multiple code points are
                  # 0
                elsif
                   (($property_type != $STRING && $property_type != $UNKNOWN)
                   || (defined $map_type && $map_type == $COMPUTE_NO_MULTI_CP)
                   || $map !~ /^ $code_point_re ( \  $code_point_re )+ $ /x)
                {
                    $map_type = 0;
                }
                else {
                    $map_type = $MULTI_CP;
                }
            }

            $property_object->add_map($low, $high,
                                        $map,
                                        Type => $map_type,
                                        Replace => $replace);
        } # End of loop through file's lines

        return;
    }
}

{ # Closure for UnicodeData.txt handling

    # This file was the first one in the UCD; its design leads to some
    # awkwardness in processing.  Here is a sample line:
    # 0041;LATIN CAPITAL LETTER A;Lu;0;L;;;;;N;;;;0061;
    # The fields in order are:
    my $i = 0;            # The code point is in field 0, and is shifted off.
    my $CHARNAME = $i++;  # character name (e.g. "LATIN CAPITAL LETTER A")
    my $CATEGORY = $i++;  # category (e.g. "Lu")
    my $CCC = $i++;       # Canonical combining class (e.g. "230")
    my $BIDI = $i++;      # directional class (e.g. "L")
    my $PERL_DECOMPOSITION = $i++;  # decomposition mapping
    my $PERL_DECIMAL_DIGIT = $i++;   # decimal digit value
    my $NUMERIC_TYPE_OTHER_DIGIT = $i++; # digit value, like a superscript
                                         # Dual-use in this program; see below
    my $NUMERIC = $i++;   # numeric value
    my $MIRRORED = $i++;  # ? mirrored
    my $UNICODE_1_NAME = $i++; # name in Unicode 1.0
    my $COMMENT = $i++;   # iso comment
    my $UPPER = $i++;     # simple uppercase mapping
    my $LOWER = $i++;     # simple lowercase mapping
    my $TITLE = $i++;     # simple titlecase mapping
    my $input_field_count = $i;

    # This routine in addition outputs these extra fields:

    my $DECOMP_TYPE = $i++; # Decomposition type

    # These fields are modifications of ones above, and are usually
    # suppressed; they must come last, as for speed, the loop upper bound is
    # normally set to ignore them
    my $NAME = $i++;        # This is the strict name field, not the one that
                            # charnames uses.
    my $DECOMP_MAP = $i++;  # Strict decomposition mapping; not the one used
                            # by Unicode::Normalize
    my $last_field = $i - 1;

    # All these are read into an array for each line, with the indices defined
    # above.  The empty fields in the example line above indicate that the
    # value is defaulted.  The handler called for each line of the input
    # changes these to their defaults.

    # Here are the official names of the properties, in a parallel array:
    my @field_names;
    $field_names[$BIDI] = 'Bidi_Class';
    $field_names[$CATEGORY] = 'General_Category';
    $field_names[$CCC] = 'Canonical_Combining_Class';
    $field_names[$CHARNAME] = 'Perl_Charnames';
    $field_names[$COMMENT] = 'ISO_Comment';
    $field_names[$DECOMP_MAP] = 'Decomposition_Mapping';
    $field_names[$DECOMP_TYPE] = 'Decomposition_Type';
    $field_names[$LOWER] = 'Lowercase_Mapping';
    $field_names[$MIRRORED] = 'Bidi_Mirrored';
    $field_names[$NAME] = 'Name';
    $field_names[$NUMERIC] = 'Numeric_Value';
    $field_names[$NUMERIC_TYPE_OTHER_DIGIT] = 'Numeric_Type';
    $field_names[$PERL_DECIMAL_DIGIT] = 'Perl_Decimal_Digit';
    $field_names[$PERL_DECOMPOSITION] = 'Perl_Decomposition_Mapping';
    $field_names[$TITLE] = 'Titlecase_Mapping';
    $field_names[$UNICODE_1_NAME] = 'Unicode_1_Name';
    $field_names[$UPPER] = 'Uppercase_Mapping';

    # Some of these need a little more explanation:
    # The $PERL_DECIMAL_DIGIT field does not lead to an official Unicode
    #   property, but is used in calculating the Numeric_Type.  Perl however,
    #   creates a file from this field, so a Perl property is created from it.
    # Similarly, the Other_Digit field is used only for calculating the
    #   Numeric_Type, and so it can be safely re-used as the place to store
    #   the value for Numeric_Type; hence it is referred to as
    #   $NUMERIC_TYPE_OTHER_DIGIT.
    # The input field named $PERL_DECOMPOSITION is a combination of both the
    #   decomposition mapping and its type.  Perl creates a file containing
    #   exactly this field, so it is used for that.  The two properties are
    #   separated into two extra output fields, $DECOMP_MAP and $DECOMP_TYPE.
    #   $DECOMP_MAP is usually suppressed (unless the lists are changed to
    #   output it), as Perl doesn't use it directly.
    # The input field named here $CHARNAME is used to construct the
    #   Perl_Charnames property, which is a combination of the Name property
    #   (which the input field contains), and the Unicode_1_Name property, and
    #   others from other files.  Since, the strict Name property is not used
    #   by Perl, this field is used for the table that Perl does use.  The
    #   strict Name property table is usually suppressed (unless the lists are
    #   changed to output it), so it is accumulated in a separate field,
    #   $NAME, which to save time is discarded unless the table is actually to
    #   be output

    # This file is processed like most in this program.  Control is passed to
    # process_generic_property_file() which calls filter_UnicodeData_line()
    # for each input line.  This filter converts the input into line(s) that
    # process_generic_property_file() understands.  There is also a setup
    # routine called before any of the file is processed, and a handler for
    # EOF processing, all in this closure.

    # A huge speed-up occurred at the cost of some added complexity when these
    # routines were altered to buffer the outputs into ranges.  Almost all the
    # lines of the input file apply to just one code point, and for most
    # properties, the map for the next code point up is the same as the
    # current one.  So instead of creating a line for each property for each
    # input line, filter_UnicodeData_line() remembers what the previous map
    # of a property was, and doesn't generate a line to pass on until it has
    # to, as when the map changes; and that passed-on line encompasses the
    # whole contiguous range of code points that have the same map for that
    # property.  This means a slight amount of extra setup, and having to
    # flush these buffers on EOF, testing if the maps have changed, plus
    # remembering state information in the closure.  But it means a lot less
    # real time in not having to change the data base for each property on
    # each line.

    # Another complication is that there are already a few ranges designated
    # in the input.  There are two lines for each, with the same maps except
    # the code point and name on each line.  This was actually the hardest
    # thing to design around.  The code points in those ranges may actually
    # have real maps not given by these two lines.  These maps will either
    # be algorithmically determinable, or be in the extracted files furnished
    # with the UCD.  In the event of conflicts between these extracted files,
    # and this one, Unicode says that this one prevails.  But it shouldn't
    # prevail for conflicts that occur in these ranges.  The data from the
    # extracted files prevails in those cases.  So, this program is structured
    # so that those files are processed first, storing maps.  Then the other
    # files are processed, generally overwriting what the extracted files
    # stored.  But just the range lines in this input file are processed
    # without overwriting.  This is accomplished by adding a special string to
    # the lines output to tell process_generic_property_file() to turn off the
    # overwriting for just this one line.
    # A similar mechanism is used to tell it that the map is of a non-default
    # type.

    sub setup_UnicodeData($file) { # Called before any lines of the input are read

        # Create a new property specially located that is a combination of
        # various Name properties: Name, Unicode_1_Name, Named Sequences, and
        # _Perl_Name_Alias properties.  (The final one duplicates elements of the
        # first, and starting in v6.1, is the same as the 'Name_Alias
        # property.)  A comment for the new property will later be constructed
        # based on the actual properties present and used
        $perl_charname = Property->new('Perl_Charnames',
                       Default_Map => "",
                       Directory => File::Spec->curdir(),
                       File => 'Name',
                       Fate => $INTERNAL_ONLY,
                       Perl_Extension => 1,
                       Range_Size_1 => \&output_perl_charnames_line,
                       Type => $STRING,
                       );
        $perl_charname->set_proxy_for('Name');

        my $Perl_decomp = Property->new('Perl_Decomposition_Mapping',
                                        Directory => File::Spec->curdir(),
                                        File => 'Decomposition',
                                        Format => $DECOMP_STRING_FORMAT,
                                        Fate => $INTERNAL_ONLY,
                                        Perl_Extension => 1,
                                        Default_Map => $CODE_POINT,

                                        # normalize.pm can't cope with these
                                        Output_Range_Counts => 0,

                                        # This is a specially formatted table
                                        # explicitly for normalize.pm, which
                                        # is expecting a particular format,
                                        # which means that mappings containing
                                        # multiple code points are in the main
                                        # body of the table
                                        Map_Type => $COMPUTE_NO_MULTI_CP,
                                        Type => $STRING,
                                        To_Output_Map => $INTERNAL_MAP,
                                        );
        $Perl_decomp->set_proxy_for('Decomposition_Mapping', 'Decomposition_Type');
        $Perl_decomp->add_comment(join_lines(<<END
This mapping is a combination of the Unicode 'Decomposition_Type' and
'Decomposition_Mapping' properties, formatted for use by normalize.pm.  It is
identical to the official Unicode 'Decomposition_Mapping' property except for
two things:
 1) It omits the algorithmically determinable Hangul syllable decompositions,
which normalize.pm handles algorithmically.
 2) It contains the decomposition type as well.  Non-canonical decompositions
begin with a word in angle brackets, like <super>, which denotes the
compatible decomposition type.  If the map does not begin with the <angle
brackets>, the decomposition is canonical.
END
        ));

        my $Decimal_Digit = Property->new("Perl_Decimal_Digit",
                                        Default_Map => "",
                                        Perl_Extension => 1,
                                        Directory => $map_directory,
                                        Type => $STRING,
                                        To_Output_Map => $OUTPUT_ADJUSTED,
                                        );
        $Decimal_Digit->add_comment(join_lines(<<END
This file gives the mapping of all code points which represent a single
decimal digit [0-9] to their respective digits, but it has ranges of 10 code
points, and the mapping of each non-initial element of each range is actually
not to "0", but to the offset that element has from its corresponding DIGIT 0.
These code points are those that have Numeric_Type=Decimal; not special
things, like subscripts nor Roman numerals.
END
        ));

        # These properties are not used for generating anything else, and are
        # usually not output.  By making them last in the list, we can just
        # change the high end of the loop downwards to avoid the work of
        # generating a table(s) that is/are just going to get thrown away.
        if (! property_ref('Decomposition_Mapping')->to_output_map
            && ! property_ref('Name')->to_output_map)
        {
            $last_field = min($NAME, $DECOMP_MAP) - 1;
        } elsif (property_ref('Decomposition_Mapping')->to_output_map) {
            $last_field = $DECOMP_MAP;
        } elsif (property_ref('Name')->to_output_map) {
            $last_field = $NAME;
        }
        return;
    }

    my $first_time = 1;                 # ? Is this the first line of the file
    my $in_range = 0;                   # ? Are we in one of the file's ranges
    my $previous_cp;                    # hex code point of previous line
    my $decimal_previous_cp = -1;       # And its decimal equivalent
    my @start;                          # For each field, the current starting
                                        # code point in hex for the range
                                        # being accumulated.
    my @fields;                         # The input fields;
    my @previous_fields;                # And those from the previous call

    sub filter_UnicodeData_line($file) {
        # Handle a single input line from UnicodeData.txt; see comments above
        # Conceptually this takes a single line from the file containing N
        # properties, and converts it into N lines with one property per line,
        # which is what the final handler expects.  But there are
        # complications due to the quirkiness of the input file, and to save
        # time, it accumulates ranges where the property values don't change
        # and only emits lines when necessary.  This is about an order of
        # magnitude fewer lines emitted.

        # $_ contains the input line.
        # -1 in split means retain trailing null fields
        (my $cp, @fields) = split /\s*;\s*/, $_, -1;

        #local $to_trace = 1 if main::DEBUG;
        trace $cp, @fields , $input_field_count if main::DEBUG && $to_trace;
        if (@fields > $input_field_count) {
            $file->carp_bad_line('Extra fields');
            $_ = "";
            return;
        }

        my $decimal_cp = hex $cp;

        # We have to output all the buffered ranges when the next code point
        # is not exactly one after the previous one, which means there is a
        # gap in the ranges.
        my $force_output = ($decimal_cp != $decimal_previous_cp + 1);

        # The decomposition mapping field requires special handling.  It looks
        # like either:
        #
        # <compat> 0032 0020
        # 0041 0300
        #
        # The decomposition type is enclosed in <brackets>; if missing, it
        # means the type is canonical.  There are two decomposition mapping
        # tables: the one for use by Perl's normalize.pm has a special format
        # which is this field intact; the other, for general use is of
        # standard format.  In either case we have to find the decomposition
        # type.  Empty fields have None as their type, and map to the code
        # point itself
        if ($fields[$PERL_DECOMPOSITION] eq "") {
            $fields[$DECOMP_TYPE] = 'None';
            $fields[$DECOMP_MAP] = $fields[$PERL_DECOMPOSITION] = $CODE_POINT;
        }
        else {
            ($fields[$DECOMP_TYPE], my $map) = $fields[$PERL_DECOMPOSITION]
                                            =~ / < ( .+? ) > \s* ( .+ ) /x;
            if (! defined $fields[$DECOMP_TYPE]) {
                $fields[$DECOMP_TYPE] = 'Canonical';
                $fields[$DECOMP_MAP] = $fields[$PERL_DECOMPOSITION];
            }
            else {
                $fields[$DECOMP_MAP] = $map;
            }
        }

        # The 3 numeric fields also require special handling.  The 2 digit
        # fields must be either empty or match the number field.  This means
        # that if it is empty, they must be as well, and the numeric type is
        # None, and the numeric value is 'Nan'.
        # The decimal digit field must be empty or match the other digit
        # field.  If the decimal digit field is non-empty, the code point is
        # a decimal digit, and the other two fields will have the same value.
        # If it is empty, but the other digit field is non-empty, the code
        # point is an 'other digit', and the number field will have the same
        # value as the other digit field.  If the other digit field is empty,
        # but the number field is non-empty, the code point is a generic
        # numeric type.
        if ($fields[$NUMERIC] eq "") {
            if ($fields[$PERL_DECIMAL_DIGIT] ne ""
                || $fields[$NUMERIC_TYPE_OTHER_DIGIT] ne ""
            ) {
                $file->carp_bad_line("Numeric values inconsistent.  Trying to process anyway");
            }
            $fields[$NUMERIC_TYPE_OTHER_DIGIT] = 'None';
            $fields[$NUMERIC] = 'NaN';
        }
        else {
            $file->carp_bad_line("'$fields[$NUMERIC]' should be a whole or rational number.  Processing as if it were") if $fields[$NUMERIC] !~ qr{ ^ -? \d+ ( / \d+ )? $ }x;
            if ($fields[$PERL_DECIMAL_DIGIT] ne "") {
                $file->carp_bad_line("$fields[$PERL_DECIMAL_DIGIT] should equal $fields[$NUMERIC].  Processing anyway") if $fields[$PERL_DECIMAL_DIGIT] != $fields[$NUMERIC];
                $file->carp_bad_line("$fields[$PERL_DECIMAL_DIGIT] should be empty since the general category ($fields[$CATEGORY]) isn't 'Nd'.  Processing as Decimal") if $fields[$CATEGORY] ne "Nd";
                $fields[$NUMERIC_TYPE_OTHER_DIGIT] = 'Decimal';
            }
            elsif ($fields[$NUMERIC_TYPE_OTHER_DIGIT] ne "") {
                $file->carp_bad_line("$fields[$NUMERIC_TYPE_OTHER_DIGIT] should equal $fields[$NUMERIC].  Processing anyway") if $fields[$NUMERIC_TYPE_OTHER_DIGIT] != $fields[$NUMERIC];
                $fields[$NUMERIC_TYPE_OTHER_DIGIT] = 'Digit';
            }
            else {
                $fields[$NUMERIC_TYPE_OTHER_DIGIT] = 'Numeric';

                # Rationals require extra effort.
                if ($fields[$NUMERIC] =~ qr{/}) {
                    reduce_fraction(\$fields[$NUMERIC]);
                    register_fraction($fields[$NUMERIC])
                }
            }
        }

        # For the properties that have empty fields in the file, and which
        # mean something different from empty, change them to that default.
        # Certain fields just haven't been empty so far in any Unicode
        # version, so don't look at those, namely $MIRRORED, $BIDI, $CCC,
        # $CATEGORY.  This leaves just the two fields, and so we hard-code in
        # the defaults; which are very unlikely to ever change.
        $fields[$UPPER] = $CODE_POINT if $fields[$UPPER] eq "";
        $fields[$LOWER] = $CODE_POINT if $fields[$LOWER] eq "";

        # UAX44 says that if title is empty, it is the same as whatever upper
        # is,
        $fields[$TITLE] = $fields[$UPPER] if $fields[$TITLE] eq "";

        # There are a few pairs of lines like:
        #   AC00;<Hangul Syllable, First>;Lo;0;L;;;;;N;;;;;
        #   D7A3;<Hangul Syllable, Last>;Lo;0;L;;;;;N;;;;;
        # that define ranges.  These should be processed after the fields are
        # adjusted above, as they may override some of them; but mostly what
        # is left is to possibly adjust the $CHARNAME field.  The names of all the
        # paired lines start with a '<', but this is also true of '<control>,
        # which isn't one of these special ones.
        if ($fields[$CHARNAME] eq '<control>') {

            # Some code points in this file have the pseudo-name
            # '<control>', but the official name for such ones is the null
            # string.
            $fields[$NAME] = $fields[$CHARNAME] = "";

            # We had better not be in between range lines.
            if ($in_range) {
                $file->carp_bad_line("Expecting a closing range line, not a $fields[$CHARNAME]'.  Trying anyway");
                $in_range = 0;
            }
        }
        elsif (substr($fields[$CHARNAME], 0, 1) ne '<') {

            # Here is a non-range line.  We had better not be in between range
            # lines.
            if ($in_range) {
                $file->carp_bad_line("Expecting a closing range line, not a $fields[$CHARNAME]'.  Trying anyway");
                $in_range = 0;
            }
            if ($fields[$CHARNAME] =~ s/- $cp $//x) {

                # These are code points whose names end in their code points,
                # which means the names are algorithmically derivable from the
                # code points.  To shorten the output Name file, the algorithm
                # for deriving these is placed in the file instead of each
                # code point, so they have map type $CP_IN_NAME
                $fields[$CHARNAME] = $CMD_DELIM
                                 . $MAP_TYPE_CMD
                                 . '='
                                 . $CP_IN_NAME
                                 . $CMD_DELIM
                                 . $fields[$CHARNAME];
            }
            $fields[$NAME] = $fields[$CHARNAME];
        }
        elsif ($fields[$CHARNAME] =~ /^<(.+), First>$/) {
            $fields[$CHARNAME] = $fields[$NAME] = $1;

            # Here we are at the beginning of a range pair.
            if ($in_range) {
                $file->carp_bad_line("Expecting a closing range line, not a beginning one, $fields[$CHARNAME]'.  Trying anyway");
            }
            $in_range = 1;

            # Because the properties in the range do not overwrite any already
            # in the db, we must flush the buffers of what's already there, so
            # they get handled in the normal scheme.
            $force_output = 1;

        }
        elsif ($fields[$CHARNAME] !~ s/^<(.+), Last>$/$1/) {
            $file->carp_bad_line("Unexpected name starting with '<' $fields[$CHARNAME].  Ignoring this line.");
            $_ = "";
            return;
        }
        else { # Here, we are at the last line of a range pair.

            if (! $in_range) {
                $file->carp_bad_line("Unexpected end of range $fields[$CHARNAME] when not in one.  Ignoring this line.");
                $_ = "";
                return;
            }
            $in_range = 0;

            $fields[$NAME] = $fields[$CHARNAME];

            # Check that the input is valid: that the closing of the range is
            # the same as the beginning.
            foreach my $i (0 .. $last_field) {
                next if $fields[$i] eq $previous_fields[$i];
                $file->carp_bad_line("Expecting '$fields[$i]' to be the same as '$previous_fields[$i]'.  Bad News.  Trying anyway");
            }

            # The processing differs depending on the type of range,
            # determined by its $CHARNAME
            if ($fields[$CHARNAME] =~ /^Hangul Syllable/) {

                # Check that the data looks right.
                if ($decimal_previous_cp != $SBase) {
                    $file->carp_bad_line("Unexpected Hangul syllable start = $previous_cp.  Bad News.  Results will be wrong");
                }
                if ($decimal_cp != $SBase + $SCount - 1) {
                    $file->carp_bad_line("Unexpected Hangul syllable end = $cp.  Bad News.  Results will be wrong");
                }

                # The Hangul syllable range has a somewhat complicated name
                # generation algorithm.  Each code point in it has a canonical
                # decomposition also computable by an algorithm.  The
                # perl decomposition map table built from these is used only
                # by normalize.pm, which has the algorithm built in it, so the
                # decomposition maps are not needed, and are large, so are
                # omitted from it.  If the full decomposition map table is to
                # be output, the decompositions are generated for it, in the
                # EOF handling code for this input file.

                $previous_fields[$DECOMP_TYPE] = 'Canonical';

                # This range is stored in our internal structure with its
                # own map type, different from all others.
                $previous_fields[$CHARNAME] = $previous_fields[$NAME]
                                        = $CMD_DELIM
                                          . $MAP_TYPE_CMD
                                          . '='
                                          . $HANGUL_SYLLABLE
                                          . $CMD_DELIM
                                          . $fields[$CHARNAME];
            }
            elsif ($fields[$CATEGORY] eq 'Lo') {    # Is a letter

                # All the CJK ranges like this have the name given as a
                # special case in the next code line.  And for the others, we
                # hope that Unicode continues to use the correct name in
                # future releases, so we don't have to make further special
                # cases.
                my $name = ($fields[$CHARNAME] =~ /^CJK/)
                           ? 'CJK UNIFIED IDEOGRAPH'
                           : uc $fields[$CHARNAME];

                # The name for these contains the code point itself, and all
                # are defined to have the same base name, regardless of what
                # is in the file.  They are stored in our internal structure
                # with a map type of $CP_IN_NAME
                $previous_fields[$CHARNAME] = $previous_fields[$NAME]
                                        = $CMD_DELIM
                                           . $MAP_TYPE_CMD
                                           . '='
                                           . $CP_IN_NAME
                                           . $CMD_DELIM
                                           . $name;

            }
            elsif ($fields[$CATEGORY] eq 'Co'
                     || $fields[$CATEGORY] eq 'Cs')
            {
                # The names of all the code points in these ranges are set to
                # null, as there are no names for the private use and
                # surrogate code points.

                $previous_fields[$CHARNAME] = $previous_fields[$NAME] = "";
            }
            else {
                $file->carp_bad_line("Unexpected code point range $fields[$CHARNAME] because category is $fields[$CATEGORY].  Attempting to process it.");
            }

            # The first line of the range caused everything else to be output,
            # and then its values were stored as the beginning values for the
            # next set of ranges, which this one ends.  Now, for each value,
            # add a command to tell the handler that these values should not
            # replace any existing ones in our database.
            foreach my $i (0 .. $last_field) {
                $previous_fields[$i] = $CMD_DELIM
                                        . $REPLACE_CMD
                                        . '='
                                        . $NO
                                        . $CMD_DELIM
                                        . $previous_fields[$i];
            }

            # And change things so it looks like the entire range has been
            # gone through with this being the final part of it.  Adding the
            # command above to each field will cause this range to be flushed
            # during the next iteration, as it guaranteed that the stored
            # field won't match whatever value the next one has.
            $previous_cp = $cp;
            $decimal_previous_cp = $decimal_cp;

            # We are now set up for the next iteration; so skip the remaining
            # code in this subroutine that does the same thing, but doesn't
            # know about these ranges.
            $_ = "";

            return;
        }

        # On the very first line, we fake it so the code below thinks there is
        # nothing to output, and initialize so that when it does get output it
        # uses the first line's values for the lowest part of the range.
        # (One could avoid this by using peek(), but then one would need to
        # know the adjustments done above and do the same ones in the setup
        # routine; not worth it)
        if ($first_time) {
            $first_time = 0;
            @previous_fields = @fields;
            @start = ($cp) x scalar @fields;
            $decimal_previous_cp = $decimal_cp - 1;
        }

        # For each field, output the stored up ranges that this code point
        # doesn't fit in.  Earlier we figured out if all ranges should be
        # terminated because of changing the replace or map type styles, or if
        # there is a gap between this new code point and the previous one, and
        # that is stored in $force_output.  But even if those aren't true, we
        # need to output the range if this new code point's value for the
        # given property doesn't match the stored range's.
        #local $to_trace = 1 if main::DEBUG;
        foreach my $i (0 .. $last_field) {
            my $field = $fields[$i];
            if ($force_output || $field ne $previous_fields[$i]) {

                # Flush the buffer of stored values.
                $file->insert_adjusted_lines("$start[$i]..$previous_cp; $field_names[$i]; $previous_fields[$i]");

                # Start a new range with this code point and its value
                $start[$i] = $cp;
                $previous_fields[$i] = $field;
            }
        }

        # Set the values for the next time.
        $previous_cp = $cp;
        $decimal_previous_cp = $decimal_cp;

        # The input line has generated whatever adjusted lines are needed, and
        # should not be looked at further.
        $_ = "";
        return;
    }

    sub EOF_UnicodeData($file) {
        # Called upon EOF to flush the buffers, and create the Hangul
        # decomposition mappings if needed.

        # Flush the buffers.
        foreach my $i (0 .. $last_field) {
            $file->insert_adjusted_lines("$start[$i]..$previous_cp; $field_names[$i]; $previous_fields[$i]");
        }

        if (-e 'Jamo.txt') {

            # The algorithm is published by Unicode, based on values in
            # Jamo.txt, (which should have been processed before this
            # subroutine), and the results left in %Jamo
            unless (%Jamo) {
                Carp::my_carp_bug("Jamo.txt should be processed before Unicode.txt.  Hangul syllables not generated.");
                return;
            }

            # If the full decomposition map table is being output, insert
            # into it the Hangul syllable mappings.  This is to avoid having
            # to publish a subroutine in it to compute them.  (which would
            # essentially be this code.)  This uses the algorithm published by
            # Unicode.  (No hangul syllables in version 1)
            if ($v_version ge v2.0.0
                && property_ref('Decomposition_Mapping')->to_output_map) {
                for (my $S = $SBase; $S < $SBase + $SCount; $S++) {
                    use integer;
                    my $SIndex = $S - $SBase;
                    my $L = $LBase + $SIndex / $NCount;
                    my $V = $VBase + ($SIndex % $NCount) / $TCount;
                    my $T = $TBase + $SIndex % $TCount;

                    trace "L=$L, V=$V, T=$T" if main::DEBUG && $to_trace;
                    my $decomposition = sprintf("%04X %04X", $L, $V);
                    $decomposition .= sprintf(" %04X", $T) if $T != $TBase;
                    $file->insert_adjusted_lines(
                                sprintf("%04X; Decomposition_Mapping; %s",
                                        $S,
                                        $decomposition));
                }
            }
        }

        return;
    }

    sub filter_v1_ucd($file) {
        # Fix UCD lines in version 1.  This is probably overkill, but this
        # fixes some glaring errors in Version 1 UnicodeData.txt.  That file:
        # 1)    had many Hangul (U+3400 - U+4DFF) code points that were later
        #       removed.  This program retains them
        # 2)    didn't include ranges, which it should have, and which are now
        #       added in @corrected_lines below.  It was hand populated by
        #       taking the data from Version 2, verified by analyzing
        #       DAge.txt.
        # 3)    There is a syntax error in the entry for U+09F8 which could
        #       cause problems for Unicode::UCD, and so is changed.  It's
        #       numeric value was simply a minus sign, without any number.
        #       (Eventually Unicode changed the code point to non-numeric.)
        # 4)    The decomposition types often don't match later versions
        #       exactly, and the whole syntax of that field is different; so
        #       the syntax is changed as well as the types to their later
        #       terminology.  Otherwise normalize.pm would be very unhappy
        # 5)    Many ccc classes are different.  These are left intact.
        # 6)    U+FF10..U+FF19 are missing their numeric values in all three
        #       fields.  These are unchanged because it doesn't really cause
        #       problems for Perl.
        # 7)    A number of code points, such as controls, don't have their
        #       Unicode Version 1 Names in this file.  These are added.
        # 8)    A number of Symbols were marked as Lm.  This changes those in
        #       the Latin1 range, so that regexes work.
        # 9)    The odd characters U+03DB .. U+03E1 weren't encoded but are
        #       referred to by their lc equivalents.  Not fixed.

        my @corrected_lines = split /\n/, <<'END';
4E00;<CJK Ideograph, First>;Lo;0;L;;;;;N;;;;;
9FA5;<CJK Ideograph, Last>;Lo;0;L;;;;;N;;;;;
E000;<Private Use, First>;Co;0;L;;;;;N;;;;;
F8FF;<Private Use, Last>;Co;0;L;;;;;N;;;;;
F900;<CJK Compatibility Ideograph, First>;Lo;0;L;;;;;N;;;;;
FA2D;<CJK Compatibility Ideograph, Last>;Lo;0;L;;;;;N;;;;;
END

        #local $to_trace = 1 if main::DEBUG;
        trace $_ if main::DEBUG && $to_trace;

        # -1 => retain trailing null fields
        my ($code_point, @fields) = split /\s*;\s*/, $_, -1;

        # At the first place that is wrong in the input, insert all the
        # corrections, replacing the wrong line.
        if ($code_point eq '4E00') {
            my @copy = @corrected_lines;
            $_ = shift @copy;
            ($code_point, @fields) = split /\s*;\s*/, $_, -1;

            $file->insert_lines(@copy);
        }
        elsif ($code_point =~ /^00/ && $fields[$CATEGORY] eq 'Lm') {

            # There are no Lm characters in Latin1; these should be 'Sk', but
            # there isn't that in V1.
            $fields[$CATEGORY] = 'So';
        }

        if ($fields[$NUMERIC] eq '-') {
            $fields[$NUMERIC] = '-1';  # This is what 2.0 made it.
        }

        if  ($fields[$PERL_DECOMPOSITION] ne "") {

            # Several entries have this change to superscript 2 or 3 in the
            # middle.  Convert these to the modern version, which is to use
            # the actual U+00B2 and U+00B3 (the superscript forms) instead.
            # So 'HHHH HHHH <+sup> 0033 <-sup> HHHH' becomes
            # 'HHHH HHHH 00B3 HHHH'.
            # It turns out that all of these that don't have another
            # decomposition defined at the beginning of the line have the
            # <square> decomposition in later releases.
            if ($code_point ne '00B2' && $code_point ne '00B3') {
                if  ($fields[$PERL_DECOMPOSITION]
                                    =~ s/<\+sup> 003([23]) <-sup>/00B$1/)
                {
                    if (substr($fields[$PERL_DECOMPOSITION], 0, 1) ne '<') {
                        $fields[$PERL_DECOMPOSITION] = '<square> '
                        . $fields[$PERL_DECOMPOSITION];
                    }
                }
            }

            # If is like '<+circled> 0052 <-circled>', convert to
            # '<circled> 0052'
            $fields[$PERL_DECOMPOSITION] =~
                            s/ < \+ ( .*? ) > \s* (.*?) \s* <-\1> /<$1> $2/xg;

            # Convert '<join> HHHH HHHH <join>' to '<medial> HHHH HHHH', etc.
            $fields[$PERL_DECOMPOSITION] =~
                            s/ <join> \s* (.*?) \s* <no-join> /<final> $1/x
            or $fields[$PERL_DECOMPOSITION] =~
                            s/ <join> \s* (.*?) \s* <join> /<medial> $1/x
            or $fields[$PERL_DECOMPOSITION] =~
                            s/ <no-join> \s* (.*?) \s* <join> /<initial> $1/x
            or $fields[$PERL_DECOMPOSITION] =~
                        s/ <no-join> \s* (.*?) \s* <no-join> /<isolated> $1/x;

            # Convert '<break> HHHH HHHH <break>' to '<break> HHHH', etc.
            $fields[$PERL_DECOMPOSITION] =~
                    s/ <(break|no-break)> \s* (.*?) \s* <\1> /<$1> $2/x;

            # Change names to modern form.
            $fields[$PERL_DECOMPOSITION] =~ s/<font variant>/<font>/g;
            $fields[$PERL_DECOMPOSITION] =~ s/<no-break>/<noBreak>/g;
            $fields[$PERL_DECOMPOSITION] =~ s/<circled>/<circle>/g;
            $fields[$PERL_DECOMPOSITION] =~ s/<break>/<fraction>/g;

            # One entry has weird braces
            $fields[$PERL_DECOMPOSITION] =~ s/[{}]//g;

            # One entry at U+2116 has an extra <sup>
            $fields[$PERL_DECOMPOSITION] =~ s/( < .*? > .* ) < .*? > \ * /$1/x;
        }

        $_ = join ';', $code_point, @fields;
        trace $_ if main::DEBUG && $to_trace;
        return;
    }

    sub filter_bad_Nd_ucd {
        # Early versions specified a value in the decimal digit field even
        # though the code point wasn't a decimal digit.  Clear the field in
        # that situation, so that the main code doesn't think it is a decimal
        # digit.

        my ($code_point, @fields) = split /\s*;\s*/, $_, -1;
        if ($fields[$PERL_DECIMAL_DIGIT] ne "" && $fields[$CATEGORY] ne 'Nd') {
            $fields[$PERL_DECIMAL_DIGIT] = "";
            $_ = join ';', $code_point, @fields;
        }
        return;
    }

    my @U1_control_names = split /\n/, <<'END';
NULL
START OF HEADING
START OF TEXT
END OF TEXT
END OF TRANSMISSION
ENQUIRY
ACKNOWLEDGE
BELL
BACKSPACE
HORIZONTAL TABULATION
LINE FEED
VERTICAL TABULATION
FORM FEED
CARRIAGE RETURN
SHIFT OUT
SHIFT IN
DATA LINK ESCAPE
DEVICE CONTROL ONE
DEVICE CONTROL TWO
DEVICE CONTROL THREE
DEVICE CONTROL FOUR
NEGATIVE ACKNOWLEDGE
SYNCHRONOUS IDLE
END OF TRANSMISSION BLOCK
CANCEL
END OF MEDIUM
SUBSTITUTE
ESCAPE
FILE SEPARATOR
GROUP SEPARATOR
RECORD SEPARATOR
UNIT SEPARATOR
DELETE
BREAK PERMITTED HERE
NO BREAK HERE
INDEX
NEXT LINE
START OF SELECTED AREA
END OF SELECTED AREA
CHARACTER TABULATION SET
CHARACTER TABULATION WITH JUSTIFICATION
LINE TABULATION SET
PARTIAL LINE DOWN
PARTIAL LINE UP
REVERSE LINE FEED
SINGLE SHIFT TWO
SINGLE SHIFT THREE
DEVICE CONTROL STRING
PRIVATE USE ONE
PRIVATE USE TWO
SET TRANSMIT STATE
CANCEL CHARACTER
MESSAGE WAITING
START OF GUARDED AREA
END OF GUARDED AREA
START OF STRING
SINGLE CHARACTER INTRODUCER
CONTROL SEQUENCE INTRODUCER
STRING TERMINATOR
OPERATING SYSTEM COMMAND
PRIVACY MESSAGE
APPLICATION PROGRAM COMMAND
END

    sub filter_early_U1_names {
        # Very early versions did not have the Unicode_1_name field specified.
        # They differed in which ones were present; make sure a U1 name
        # exists, so that Unicode::UCD::charinfo will work

        my ($code_point, @fields) = split /\s*;\s*/, $_, -1;


        # @U1_control names above are entirely positional, so we pull them out
        # in the exact order required, with gaps for the ones that don't have
        # names.
        if ($code_point =~ /^00[01]/
            || $code_point eq '007F'
            || $code_point =~ /^008[2-9A-F]/
            || $code_point =~ /^009[0-8A-F]/)
        {
            my $u1_name = shift @U1_control_names;
            $fields[$UNICODE_1_NAME] = $u1_name unless $fields[$UNICODE_1_NAME];
            $_ = join ';', $code_point, @fields;
        }
        return;
    }

    sub filter_v2_1_5_ucd {
        # A dozen entries in this 2.1.5 file had the mirrored and numeric
        # columns swapped;  These all had mirrored be 'N'.  So if the numeric
        # column appears to be N, swap it back.

        my ($code_point, @fields) = split /\s*;\s*/, $_, -1;
        if ($fields[$NUMERIC] eq 'N') {
            $fields[$NUMERIC] = $fields[$MIRRORED];
            $fields[$MIRRORED] = 'N';
            $_ = join ';', $code_point, @fields;
        }
        return;
    }

    sub filter_v6_ucd {

        # Unicode 6.0 co-opted the name BELL for U+1F514, but until 5.17,
        # it wasn't accepted, to allow for some deprecation cycles.  This
        # function is not called after 5.16

        return if $_ !~ /^(?:0007|1F514|070F);/;

        my ($code_point, @fields) = split /\s*;\s*/, $_, -1;
        if ($code_point eq '0007') {
            $fields[$CHARNAME] = "";
        }
        elsif ($code_point eq '070F') { # Unicode Corrigendum #8; see
                            # http://www.unicode.org/versions/corrigendum8.html
            $fields[$BIDI] = "AL";
        }
        elsif ($^V lt v5.18.0) { # For 5.18 will convert to use Unicode's name
            $fields[$CHARNAME] = "";
        }

        $_ = join ';', $code_point, @fields;

        return;
    }
} # End closure for UnicodeData

sub process_GCB_test($file) {

    while ($file->next_line) {
        push @backslash_X_tests, $_;
    }

    return;
}

sub process_LB_test($file) {

    while ($file->next_line) {
        push @LB_tests, $_;
    }

    return;
}

sub process_SB_test($file) {

    while ($file->next_line) {
        push @SB_tests, $_;
    }

    return;
}

sub process_WB_test($file) {

    while ($file->next_line) {
        push @WB_tests, $_;
    }

    return;
}

sub process_NamedSequences($file) {
    # NamedSequences.txt entries are just added to an array.  Because these
    # don't look like the other tables, they have their own handler.
    # An example:
    # LATIN CAPITAL LETTER A WITH MACRON AND GRAVE;0100 0300
    #
    # This just adds the sequence to an array for later handling

    while ($file->next_line) {
        my ($name, $sequence, @remainder) = split /\s*;\s*/, $_, -1;
        if (@remainder) {
            $file->carp_bad_line(
                "Doesn't look like 'KHMER VOWEL SIGN OM;17BB 17C6'");
            next;
        }

        # Code points need to be 5 digits long like the other entries in
        # Name.pl, for regcomp.c parsing; and the ones below 0x0100 need to be
        # converted to native
        $sequence = join " ", map { sprintf("%05X",
                                    utf8::unicode_to_native(hex $_))
                                  } split / /, $sequence;
        push @named_sequences, "$sequence\n$name\n";
    }
    return;
}

{ # Closure

    my $first_range;

    sub  filter_early_ea_lb {
        # Fixes early EastAsianWidth.txt and LineBreak.txt files.  These had a
        # third field be the name of the code point, which can be ignored in
        # most cases.  But it can be meaningful if it marks a range:
        # 33FE;W;IDEOGRAPHIC TELEGRAPH SYMBOL FOR DAY THIRTY-ONE
        # 3400;W;<CJK Ideograph Extension A, First>
        #
        # We need to see the First in the example above to know it's a range.
        # They did not use the later range syntaxes.  This routine changes it
        # to use the modern syntax.
        # $1 is the Input_file object.

        my @fields = split /\s*;\s*/;
        if ($fields[2] =~ /^<.*, First>/) {
            $first_range = $fields[0];
            $_ = "";
        }
        elsif ($fields[2] =~ /^<.*, Last>/) {
            $_ = $_ = "$first_range..$fields[0]; $fields[1]";
        }
        else {
            undef $first_range;
            $_ = "$fields[0]; $fields[1]";
        }

        return;
    }
}

sub filter_substitute_lb {
    # Used on Unicodes that predate the LB property, where there is a
    # substitute file.  This just does the regular ea_lb handling for such
    # files, and then substitutes the long property value name for the short
    # one that comes with the file.  (The other break files have the long
    # names in them, so this is the odd one out.)  The reason for doing this
    # kludge is that regen/mk_invlists.pl is expecting the long name.  This
    # also fixes the typo 'Inseperable' that leads to problems.

    filter_early_ea_lb;
    return unless $_;

    my @fields = split /\s*;\s*/;
    $fields[1] = property_ref('_Perl_LB')->table($fields[1])->full_name;
    $fields[1] = 'Inseparable' if lc $fields[1] eq 'inseperable';
    $_ = join '; ', @fields;
}

sub filter_old_style_arabic_shaping {
    # Early versions used a different term for the later one.

    my @fields = split /\s*;\s*/;
    $fields[3] =~ s/<no shaping>/No_Joining_Group/;
    $fields[3] =~ s/\s+/_/g;                # Change spaces to underscores
    $_ = join ';', @fields;
    return;
}

{ # Closure
    my $lc; # Table for lowercase mapping
    my $tc;
    my $uc;
    my %special_casing_code_points;

    sub setup_special_casing($file) {
        # SpecialCasing.txt contains the non-simple case change mappings.  The
        # simple ones are in UnicodeData.txt, which should already have been
        # read in to the full property data structures, so as to initialize
        # these with the simple ones.  Then the SpecialCasing.txt entries
        # add or overwrite the ones which have different full mappings.

        # This routine sees if the simple mappings are to be output, and if
        # so, copies what has already been put into the full mapping tables,
        # while they still contain only the simple mappings.

        # The reason it is done this way is that the simple mappings are
        # probably not going to be output, so it saves work to initialize the
        # full tables with the simple mappings, and then overwrite those
        # relatively few entries in them that have different full mappings,
        # and thus skip the simple mapping tables altogether.

        $lc = property_ref('lc');
        $tc = property_ref('tc');
        $uc = property_ref('uc');

        # For each of the case change mappings...
        foreach my $full_casing_table ($lc, $tc, $uc) {
            my $full_casing_name = $full_casing_table->name;
            my $full_casing_full_name = $full_casing_table->full_name;
            unless (defined $full_casing_table
                    && ! $full_casing_table->is_empty)
            {
                Carp::my_carp_bug("Need to process UnicodeData before SpecialCasing.  Only special casing will be generated.");
            }

            $full_casing_table->add_comment(join_lines( <<END
This file includes both the simple and full case changing maps.  The simple
ones are in the main body of the table below, and the full ones adding to or
overriding them are in the hash.
END
            ));

            # The simple version's name in each mapping merely has an 's' in
            # front of the full one's
            my $simple_name = 's' . $full_casing_name;
            my $simple = property_ref($simple_name);
            $simple->initialize($full_casing_table) if $simple->to_output_map();
        }

        return;
    }

    sub filter_2_1_8_special_casing_line {

        # This version had duplicate entries in this file.  Delete all but the
        # first one
        my @fields = split /\s*;\s*/, $_, -1; # -1 => retain trailing null
                                              # fields
        if (exists $special_casing_code_points{$fields[0]}) {
            $_ = "";
            return;
        }

        $special_casing_code_points{$fields[0]} = 1;
        filter_special_casing_line(@_);
    }

    sub filter_special_casing_line($file) {
        # Change the format of $_ from SpecialCasing.txt into something that
        # the generic handler understands.  Each input line contains three
        # case mappings.  This will generate three lines to pass to the
        # generic handler for each of those.

        # The input syntax (after stripping comments and trailing white space
        # is like one of the following (with the final two being entries that
        # we ignore):
        # 00DF; 00DF; 0053 0073; 0053 0053; # LATIN SMALL LETTER SHARP S
        # 03A3; 03C2; 03A3; 03A3; Final_Sigma;
        # 0307; ; 0307; 0307; tr After_I; # COMBINING DOT ABOVE
        # Note the trailing semi-colon, unlike many of the input files.  That
        # means that there will be an extra null field generated by the split

        my @fields = split /\s*;\s*/, $_, -1; # -1 => retain trailing null
                                              # fields

        # field #4 is when this mapping is conditional.  If any of these get
        # implemented, it would be by hard-coding in the casing functions in
        # the Perl core, not through tables.  But if there is a new condition
        # we don't know about, output a warning.  We know about all the
        # conditions through 6.0
        if ($fields[4] ne "") {
            my @conditions = split ' ', $fields[4];
            if ($conditions[0] ne 'tr'  # We know that these languages have
                                        # conditions, and some are multiple
                && $conditions[0] ne 'az'
                && $conditions[0] ne 'lt'

                # And, we know about a single condition Final_Sigma, but
                # nothing else.
                && ($v_version gt v5.2.0
                    && (@conditions > 1 || $conditions[0] ne 'Final_Sigma')))
            {
                $file->carp_bad_line("Unknown condition '$fields[4]'.  You should inspect it and either add code to handle it, or add to list of those that are to ignore");
            }
            elsif ($conditions[0] ne 'Final_Sigma') {

                    # Don't print out a message for Final_Sigma, because we
                    # have hard-coded handling for it.  (But the standard
                    # could change what the rule should be, but it wouldn't
                    # show up here anyway.

                    print "# SKIPPING Special Casing: $_\n"
                                                    if $verbosity >= $VERBOSE;
            }
            $_ = "";
            return;
        }
        elsif (@fields > 6 || (@fields == 6 && $fields[5] ne "" )) {
            $file->carp_bad_line('Extra fields');
            $_ = "";
            return;
        }

        my $decimal_code_point = hex $fields[0];

        # Loop to handle each of the three mappings in the input line, in
        # order, with $i indicating the current field number.
        my $i = 0;
        for my $object ($lc, $tc, $uc) {
            $i++;   # First time through, $i = 0 ... 3rd time = 3

            my $value = $object->value_of($decimal_code_point);
            $value = ($value eq $CODE_POINT)
                      ? $decimal_code_point
                      : hex $value;

            # If this isn't a multi-character mapping, it should already have
            # been read in.
            if ($fields[$i] !~ / /) {
                if ($value != hex $fields[$i]) {
                    Carp::my_carp("Bad news. UnicodeData.txt thinks "
                                  . $object->name
                                  . "(0x$fields[0]) is $value"
                                  . " and SpecialCasing.txt thinks it is "
                                  . hex($fields[$i])
                                  . ".  Good luck.  Retaining UnicodeData value, and proceeding anyway.");
                }
            }
            else {

                # The mapping is additional, beyond the simple mapping.
                $file->insert_adjusted_lines("$fields[0]; "
                                             . $object->name
                                            . "; "
                                            . $CMD_DELIM
                                            . "$REPLACE_CMD=$MULTIPLE_BEFORE"
                                            . $CMD_DELIM
                                            . $fields[$i]);
            }
        }

        # Everything has been handled by the insert_adjusted_lines()
        $_ = "";

        return;
    }
}

sub filter_old_style_case_folding($file) {
    # This transforms $_ containing the case folding style of 3.0.1, to 3.1
    # and later style.  Different letters were used in the earlier.

    my @fields = split /\s*;\s*/;

    if ($fields[1] eq 'L') {
        $fields[1] = 'C';             # L => C always
    }
    elsif ($fields[1] eq 'E') {
        if ($fields[2] =~ / /) {      # E => C if one code point; F otherwise
            $fields[1] = 'F'
        }
        else {
            $fields[1] = 'C'
        }
    }
    else {
        $file->carp_bad_line("Expecting L or E in second field");
        $_ = "";
        return;
    }
    $_ = join("; ", @fields) . ';';
    return;
}

{ # Closure for case folding

    # Create the map for simple only if are going to output it, for otherwise
    # it takes no part in anything we do.
    my $to_output_simple;

    sub setup_case_folding {
        # Read in the case foldings in CaseFolding.txt.  This handles both
        # simple and full case folding.

        $to_output_simple
                        = property_ref('Simple_Case_Folding')->to_output_map;

        if (! $to_output_simple) {
            property_ref('Case_Folding')->set_proxy_for('Simple_Case_Folding');
        }

        # If we ever wanted to show that these tables were combined, a new
        # property method could be created, like set_combined_props()
        property_ref('Case_Folding')->add_comment(join_lines( <<END
This file includes both the simple and full case folding maps.  The simple
ones are in the main body of the table below, and the full ones adding to or
overriding them are in the hash.
END
        ));
        return;
    }

    sub filter_case_folding_line($file) {
        # Called for each line in CaseFolding.txt
        # Input lines look like:
        # 0041; C; 0061; # LATIN CAPITAL LETTER A
        # 00DF; F; 0073 0073; # LATIN SMALL LETTER SHARP S
        # 1E9E; S; 00DF; # LATIN CAPITAL LETTER SHARP S
        #
        # 'C' means that folding is the same for both simple and full
        # 'F' that it is only for full folding
        # 'S' that it is only for simple folding
        # 'T' is locale-dependent, and ignored
        # 'I' is a type of 'F' used in some early releases.
        # Note the trailing semi-colon, unlike many of the input files.  That
        # means that there will be an extra null field generated by the split
        # below, which we ignore and hence is not an error.

        my ($range, $type, $map, @remainder) = split /\s*;\s*/, $_, -1;
        if (@remainder > 1 || (@remainder == 1 && $remainder[0] ne "" )) {
            $file->carp_bad_line('Extra fields');
            $_ = "";
            return;
        }

        if ($type =~ / ^ [IT] $/x) {   # Skip Turkic case folding, is locale dependent
            $_ = "";
            return;
        }

        # C: complete, F: full, or I: dotted uppercase I -> dotless lowercase
        # I are all full foldings; S is single-char.  For S, there is always
        # an F entry, so we must allow multiple values for the same code
        # point.  Fortunately this table doesn't need further manipulation
        # which would preclude using multiple-values.  The S is now included
        # so that _swash_inversion_hash() is able to construct closures
        # without having to worry about F mappings.
        if ($type eq 'C' || $type eq 'F' || $type eq 'I' || $type eq 'S') {
            $_ = "$range; Case_Folding; "
                 . "$CMD_DELIM$REPLACE_CMD=$MULTIPLE_BEFORE$CMD_DELIM$map";
        }
        else {
            $_ = "";
            $file->carp_bad_line('Expecting C F I S or T in second field');
        }

        # C and S are simple foldings, but simple case folding is not needed
        # unless we explicitly want its map table output.
        if ($to_output_simple && $type eq 'C' || $type eq 'S') {
            $file->insert_adjusted_lines("$range; Simple_Case_Folding; $map");
        }

        return;
    }

} # End case fold closure

sub filter_jamo_line {
    # Filter Jamo.txt lines.  This routine mainly is used to populate hashes
    # from this file that is used in generating the Name property for Jamo
    # code points.  But, it also is used to convert early versions' syntax
    # into the modern form.  Here are two examples:
    # 1100; G   # HANGUL CHOSEONG KIYEOK            # Modern syntax
    # U+1100; G; HANGUL CHOSEONG KIYEOK             # 2.0 syntax
    #
    # The input is $_, the output is $_ filtered.

    my @fields = split /\s*;\s*/, $_, -1;  # -1 => retain trailing null fields

    # Let the caller handle unexpected input.  In earlier versions, there was
    # a third field which is supposed to be a comment, but did not have a '#'
    # before it.
    return if @fields > (($v_version gt v3.0.0) ? 2 : 3);

    $fields[0] =~ s/^U\+//;     # Also, early versions had this extraneous
                                # beginning.

    # Some 2.1 versions had this wrong.  Causes havoc with the algorithm.
    $fields[1] = 'R' if $fields[0] eq '1105';

    # Add to structure so can generate Names from it.
    my $cp = hex $fields[0];
    my $short_name = $fields[1];
    $Jamo{$cp} = $short_name;
    if ($cp <= $LBase + $LCount) {
        $Jamo_L{$short_name} = $cp - $LBase;
    }
    elsif ($cp <= $VBase + $VCount) {
        $Jamo_V{$short_name} = $cp - $VBase;
    }
    elsif ($cp <= $TBase + $TCount) {
        $Jamo_T{$short_name} = $cp - $TBase;
    }
    else {
        Carp::my_carp_bug("Unexpected Jamo code point in $_");
    }


    # Reassemble using just the first two fields to look like a typical
    # property file line
    $_ = "$fields[0]; $fields[1]";

    return;
}

sub register_fraction($rational) {
    # This registers the input rational number so that it can be passed on to
    # Unicode::UCD, both in rational and floating forms.

    my $floating = eval $rational;

    my @floats = sprintf "%.*e", $E_FLOAT_PRECISION, $floating;

    # See if the denominator is a power of 2.
    $rational =~ m!.*/(.*)!;
    my $denominator = $1;
    if (defined $denominator && (($denominator & ($denominator - 1)) == 0)) {

        # Here the denominator is a power of 2.  This means it has an exact
        # representation in binary, so rounding could go either way.  It turns
        # out that Windows doesn't necessarily round towards even, so output
        # an extra entry.  This happens when the final digit we output is even
        # and the next digits would be 50* to the precision of the machine.
        my $extra_digit_float = sprintf "%e", $floating;
        my $q = $E_FLOAT_PRECISION - 1;
        if ($extra_digit_float =~ / ( .* \. \d{$q} )
                                    ( [02468] ) 5 0* ( e .*)
                                  /ix)
        {
            push @floats, $1 . ($2 + 1) . $3;
        }
    }

    foreach my $float (@floats) {
        # Strip off any leading zeros beyond 2 digits to make it C99
        # compliant.  (Windows has 3 digit exponents, contrary to C99)
        $float =~ s/ ( .* e [-+] ) 0* ( \d{2,}? ) /$1$2/x;

        if (   defined $nv_floating_to_rational{$float}
            && $nv_floating_to_rational{$float} ne $rational)
        {
            die Carp::my_carp_bug("Both '$rational' and"
                            . " '$nv_floating_to_rational{$float}' evaluate to"
                            . " the same floating point number."
                            . "  \$E_FLOAT_PRECISION must be increased");
        }
        $nv_floating_to_rational{$float} = $rational;
    }
    return;
}

sub gcd($a, $b) {   # Greatest-common-divisor; from
                # http://en.wikipedia.org/wiki/Euclidean_algorithm
    use integer;

    while ($b != 0) {
       my $temp = $b;
       $b = $a % $b;
       $a = $temp;
    }
    return $a;
}

sub reduce_fraction($fraction_ref) {
    # Reduce a fraction to lowest terms.  The Unicode data may be reducible,
    # hence this is needed.  The argument is a reference to the
    # string denoting the fraction, which must be of the form:
    if ($$fraction_ref !~ / ^ (-?) (\d+) \/ (\d+) $ /ax) {
        Carp::my_carp_bug("Non-fraction input '$$fraction_ref'.  Unchanged");
        return;
    }

    my $sign = $1;
    my $numerator = $2;
    my $denominator = $3;

    use integer;

    # Find greatest common divisor
    my $gcd = gcd($numerator, $denominator);

    # And reduce using the gcd.
    if ($gcd != 1) {
        $numerator    /= $gcd;
        $denominator  /= $gcd;
        $$fraction_ref = "$sign$numerator/$denominator";
    }

    return;
}

sub filter_numeric_value_line($file) {
    # DNumValues contains lines of a different syntax than the typical
    # property file:
    # 0F33          ; -0.5 ; ; -1/2 # No       TIBETAN DIGIT HALF ZERO
    #
    # This routine transforms $_ containing the anomalous syntax to the
    # typical, by filtering out the extra columns, and convert early version
    # decimal numbers to strings that look like rational numbers.

    # Starting in 5.1, there is a rational field.  Just use that, omitting the
    # extra columns.  Otherwise convert the decimal number in the second field
    # to a rational, and omit extraneous columns.
    my @fields = split /\s*;\s*/, $_, -1;
    my $rational;

    if ($v_version ge v5.1.0) {
        if (@fields != 4) {
            $file->carp_bad_line('Not 4 semi-colon separated fields');
            $_ = "";
            return;
        }
        reduce_fraction(\$fields[3]) if $fields[3] =~ qr{/};
        $rational = $fields[3];

        $_ = join '; ', @fields[ 0, 3 ];
    }
    else {

        # Here, is an older Unicode file, which has decimal numbers instead of
        # rationals in it.  Use the fraction to calculate the denominator and
        # convert to rational.

        if (@fields != 2 && @fields != 3) {
            $file->carp_bad_line('Not 2 or 3 semi-colon separated fields');
            $_ = "";
            return;
        }

        my $codepoints = $fields[0];
        my $decimal = $fields[1];
        if ($decimal =~ s/\.0+$//) {

            # Anything ending with a decimal followed by nothing but 0's is an
            # integer
            $_ = "$codepoints; $decimal";
            $rational = $decimal;
        }
        else {

            my $denominator;
            if ($decimal =~ /\.50*$/) {
                $denominator = 2;
            }

            # Here have the hardcoded repeating decimals in the fraction, and
            # the denominator they imply.  There were only a few denominators
            # in the older Unicode versions of this file which this code
            # handles, so it is easy to convert them.

            # The 4 is because of a round-off error in the Unicode 3.2 files
            elsif ($decimal =~ /\.33*[34]$/ || $decimal =~ /\.6+7$/) {
                $denominator = 3;
            }
            elsif ($decimal =~ /\.[27]50*$/) {
                $denominator = 4;
            }
            elsif ($decimal =~ /\.[2468]0*$/) {
                $denominator = 5;
            }
            elsif ($decimal =~ /\.16+7$/ || $decimal =~ /\.83+$/) {
                $denominator = 6;
            }
            elsif ($decimal =~ /\.(12|37|62|87)50*$/) {
                $denominator = 8;
            }
            if ($denominator) {
                my $sign = ($decimal < 0) ? "-" : "";
                my $numerator = int((abs($decimal) * $denominator) + .5);
                $rational = "$sign$numerator/$denominator";
                $_ = "$codepoints; $rational";
            }
            else {
                $file->carp_bad_line("Can't cope with number '$decimal'.");
                $_ = "";
                return;
            }
        }
    }

    register_fraction($rational) if $rational =~ qr{/};
    return;
}

{ # Closure
    my %unihan_properties;

    sub construct_unihan($file_object) {

        return unless file_exists($file_object->file);

        if ($v_version lt v4.0.0) {
            push @cjk_properties, 'URS ; Unicode_Radical_Stroke';
            push @cjk_property_values, split "\n", <<'END';
# @missing: 0000..10FFFF; Unicode_Radical_Stroke; <none>
END
        }

        if ($v_version ge v3.0.0) {
            push @cjk_properties, split "\n", <<'END';
cjkIRG_GSource; kIRG_GSource
cjkIRG_JSource; kIRG_JSource
cjkIRG_KSource; kIRG_KSource
cjkIRG_TSource; kIRG_TSource
cjkIRG_VSource; kIRG_VSource
END
        push @cjk_property_values, split "\n", <<'END';
# @missing: 0000..10FFFF; cjkIRG_GSource; <none>
# @missing: 0000..10FFFF; cjkIRG_JSource; <none>
# @missing: 0000..10FFFF; cjkIRG_KSource; <none>
# @missing: 0000..10FFFF; cjkIRG_TSource; <none>
# @missing: 0000..10FFFF; cjkIRG_VSource; <none>
END
        }
        if ($v_version ge v3.1.0) {
            push @cjk_properties, 'cjkIRG_HSource; kIRG_HSource';
            push @cjk_property_values, '# @missing: 0000..10FFFF; cjkIRG_HSource; <none>';
        }
        if ($v_version ge v3.1.1) {
            push @cjk_properties, 'cjkIRG_KPSource; kIRG_KPSource';
            push @cjk_property_values, '# @missing: 0000..10FFFF; cjkIRG_KPSource; <none>';
        }
        if ($v_version ge v3.2.0) {
            push @cjk_properties, split "\n", <<'END';
cjkAccountingNumeric; kAccountingNumeric
cjkCompatibilityVariant; kCompatibilityVariant
cjkOtherNumeric; kOtherNumeric
cjkPrimaryNumeric; kPrimaryNumeric
END
            push @cjk_property_values, split "\n", <<'END';
# @missing: 0000..10FFFF; cjkAccountingNumeric; NaN
# @missing: 0000..10FFFF; cjkCompatibilityVariant; <code point>
# @missing: 0000..10FFFF; cjkOtherNumeric; NaN
# @missing: 0000..10FFFF; cjkPrimaryNumeric; NaN
END
        }
        if ($v_version gt v4.0.0) {
            push @cjk_properties, 'cjkIRG_USource; kIRG_USource';
            push @cjk_property_values, '# @missing: 0000..10FFFF; cjkIRG_USource; <none>';
        }

        if ($v_version ge v4.1.0) {
            push @cjk_properties, 'cjkIICore ; kIICore';
            push @cjk_property_values, '# @missing: 0000..10FFFF; cjkIICore; <none>';
        }
    }

    sub setup_unihan {
        # Do any special setup for Unihan properties.

        # This property gives the wrong computed type, so override.
        my $usource = property_ref('kIRG_USource');
        $usource->set_type($STRING) if defined $usource;

        # This property is to be considered binary (it says so in
        # http://www.unicode.org/reports/tr38/)
        my $iicore = property_ref('kIICore');
        if (defined $iicore) {
            $iicore->set_type($FORCED_BINARY);
            $iicore->table("Y")->add_note("Matches any code point which has a non-null value for this property; see unicode.org UAX #38.");

            # Unicode doesn't include the maps for this property, so don't
            # warn that they are missing.
            $iicore->set_pre_declared_maps(0);
            $iicore->add_comment(join_lines( <<END
This property contains string values, but any non-empty ones are considered to
be 'core', so Perl creates tables for both: 1) its string values, plus 2)
tables so that \\p{kIICore} matches any code point which has a non-empty
value for this property.
END
            ));
        }

        return;
    }

    sub filter_unihan_line {
        # Change unihan db lines to look like the others in the db.  Here is
        # an input sample:
        #   U+341C        kCangjie        IEKN

        # Tabs are used instead of semi-colons to separate fields; therefore
        # they may have semi-colons embedded in them.  Change these to periods
        # so won't screw up the rest of the code.
        s/;/./g;

        # Remove lines that don't look like ones we accept.
        if ($_ !~ /^ [^\t]* \t ( [^\t]* ) /x) {
            $_ = "";
            return;
        }

        # Extract the property, and save a reference to its object.
        my $property = $1;
        if (! exists $unihan_properties{$property}) {
            $unihan_properties{$property} = property_ref($property);
        }

        # Don't do anything unless the property is one we're handling, which
        # we determine by seeing if there is an object defined for it or not
        if (! defined $unihan_properties{$property}) {
            $_ = "";
            return;
        }

        # Convert the tab separators to our standard semi-colons, and convert
        # the U+HHHH notation to the rest of the standard's HHHH
        s/\t/;/g;
        s/\b U \+ (?= $code_point_re )//xg;

        #local $to_trace = 1 if main::DEBUG;
        trace $_ if main::DEBUG && $to_trace;

        return;
    }
}

sub filter_blocks_lines($file) {
    # In the Blocks.txt file, the names of the blocks don't quite match the
    # names given in PropertyValueAliases.txt, so this changes them so they
    # do match:  Blanks and hyphens are changed into underscores.  Also makes
    # early release versions look like later ones
    #
    # $_ is transformed to the correct value.

    if ($v_version lt v3.2.0) {
        if (/FEFF.*Specials/) { # Bug in old versions: line wrongly inserted
            $_ = "";
            return;
        }

        # Old versions used a different syntax to mark the range.
        $_ =~ s/;\s+/../ if $v_version lt v3.1.0;
    }

    my @fields = split /\s*;\s*/, $_, -1;
    if (@fields != 2) {
        $file->carp_bad_line("Expecting exactly two fields");
        $_ = "";
        return;
    }

    # Change hyphens and blanks in the block name field only
    $fields[1] =~ s/[ -]/_/g;
    $fields[1] =~ s/_ ( [a-z] ) /_\u$1/xg;   # Capitalize first letter of word

    $_ = join("; ", @fields);
    return;
}

{ # Closure
    my $current_property;

    sub filter_old_style_proplist {
        # PropList.txt has been in Unicode since version 2.0.  Until 3.1, it
        # was in a completely different syntax.  Ken Whistler of Unicode says
        # that it was something he used as an aid for his own purposes, but
        # was never an official part of the standard.  Many of the properties
        # in it were incorporated into the later PropList.txt, but some were
        # not.  This program uses this early file to generate property tables
        # that are otherwise not accessible in the early UCD's.  It does this
        # for the ones that eventually became official, and don't appear to be
        # too different in their contents from the later official version, and
        # throws away the rest.  It could be argued that the ones it generates
        # were probably not really official at that time, so should be
        # ignored.  You can easily modify things to skip all of them by
        # changing this function to just set $_ to "", and return; and to skip
        # certain of them by simply removing their declarations from
        # get_old_property_aliases().
        #
        # Here is a list of all the ones that are thrown away:
        #   Alphabetic                   The definitions for this are very
        #                                defective, so better to not mislead
        #                                people into thinking it works.
        #                                Instead the Perl extension of the
        #                                same name is constructed from first
        #                                principles.
        #   Bidi=*                       duplicates UnicodeData.txt
        #   Combining                    never made into official property;
        #                                is \P{ccc=0}
        #   Composite                    never made into official property.
        #   Currency Symbol              duplicates UnicodeData.txt: gc=sc
        #   Decimal Digit                duplicates UnicodeData.txt: gc=nd
        #   Delimiter                    never made into official property;
        #                                removed in 3.0.1
        #   Format Control               never made into official property;
        #                                similar to gc=cf
        #   High Surrogate               duplicates Blocks.txt
        #   Ignorable Control            never made into official property;
        #                                similar to di=y
        #   ISO Control                  duplicates UnicodeData.txt: gc=cc
        #   Left of Pair                 never made into official property;
        #   Line Separator               duplicates UnicodeData.txt: gc=zl
        #   Low Surrogate                duplicates Blocks.txt
        #   Non-break                    was actually listed as a property
        #                                in 3.2, but without any code
        #                                points.  Unicode denies that this
        #                                was ever an official property
        #   Non-spacing                  duplicate UnicodeData.txt: gc=mn
        #   Numeric                      duplicates UnicodeData.txt: gc=cc
        #   Paired Punctuation           never made into official property;
        #                                appears to be gc=ps + gc=pe
        #   Paragraph Separator          duplicates UnicodeData.txt: gc=cc
        #   Private Use                  duplicates UnicodeData.txt: gc=co
        #   Private Use High Surrogate   duplicates Blocks.txt
        #   Punctuation                  duplicates UnicodeData.txt: gc=p
        #   Space                        different definition than eventual
        #                                one.
        #   Titlecase                    duplicates UnicodeData.txt: gc=lt
        #   Unassigned Code Value        duplicates UnicodeData.txt: gc=cn
        #   Zero-width                   never made into official property;
        #                                subset of gc=cf
        # Most of the properties have the same names in this file as in later
        # versions, but a couple do not.
        #
        # This subroutine filters $_, converting it from the old style into
        # the new style.  Here's a sample of the old-style
        #
        #   *******************************************
        #
        #   Property dump for: 0x100000A0 (Join Control)
        #
        #   200C..200D  (2 chars)
        #
        # In the example, the property is "Join Control".  It is kept in this
        # closure between calls to the subroutine.  The numbers beginning with
        # 0x were internal to Ken's program that generated this file.

        # If this line contains the property name, extract it.
        if (/^Property dump for: [^(]*\((.*)\)/) {
            $_ = $1;

            # Convert white space to underscores.
            s/ /_/g;

            # Convert the few properties that don't have the same name as
            # their modern counterparts
            s/Identifier_Part/ID_Continue/
            or s/Not_a_Character/NChar/;

            # If the name matches an existing property, use it.
            if (defined property_ref($_)) {
                trace "new property=", $_ if main::DEBUG && $to_trace;
                $current_property = $_;
            }
            else {        # Otherwise discard it
                trace "rejected property=", $_ if main::DEBUG && $to_trace;
                undef $current_property;
            }
            $_ = "";    # The property is saved for the next lines of the
                        # file, but this defining line is of no further use,
                        # so clear it so that the caller won't process it
                        # further.
        }
        elsif (! defined $current_property || $_ !~ /^$code_point_re/) {

            # Here, the input line isn't a header defining a property for the
            # following section, and either we aren't in such a section, or
            # the line doesn't look like one that defines the code points in
            # such a section.  Ignore this line.
            $_ = "";
        }
        else {

            # Here, we have a line defining the code points for the current
            # stashed property.  Anything starting with the first blank is
            # extraneous.  Otherwise, it should look like a normal range to
            # the caller.  Append the property name so that it looks just like
            # a modern PropList entry.

            $_ =~ s/\s.*//;
            $_ .= "; $current_property";
        }
        trace $_ if main::DEBUG && $to_trace;
        return;
    }
} # End closure for old style proplist

sub filter_old_style_normalization_lines {
    # For early releases of Unicode, the lines were like:
    #        74..2A76    ; NFKD_NO
    # For later releases this became:
    #        74..2A76    ; NFKD_QC; N
    # Filter $_ to look like those in later releases.
    # Similarly for MAYBEs

    s/ _NO \b /_QC; N/x || s/ _MAYBE \b /_QC; M/x;

    # Also, the property FC_NFKC was abbreviated to FNC
    s/FNC/FC_NFKC/;
    return;
}

sub setup_script_extensions {
    # The Script_Extensions property starts out with a clone of the Script
    # property.

    $scx = property_ref("Script_Extensions");
    return unless defined $scx;

    $scx->_set_format($STRING_WHITE_SPACE_LIST);
    $scx->initialize($script);
    $scx->set_default_map($script->default_map);
    $scx->set_pre_declared_maps(0);     # PropValueAliases doesn't list these
    $scx->add_comment(join_lines( <<END
The values for code points that appear in one script are just the same as for
the 'Script' property.  Likewise the values for those that appear in many
scripts are either 'Common' or 'Inherited', same as with 'Script'.  But the
values of code points that appear in a few scripts are a space separated list
of those scripts.
END
    ));

    # Initialize scx's tables and the aliases for them to be the same as sc's
    foreach my $table ($script->tables) {
        my $scx_table = $scx->add_match_table($table->name,
                                Full_Name => $table->full_name);
        foreach my $alias ($table->aliases) {
            $scx_table->add_alias($alias->name);
        }
    }
}

sub  filter_script_extensions_line {
    # The Scripts file comes with the full name for the scripts; the
    # ScriptExtensions, with the short name.  The final mapping file is a
    # combination of these, and without adjustment, would have inconsistent
    # entries.  This filters the latter file to convert to full names.
    # Entries look like this:
    # 064B..0655    ; Arab Syrc # Mn  [11] ARABIC FATHATAN..ARABIC HAMZA BELOW

    my @fields = split /\s*;\s*/;

    # This script was erroneously omitted in this Unicode version.
    $fields[1] .= ' Takr' if $v_version eq v6.1.0 && $fields[0] =~ /^0964/;

    my @full_names;
    foreach my $short_name (split " ", $fields[1]) {
        push @full_names, $script->table($short_name)->full_name;
    }
    $fields[1] = join " ", @full_names;
    $_ = join "; ", @fields;

    return;
}

sub setup_emojidata {
    my $prop_ref = Property->new('ExtPict',
                                 Full_Name => 'Extended_Pictographic',
    );
    $prop_ref->set_fate($PLACEHOLDER,
                        "Not part of the Unicode Character Database");
}

sub filter_emojidata_line {
    # We only are interested in this single property from this non-UCD data
    # file, and we turn it into a Perl property, so that it isn't accessible
    # to the users

    $_ = "" unless /\bExtended_Pictographic\b/;

    return;
}

sub setup_IdStatus {
    my $ids = Property->new('Identifier_Status',
                            Match_SubDir => 'IdStatus',
                            Default_Map => 'Restricted',
                           );
    $ids->add_match_table('Allowed');
}

sub setup_IdType {
    $idt = Property->new('Identifier_Type',
                            Match_SubDir => 'IdType',
                            Default_Map => 'Not_Character',
                            Format => $STRING_WHITE_SPACE_LIST,
                           );
}

sub  filter_IdType_line {

    # Some code points have more than one type, separated by spaces on the
    # input.  For now, we just add everything as a property value.  Later when
    # we look for properties with format $STRING_WHITE_SPACE_LIST, we resolve
    # things

    my @fields = split /\s*;\s*/;
    my $types = $fields[1];
    $idt->add_match_table($types) unless defined $idt->table($types);

    return;
}

sub generate_hst($file) {

    # Populates the Hangul Syllable Type property from first principles

    # These few ranges are hard-coded in.
    $file->insert_lines(split /\n/, <<'END'
1100..1159    ; L
115F          ; L
1160..11A2    ; V
11A8..11F9    ; T
END
);

    # The Hangul syllables in version 1 are at different code points than
    # those that came along starting in version 2, and have different names;
    # they comprise about 60% of the code points of the later version.
    # From my (khw) research on them (see <558493EB.4000807@att.net>), the
    # initial set is a subset of the later version, with different English
    # transliterations.  I did not see an easy mapping between them.  The
    # later set includes essentially all possibilities, even ones that aren't
    # in modern use (if they ever were), and over 96% of the new ones are type
    # LVT.  Mathematically, the early set must also contain a preponderance of
    # LVT values.  In lieu of doing nothing, we just set them all to LVT, and
    # expect that this will be right most of the time, which is better than
    # not being right at all.
    if ($v_version lt v2.0.0) {
        my $property = property_ref($file->property);
        $file->insert_lines(sprintf("%04X..%04X; LVT\n",
                                    $FIRST_REMOVED_HANGUL_SYLLABLE,
                                    $FINAL_REMOVED_HANGUL_SYLLABLE));
        push @tables_that_may_be_empty, $property->table('LV')->complete_name;
        return;
    }

    # The algorithmically derived syllables are almost all LVT ones, so
    # initialize the whole range with that.
    $file->insert_lines(sprintf "%04X..%04X; LVT\n",
                        $SBase, $SBase + $SCount -1);

    # Those ones that aren't LVT are LV, and they occur at intervals of
    # $TCount code points, starting with the first code point, at $SBase.
    for (my $i = $SBase; $i < $SBase + $SCount; $i += $TCount) {
        $file->insert_lines(sprintf "%04X..%04X; LV\n", $i, $i);
    }

    return;
}

sub generate_GCB($file) {

    # Populates the Grapheme Cluster Break property from first principles

    # All these definitions are from
    # http://www.unicode.org/reports/tr29/tr29-3.html with confirmation
    # from http://www.unicode.org/reports/tr29/tr29-4.html

    foreach my $range ($gc->ranges) {

        # Extend includes gc=Me and gc=Mn, while Control includes gc=Cc
        # and gc=Cf
        if ($range->value =~ / ^ M [en] $ /x) {
            $file->insert_lines(sprintf "%04X..%04X; Extend",
                                $range->start,  $range->end);
        }
        elsif ($range->value =~ / ^ C [cf] $ /x) {
            $file->insert_lines(sprintf "%04X..%04X; Control",
                                $range->start,  $range->end);
        }
    }
    $file->insert_lines("2028; Control"); # Line Separator
    $file->insert_lines("2029; Control"); # Paragraph Separator

    $file->insert_lines("000D; CR");
    $file->insert_lines("000A; LF");

    # Also from http://www.unicode.org/reports/tr29/tr29-3.html.
    foreach my $code_point ( qw{
                                09BE 09D7 0B3E 0B57 0BBE 0BD7 0CC2 0CD5 0CD6
                                0D3E 0D57 0DCF 0DDF FF9E FF9F 1D165 1D16E 1D16F
                                }
    ) {
        my $category = $gc->value_of(hex $code_point);
        next if ! defined $category || $category eq 'Cn'; # But not if
                                                          # unassigned in this
                                                          # release
        $file->insert_lines("$code_point; Extend");
    }

    my $hst = property_ref('Hangul_Syllable_Type');
    if ($hst->count > 0) {
        foreach my $range ($hst->ranges) {
            $file->insert_lines(sprintf "%04X..%04X; %s",
                                    $range->start, $range->end, $range->value);
        }
    }
    else {
        generate_hst($file);
    }

    main::process_generic_property_file($file);
}


sub fixup_early_perl_name_alias($file) {

    # Different versions of Unicode have varying support for the name synonyms
    # below.  Just include everything.  As of 6.1, all these are correct in
    # the Unicode-supplied file.

    # ALERT did not come along until 6.0, at which point it became preferred
    # over BELL.  By inserting it last in early releases, BELL is preferred
    # over it; and vice-vers in 6.0
    my $type_for_bell = ($v_version lt v6.0.0)
               ? 'correction'
               : 'alternate';
    $file->insert_lines(split /\n/, <<END
0007;BELL; $type_for_bell
000A;LINE FEED (LF);alternate
000C;FORM FEED (FF);alternate
000D;CARRIAGE RETURN (CR);alternate
0085;NEXT LINE (NEL);alternate
END

    );

    # One might think that the 'Unicode_1_Name' field, could work for most
    # of the above names, but sadly that field varies depending on the
    # release.  Version 1.1.5 had no names for any of the controls; Version
    # 2.0 introduced names for the C0 controls, and 3.0 introduced C1 names.
    # 3.0.1 removed the name INDEX; and 3.2 changed some names:
    #   changed to parenthesized versions like "NEXT LINE" to
    #       "NEXT LINE (NEL)";
    #   changed PARTIAL LINE DOWN to PARTIAL LINE FORWARD
    #   changed PARTIAL LINE UP to PARTIAL LINE BACKWARD;;
    #   changed e.g. FILE SEPARATOR to INFORMATION SEPARATOR FOUR
    #
    # All these are present in the 6.1 NameAliases.txt

    return;
}

sub filter_later_version_name_alias_line {

    # This file has an extra entry per line for the alias type.  This is
    # handled by creating a compound entry: "$alias: $type";  First, split
    # the line into components.
    my ($range, $alias, $type, @remainder)
        = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields

    # This file contains multiple entries for some components, so tell the
    # downstream code to allow this in our internal tables; the
    # $MULTIPLE_AFTER preserves the input ordering.
    $_ = join ";", $range, $CMD_DELIM
                           . $REPLACE_CMD
                           . '='
                           . $MULTIPLE_AFTER
                           . $CMD_DELIM
                           . "$alias: $type",
                   @remainder;
    return;
}

sub filter_early_version_name_alias_line {

    # Early versions did not have the trailing alias type field; implicitly it
    # was 'correction'.
    $_ .= "; correction";

    filter_later_version_name_alias_line;
    return;
}

sub filter_all_caps_script_names {

    # Some early Unicode releases had the script names in all CAPS.  This
    # converts them to just the first letter of each word being capital.

    my ($range, $script, @remainder)
        = split /\s*;\s*/, $_, -1; # -1 => retain trailing null fields
    my @words = split /[_-]/, $script;
    for my $word (@words) {
        $word =
            ucfirst(lc($word)) if $word ne 'CJK';
    }
    $script = join "_", @words;
    $_ = join ";", $range, $script, @remainder;
}

sub finish_Unicode() {
    # This routine should be called after all the Unicode files have been read
    # in.  It:
    # 1) Creates properties that are missing from the version of Unicode being
    #    compiled, and which, for whatever reason, are needed for the Perl
    #    core to function properly.  These are minimally populated as
    #    necessary.
    # 2) Adds the mappings for code points missing from the files which have
    #    defaults specified for them.
    # 3) At this point all mappings are known, so it computes the type of
    #    each property whose type hasn't been determined yet.
    # 4) Calculates all the regular expression match tables based on the
    #    mappings.
    # 5) Calculates and adds the tables which are defined by Unicode, but
    #    which aren't derived by them, and certain derived tables that Perl
    #    uses.

    # Folding information was introduced later into Unicode data.  To get
    # Perl's case ignore (/i) to work at all in releases that don't have
    # folding, use the best available alternative, which is lower casing.
    my $fold = property_ref('Case_Folding');
    if ($fold->is_empty) {
        $fold->initialize(property_ref('Lowercase_Mapping'));
        $fold->add_note(join_lines(<<END
WARNING: This table uses lower case as a substitute for missing fold
information
END
        ));
    }

    # Multiple-character mapping was introduced later into Unicode data, so it
    # is by default the simple version.  If to output the simple versions and
    # not present, just use the regular (which in these Unicode versions is
    # the simple as well).
    foreach my $map (qw {   Uppercase_Mapping
                            Lowercase_Mapping
                            Titlecase_Mapping
                            Case_Folding
                        } )
    {
        my $comment = <<END;

Note that although the Perl core uses this file, it has the standard values
for code points from U+0000 to U+00FF compiled in, so changing this table will
not change the core's behavior with respect to these code points.  Use
Unicode::Casing to override this table.
END
        if ($map eq 'Case_Folding') {
            $comment .= <<END;
(/i regex matching is not overridable except by using a custom regex engine)
END
        }
        property_ref($map)->add_comment(join_lines($comment));
        my $simple = property_ref("Simple_$map");
        next if ! $simple->is_empty;
        if ($simple->to_output_map) {
            $simple->initialize(property_ref($map));
        }
        else {
            property_ref($map)->set_proxy_for($simple->name);
        }
    }

    # For each property, fill in any missing mappings, and calculate the re
    # match tables.  If a property has more than one missing mapping, the
    # default is a reference to a data structure, and may require data from
    # other properties to resolve.  The sort is used to cause these to be
    # processed last, after all the other properties have been calculated.
    # (Fortunately, the missing properties so far don't depend on each other.)
    foreach my $property
        (sort { (defined $a->default_map && ref $a->default_map) ? 1 : -1 }
        property_ref('*'))
    {
        # $perl has been defined, but isn't one of the Unicode properties that
        # need to be finished up.
        next if $property == $perl;

        # Nor do we need to do anything with properties that aren't going to
        # be output.
        next if $property->fate == $SUPPRESSED;

        # Handle the properties that have more than one possible default
        if (ref $property->default_map) {
            my $default_map = $property->default_map;

            # These properties have stored in the default_map:
            # One or more of:
            #   1)  A default map which applies to all code points in a
            #       certain class
            #   2)  an expression which will evaluate to the list of code
            #       points in that class
            # And
            #   3) the default map which applies to every other missing code
            #      point.
            #
            # Go through each list.
            while (my ($default, $eval) = $default_map->get_next_defaults) {
                last unless defined $eval;

                # Get the class list, and intersect it with all the so-far
                # unspecified code points yielding all the code points
                # in the class that haven't been specified.
                my $list = eval $eval;
                if ($@) {
                    Carp::my_carp("Can't set some defaults for missing code points for $property because eval '$eval' failed with '$@'");
                    last;
                }

                # Narrow down the list to just those code points we don't have
                # maps for yet.
                $list = $list & $property->inverse_list;

                # Add mappings to the property for each code point in the list
                foreach my $range ($list->ranges) {
                    $property->add_map($range->start, $range->end, $default,
                    Replace => $NO);
                }
            }

            # All remaining code points have the other mapping.  Set that up
            # so the normal single-default mapping code will work on them
            $property->set_default_map($default_map->other_default);

            # And fall through to do that
        }

        # We should have enough data now to compute the type of the property.
        my $property_name = $property->name;
        $property->compute_type;
        my $property_type = $property->type;

        next if ! $property->to_create_match_tables;

        # Here want to create match tables for this property

        # The Unicode db always (so far, and they claim into the future) have
        # the default for missing entries in binary properties be 'N' (unless
        # there is a '@missing' line that specifies otherwise)
        if (! defined $property->default_map) {
            if ($property_type == $BINARY) {
                $property->set_default_map('N');
            }
            elsif ($property_type == $ENUM) {
                Carp::my_carp("Property '$property_name doesn't have a default mapping.  Using a fake one");
                $property->set_default_map('XXX This makes sure there is a default map');
            }
        }

        # Add any remaining code points to the mapping, using the default for
        # missing code points.
        my $default_table;
        my $default_map = $property->default_map;
        if ($property_type == $FORCED_BINARY) {

            # A forced binary property creates a 'Y' table that matches all
            # non-default values.  The actual string values are also written out
            # as a map table.  (The default value will almost certainly be the
            # empty string, so the pod glosses over the distinction, and just
            # talks about empty vs non-empty.)
            my $yes = $property->table("Y");
            foreach my $range ($property->ranges) {
                next if $range->value eq $default_map;
                $yes->add_range($range->start, $range->end);
            }
            $property->table("N")->set_complement($yes);
        }
        else {
            if (defined $default_map) {

                # Make sure there is a match table for the default
                if (! defined ($default_table = $property->table($default_map)))
                {
                    $default_table = $property->add_match_table($default_map);
                }

                # And, if the property is binary, the default table will just
                # be the complement of the other table.
                if ($property_type == $BINARY) {
                    my $non_default_table;

                    # Find the non-default table.
                    for my $table ($property->tables) {
                        if ($table == $default_table) {
                            if ($v_version le v5.0.0) {
                                $table->add_alias($_) for qw(N No F False);
                            }
                            next;
                        } elsif ($v_version le v5.0.0) {
                            $table->add_alias($_) for qw(Y Yes T True);
                        }
                        $non_default_table = $table;
                    }
                    $default_table->set_complement($non_default_table);
                }
                else {

                    # This fills in any missing values with the default.  It's
                    # not necessary to do this with binary properties, as the
                    # default is defined completely in terms of the Y table.
                    $property->add_map(0, $MAX_WORKING_CODEPOINT,
                                    $default_map, Replace => $NO);
                }
            }

            # Have all we need to populate the match tables.
            my $maps_should_be_defined = $property->pre_declared_maps;
            foreach my $range ($property->ranges) {
                my $map = $range->value;
                my $table = $property->table($map);
                if (! defined $table) {

                    # Integral and rational property values are not
                    # necessarily defined in PropValueAliases, but whether all
                    # the other ones should be depends on the property.
                    if ($maps_should_be_defined
                        && $map !~ /^ -? \d+ ( \/ \d+ )? $/x)
                    {
                        Carp::my_carp("Table '$property_name=$map' should "
                                    . "have been defined.  Defining it now.")
                    }
                    $table = $property->add_match_table($map);
                }

                next if $table->complement != 0; # Don't need to populate these
                $table->add_range($range->start, $range->end);
            }
        }

        # For Perl 5.6 compatibility, all properties matchable in regexes can
        # have an optional 'Is_' prefix.  This is now done in Unicode::UCD.
        # But warn if this creates a conflict with a (new) Unicode property
        # name, although it appears that Unicode has made a decision never to
        # begin a property name with 'Is_', so this shouldn't happen.
        foreach my $alias ($property->aliases) {
            my $Is_name = 'Is_' . $alias->name;
            if (defined (my $pre_existing = property_ref($Is_name))) {
                Carp::my_carp(<<END
There is already an alias named $Is_name (from " . $pre_existing . "), so
creating one for $property won't work.  This is bad news.  If it is not too
late, get Unicode to back off.  Otherwise go back to the old scheme (findable
from the git blame log for this area of the code that suppressed individual
aliases that conflict with the new Unicode names.  Proceeding anyway.
END
                );
            }
        } # End of loop through aliases for this property


        # Properties that have sets of values for some characters are now
        # converted.  For example, the Script_Extensions property started out
        # as a clone of the Script property.  But processing its data file
        # caused some elements to be replaced with different data.  (These
        # elements were for the Common and Inherited properties.)  This data
        # is a qw() list of all the scripts that the code points in the given
        # range are in.  An example line is:
        #
        # 060C          ; Arab Syrc Thaa # Po       ARABIC COMMA
        #
        # Code executed earlier has created a new match table named "Arab Syrc
        # Thaa" which contains 060C.  (The cloned table started out with this
        # code point mapping to "Common".)  Now we add 060C to each of the
        # Arab, Syrc, and Thaa match tables.  Then we delete the now spurious
        # "Arab Syrc Thaa" match table.  This is repeated for all these tables
        # and ranges.  The map data is retained in the map table for
        # reference, but the spurious match tables are deleted.
        my $format = $property->format;
        if (defined $format && $format eq $STRING_WHITE_SPACE_LIST) {
            foreach my $table ($property->tables) {

                # Space separates the entries which should go in multiple
                # tables
                next unless $table->name =~ /\s/;

                # The list of the entries, hence the names of the tables that
                # everything in this combo table should be added to.
                my @list = split /\s+/, $table->name;

                # Add the entries from the combo table to each individual
                # table
                foreach my $individual (@list) {
                    my $existing_table = $property->table($individual);

                    # This should only be necessary if this particular entry
                    # occurs only in combo with others.
                    $existing_table = $property->add_match_table($individual)
                                                unless defined $existing_table;
                    $existing_table += $table;
                }
                $property->delete_match_table($table);
            }
        }
    } # End of loop through all Unicode properties.

    # Fill in the mappings that Unicode doesn't completely furnish.  First the
    # single letter major general categories.  If Unicode were to start
    # delivering the values, this would be redundant, but better that than to
    # try to figure out if should skip and not get it right.  Ths could happen
    # if a new major category were to be introduced, and the hard-coded test
    # wouldn't know about it.
    # This routine depends on the standard names for the general categories
    # being what it thinks they are, like 'Cn'.  The major categories are the
    # union of all the general category tables which have the same first
    # letters. eg. L = Lu + Lt + Ll + Lo + Lm
    foreach my $minor_table ($gc->tables) {
        my $minor_name = $minor_table->name;
        next if length $minor_name == 1;
        if (length $minor_name != 2) {
            Carp::my_carp_bug("Unexpected general category '$minor_name'.  Skipped.");
            next;
        }

        my $major_name = uc(substr($minor_name, 0, 1));
        my $major_table = $gc->table($major_name);
        $major_table += $minor_table;
    }

    # LC is Ll, Lu, and Lt.  (used to be L& or L_, but PropValueAliases.txt
    # defines it as LC)
    my $LC = $gc->table('LC');
    $LC->add_alias('L_', Status => $DISCOURAGED);   # For backwards...
    $LC->add_alias('L&', Status => $DISCOURAGED);   # compatibility.


    if ($LC->is_empty) { # Assume if not empty that Unicode has started to
                         # deliver the correct values in it
        $LC->initialize($gc->table('Ll') + $gc->table('Lu'));

        # Lt not in release 1.
        if (defined $gc->table('Lt')) {
            $LC += $gc->table('Lt');
            $gc->table('Lt')->set_caseless_equivalent($LC);
        }
    }
    $LC->add_description('[\p{Ll}\p{Lu}\p{Lt}]');

    $gc->table('Ll')->set_caseless_equivalent($LC);
    $gc->table('Lu')->set_caseless_equivalent($LC);

    # Make sure this assumption in perl core code is valid in this Unicode
    # release, with known exceptions
    foreach my $range (property_ref('Numeric-Type')->table('Decimal')->ranges) {
        next if $range->end - $range->start == 9;
        next if $range->start == 0x1D7CE;   # This whole range was added in 3.1
        next if $range->end == 0x19DA && $v_version eq v5.2.0;
        next if $range->end - $range->start < 9 && $v_version le 4.0.0;

        # 2 sequential series of 10 each were added in 16.0
        next if $range->start == 0x116D0 && $range->end == 0x116E3;

        Carp::my_carp("Range $range unexpectedly doesn't contain 10"
                    . " decimal digits.  Code in regcomp.c assumes it does,"
                    . " and will have to be fixed.  Proceeding anyway.");
    }

    # Mark the scx table as the parent of the corresponding sc table for those
    # which are identical.  This causes the pod for the script table to refer
    # to the corresponding scx one.  This is done after everything, so as to
    # wait until the tables are stabilized before checking for equivalency.
    if (defined $scx) {
        if (defined $pod_directory) {
            foreach my $table ($scx->tables) {
                my $plain_sc_equiv = $script->table($table->name);
                if ($table->matches_identically_to($plain_sc_equiv)) {
                    $plain_sc_equiv->set_equivalent_to($table, Related => 1);
                }
            }
        }
    }

    return;
}

sub pre_3_dot_1_Nl () {

    # Return a range list for gc=nl for Unicode versions prior to 3.1, which
    # is when Unicode's became fully usable.  These code points were
    # determined by inspection and experimentation.  gc=nl is important for
    # certain Perl-extension properties that should be available in all
    # releases.

    my $Nl = Range_List->new();
    if (defined (my $official = $gc->table('Nl'))) {
        $Nl += $official;
    }
    else {
        $Nl->add_range(0x2160, 0x2182);
        $Nl->add_range(0x3007, 0x3007);
        $Nl->add_range(0x3021, 0x3029);
    }
    $Nl->add_range(0xFE20, 0xFE23);
    $Nl->add_range(0x16EE, 0x16F0) if $v_version ge v3.0.0; # 3.0 was when
                                                            # these were added
    return $Nl;
}

sub calculate_Assigned() {  # Set $Assigned to the gc != Cn code points; may be
                            # called before the Cn's are completely filled.
                            # Works on Unicodes earlier than ones that
                            # explicitly specify Cn.
    return if defined $Assigned;

    if (! defined $gc || $gc->is_empty()) {
        Carp::my_carp_bug("calculate_Assigned() called before $gc is populated");
    }

    $Assigned = $perl->add_match_table('Assigned',
                                Description  => "All assigned code points",
                                );
    while (defined (my $range = $gc->each_range())) {
        my $standard_value = standardize($range->value);
        next if $standard_value eq 'cn' || $standard_value eq 'unassigned';
        $Assigned->add_range($range->start, $range->end);
    }
}

sub calculate_DI() {    # Set $DI to a Range_List equivalent to the
                        # Default_Ignorable_Code_Point property.  Works on
                        # Unicodes earlier than ones that explicitly specify
                        # DI.
    return if defined $DI;

    if (defined (my $di = property_ref('Default_Ignorable_Code_Point'))) {
        $DI = $di->table('Y');
    }
    else {
        $DI = Range_List->new(Initialize => [ 0x180B .. 0x180D,
                                              0x2060 .. 0x206F,
                                              0xFE00 .. 0xFE0F,
                                              0xFFF0 .. 0xFFFB,
                                            ]);
        if ($v_version ge v2.0) {
            $DI += $gc->table('Cf')
                +  $gc->table('Cs');

            # These are above the Unicode version 1 max
            $DI->add_range(0xE0000, 0xE0FFF);
        }
        $DI += $gc->table('Cc')
             - ord("\t")
             - utf8::unicode_to_native(0x0A)  # LINE FEED
             - utf8::unicode_to_native(0x0B)  # VERTICAL TAB
             - ord("\f")
             - utf8::unicode_to_native(0x0D)  # CARRIAGE RETURN
             - utf8::unicode_to_native(0x85); # NEL
    }
}

sub calculate_NChar() {  # Create a Perl extension match table which is the
                         # same as the Noncharacter_Code_Point property, and
                         # set $NChar to point to it.  Works on Unicodes
                         # earlier than ones that explicitly specify NChar
    return if defined $NChar;

    $NChar = $perl->add_match_table('_Perl_Nchar',
                                    Perl_Extension => 1,
                                    Fate => $INTERNAL_ONLY);
    if (defined (my $off_nchar = property_ref('NChar'))) {
        $NChar->initialize($off_nchar->table('Y'));
    }
    else {
        $NChar->initialize([ 0xFFFE .. 0xFFFF ]);
        if ($v_version ge v2.0) {   # First release with these nchars
            for (my $i = 0x1FFFE; $i <= 0x10FFFE; $i += 0x10000) {
                $NChar += [ $i .. $i+1 ];
            }
        }
    }
}

sub handle_compare_versions () {
    # This fixes things up for the $compare_versions capability, where we
    # compare Unicode version X with version Y (with Y > X), and we are
    # running it on the Unicode Data for version Y.
    #
    # It works by calculating the code points whose meaning has been specified
    # after release X, by using the Age property.  The complement of this set
    # is the set of code points whose meaning is unchanged between the
    # releases.  This is the set the program restricts itself to.  It includes
    # everything whose meaning has been specified by the time version X came
    # along, plus those still unassigned by the time of version Y.  (We will
    # continue to use the word 'assigned' to mean 'meaning has been
    # specified', as it's shorter and is accurate in all cases except the
    # Noncharacter code points.)
    #
    # This function is run after all the properties specified by Unicode have
    # been calculated for release Y.  This makes sure we get all the nuances
    # of Y's rules.  (It is done before the Perl extensions are calculated, as
    # those are based entirely on the Unicode ones.)  But doing it after the
    # Unicode table calculations means we have to fix up the Unicode tables.
    # We do this by subtracting the code points that have been assigned since
    # X (which is actually done by ANDing each table of assigned code points
    # with the set of unchanged code points).  Most Unicode properties are of
    # the form such that all unassigned code points have a default, grab-bag,
    # property value which is changed when the code point gets assigned.  For
    # these, we just remove the changed code points from the table for the
    # latter property value, and add them back in to the grab-bag one.  A few
    # other properties are not entirely of this form and have values for some
    # or all unassigned code points that are not the grab-bag one.  These have
    # to be handled specially, and are hard-coded in to this routine based on
    # manual inspection of the Unicode character database.  A list of the
    # outlier code points is made for each of these properties, and those
    # outliers are excluded from adding and removing from tables.
    #
    # Note that there are glitches when comparing against Unicode 1.1, as some
    # Hangul syllables in it were later ripped out and eventually replaced
    # with other things.

    print "Fixing up for version comparison\n" if $verbosity >= $PROGRESS;

    my $after_first_version = "All matching code points were added after "
                            . "Unicode $string_compare_versions";

    # Calculate the delta as those code points that have been newly assigned
    # since the first compare version.
    my $delta = Range_List->new();
    foreach my $table ($age->tables) {
        use version;
        next if $table == $age->table('Unassigned');
        next if version->parse($table->name)
             le version->parse($string_compare_versions);
        $delta += $table;
    }
    if ($delta->is_empty) {
        die ("No changes; perhaps you need a 'DAge.txt' file?");
    }

    my $unchanged = ~ $delta;

    calculate_Assigned() if ! defined $Assigned;
    $Assigned &= $unchanged;

    # $Assigned now contains the code points that were assigned as of Unicode
    # version X.

    # A block is all or nothing.  If nothing is assigned in it, it all goes
    # back to the No_Block pool; but if even one code point is assigned, the
    # block is retained.
    my $no_block = $block->table('No_Block');
    foreach my $this_block ($block->tables) {
        next if     $this_block == $no_block
                ||  ! ($this_block & $Assigned)->is_empty;
        $this_block->set_fate($SUPPRESSED, $after_first_version);
        foreach my $range ($this_block->ranges) {
            $block->replace_map($range->start, $range->end, 'No_Block')
        }
        $no_block += $this_block;
    }

    my @special_delta_properties;   # List of properties that have to be
                                    # handled specially.
    my %restricted_delta;           # Keys are the entries in
                                    # @special_delta_properties;  values
                                    # are the range list of the code points
                                    # that behave normally when they get
                                    # assigned.

    # In the next three properties, the Default Ignorable code points are
    # outliers.
    calculate_DI();
    $DI &= $unchanged;

    push @special_delta_properties, property_ref('_Perl_GCB');
    $restricted_delta{$special_delta_properties[-1]} = ~ $DI;

    if (defined (my $cwnfkcc = property_ref('Changes_When_NFKC_Casefolded')))
    {
        push @special_delta_properties, $cwnfkcc;
        $restricted_delta{$special_delta_properties[-1]} = ~ $DI;
    }

    calculate_NChar();      # Non-character code points
    $NChar &= $unchanged;

    # This may have to be updated from time-to-time to get the most accurate
    # results.
    my $default_BC_non_LtoR = Range_List->new(Initialize =>
                        # These came from the comments in v8.0 DBidiClass.txt
                                                        [ # AL
                                                            0x0600 .. 0x07BF,
                                                            0x08A0 .. 0x08FF,
                                                            0xFB50 .. 0xFDCF,
                                                            0xFDF0 .. 0xFDFF,
                                                            0xFE70 .. 0xFEFF,
                                                            0x1EE00 .. 0x1EEFF,
                                                           # R
                                                            0x0590 .. 0x05FF,
                                                            0x07C0 .. 0x089F,
                                                            0xFB1D .. 0xFB4F,
                                                            0x10800 .. 0x10FFF,
                                                            0x1E800 .. 0x1EDFF,
                                                            0x1EF00 .. 0x1EFFF,
                                                           # ET
                                                            0x20A0 .. 0x20CF,
                                                         ]
                                          );
    $default_BC_non_LtoR += $DI + $NChar;
    push @special_delta_properties, property_ref('BidiClass');
    $restricted_delta{$special_delta_properties[-1]} = ~ $default_BC_non_LtoR;

    if (defined (my $eaw = property_ref('East_Asian_Width'))) {

        my $default_EA_width_W = Range_List->new(Initialize =>
                                    # From comments in v8.0 EastAsianWidth.txt
                                                [
                                                    0x3400 .. 0x4DBF,
                                                    0x4E00 .. 0x9FFF,
                                                    0xF900 .. 0xFAFF,
                                                    0x20000 .. 0x2A6DF,
                                                    0x2A700 .. 0x2B73F,
                                                    0x2B740 .. 0x2B81F,
                                                    0x2B820 .. 0x2CEAF,
                                                    0x2F800 .. 0x2FA1F,
                                                    0x20000 .. 0x2FFFD,
                                                    0x30000 .. 0x3FFFD,
                                                ]
                                             );
        push @special_delta_properties, $eaw;
        $restricted_delta{$special_delta_properties[-1]}
                                                       = ~ $default_EA_width_W;

        # Line break came along in the same release as East_Asian_Width, and
        # the non-grab-bag default set is a superset of the EAW one.
        if (defined (my $lb = property_ref('Line_Break'))) {
            my $default_LB_non_XX = Range_List->new(Initialize =>
                                        # From comments in v8.0 LineBreak.txt
                                                        [ 0x20A0 .. 0x20CF ]);
            $default_LB_non_XX += $default_EA_width_W;
            push @special_delta_properties, $lb;
            $restricted_delta{$special_delta_properties[-1]}
                                                        = ~ $default_LB_non_XX;
        }
    }

    # Go through every property, skipping those we've already worked on, those
    # that are immutable, and the perl ones that will be calculated after this
    # routine has done its fixup.
    foreach my $property (property_ref('*')) {
        next if    $property == $perl     # Done later in the program
                || $property == $block    # Done just above
                || $property == $DI       # Done just above
                || $property == $NChar    # Done just above

                   # The next two are invariant across Unicode versions
                || $property == property_ref('Pattern_Syntax')
                || $property == property_ref('Pattern_White_Space');

        #  Find the grab-bag value.
        my $default_map = $property->default_map;

        if (! $property->to_create_match_tables) {

            # Here there aren't any match tables.  So far, all such properties
            # have a default map, and don't require special handling.  Just
            # change each newly assigned code point back to the default map,
            # as if they were unassigned.
            foreach my $range ($delta->ranges) {
                $property->add_map($range->start,
                                $range->end,
                                $default_map,
                                Replace => $UNCONDITIONALLY);
            }
        }
        else {  # Here there are match tables.  Find the one (if any) for the
                # grab-bag value that unassigned code points go to.
            my $default_table;
            if (defined $default_map) {
                $default_table = $property->table($default_map);
            }

            # If some code points don't go back to the grab-bag when they
            # are considered unassigned, exclude them from the list that does
            # that.
            my $this_delta = $delta;
            my $this_unchanged = $unchanged;
            if (grep { $_ == $property } @special_delta_properties) {
                $this_delta = $delta & $restricted_delta{$property};
                $this_unchanged = ~ $this_delta;
            }

            # Fix up each match table for this property.
            foreach my $table ($property->tables) {
                if (defined $default_table && $table == $default_table) {

                    # The code points assigned after release X (the ones we
                    # are excluding in this routine) go back on to the default
                    # (grab-bag) table.  However, some of these tables don't
                    # actually exist, but are specified solely by the other
                    # tables.  (In a binary property, we don't need to
                    # actually have an 'N' table, as it's just the complement
                    # of the 'Y' table.)  Such tables will be locked, so just
                    # skip those.
                    $table += $this_delta unless $table->locked;
                }
                else {

                    # Here the table is not for the default value.  We need to
                    # subtract the code points we are ignoring for this
                    # comparison (the deltas) from it.  But if the table
                    # started out with nothing, no need to exclude anything,
                    # and want to skip it here anyway, so it gets listed
                    # properly in the pod.
                    next if $table->is_empty;

                    # Save the deltas for later, before we do the subtraction
                    my $deltas = $table & $this_delta;

                    $table &= $this_unchanged;

                    # Suppress the table if the subtraction left it with
                    # nothing in it
                    if ($table->is_empty) {
                        if ($property->type == $BINARY) {
                            push @tables_that_may_be_empty, $table->complete_name;
                        }
                        else {
                            $table->set_fate($SUPPRESSED, $after_first_version);
                        }
                    }

                    # Now we add the removed code points to the property's
                    # map, as they should now map to the grab-bag default
                    # property (which they did in the first comparison
                    # version).  But we don't have to do this if the map is
                    # only for internal use.
                    if (defined $default_map && $property->to_output_map) {

                        # The gc property has pseudo property values whose names
                        # have length 1.  These are the union of all the
                        # property values whose name is longer than 1 and
                        # whose first letter is all the same.  The replacement
                        # is done once for the longer-named tables.
                        next if $property == $gc && length $table->name == 1;

                        foreach my $range ($deltas->ranges) {
                            $property->add_map($range->start,
                                            $range->end,
                                            $default_map,
                                            Replace => $UNCONDITIONALLY);
                        }
                    }
                }
            }
        }
    }

    # The above code doesn't work on 'gc=C', as it is a superset of the default
    # ('Cn') table.  It's easiest to just special case it here.
    my $C = $gc->table('C');
    $C += $gc->table('Cn');

    return;
}

sub compile_perl() {
    # Create perl-defined tables.  Almost all are part of the pseudo-property
    # named 'perl' internally to this program.  Many of these are recommended
    # in UTS#18 "Unicode Regular Expressions", and their derivations are based
    # on those found there.
    # Almost all of these are equivalent to some Unicode property.
    # A number of these properties have equivalents restricted to the ASCII
    # range, with their names prefaced by 'Posix', to signify that these match
    # what the Posix standard says they should match.  A couple are
    # effectively this, but the name doesn't have 'Posix' in it because there
    # just isn't any Posix equivalent.  'XPosix' are the Posix tables extended
    # to the full Unicode range, by our guesses as to what is appropriate.

    # 'All' is all code points.  As an error check, instead of just setting it
    # to be that, construct it to be the union of all the major categories
    $All = $perl->add_match_table('All',
      Description
        => "All code points, including those above Unicode.  Same as qr/./s",
      Matches_All => 1);

    foreach my $major_table ($gc->tables) {

        # Major categories are the ones with single letter names.
        next if length($major_table->name) != 1;

        $All += $major_table;
    }

    if ($All->max != $MAX_WORKING_CODEPOINT) {
        Carp::my_carp_bug("Generated highest code point ("
           . sprintf("%X", $All->max)
           . ") doesn't match expected value $MAX_WORKING_CODEPOINT_STRING.")
    }
    if ($All->range_count != 1 || $All->min != 0) {
     Carp::my_carp_bug("Generated table 'All' doesn't match all code points.")
    }

    my $Any = $perl->add_match_table('Any',
                                    Description  => "All Unicode code points");
    $Any->add_range(0, $MAX_UNICODE_CODEPOINT);
    $Any->add_alias('Unicode');

    calculate_Assigned();

    my $ASCII = $perl->add_match_table('ASCII');
    if (defined $block) {   # This is equivalent to the block if have it.
        my $Unicode_ASCII = $block->table('Basic_Latin');
        if (defined $Unicode_ASCII && ! $Unicode_ASCII->is_empty) {
            $ASCII->set_equivalent_to($Unicode_ASCII, Related => 1);
        }
    }

    # Very early releases didn't have blocks, so initialize ASCII ourselves if
    # necessary
    if ($ASCII->is_empty) {
        if (! NON_ASCII_PLATFORM) {
            $ASCII->add_range(0, 127);
        }
        else {
            for my $i (0 .. 127) {
                $ASCII->add_range(utf8::unicode_to_native($i),
                                  utf8::unicode_to_native($i));
            }
        }
    }

    # Get the best available case definitions.  Early Unicode versions didn't
    # have Uppercase and Lowercase defined, so use the general category
    # instead for them, modified by hard-coding in the code points each is
    # missing.
    my $Lower = $perl->add_match_table('XPosixLower');
    my $Unicode_Lower = property_ref('Lowercase');
    if (defined $Unicode_Lower && ! $Unicode_Lower->is_empty) {
        $Lower->set_equivalent_to($Unicode_Lower->table('Y'), Related => 1);

    }
    else {
        $Lower += $gc->table('Lowercase_Letter');

        # There are quite a few code points in Lower, that aren't in gc=lc,
        # and not all are in all releases.
        my $temp = Range_List->new(Initialize => [
                                                utf8::unicode_to_native(0xAA),
                                                utf8::unicode_to_native(0xBA),
                                                0x02B0 .. 0x02B8,
                                                0x02C0 .. 0x02C1,
                                                0x02E0 .. 0x02E4,
                                                0x0345,
                                                0x037A,
                                                0x1D2C .. 0x1D6A,
                                                0x1D78,
                                                0x1D9B .. 0x1DBF,
                                                0x2071,
                                                0x207F,
                                                0x2090 .. 0x209C,
                                                0x2170 .. 0x217F,
                                                0x24D0 .. 0x24E9,
                                                0x2C7C .. 0x2C7D,
                                                0xA770,
                                                0xA7F8 .. 0xA7F9,
                                ]);
        $Lower += $temp & $Assigned;
    }
    my $Posix_Lower = $perl->add_match_table("PosixLower",
                            Initialize => $Lower & $ASCII,
                            );

    my $Upper = $perl->add_match_table("XPosixUpper");
    my $Unicode_Upper = property_ref('Uppercase');
    if (defined $Unicode_Upper && ! $Unicode_Upper->is_empty) {
        $Upper->set_equivalent_to($Unicode_Upper->table('Y'), Related => 1);
    }
    else {

        # Unlike Lower, there are only two ranges in Upper that aren't in
        # gc=Lu, and all code points were assigned in all releases.
        $Upper += $gc->table('Uppercase_Letter');
        $Upper->add_range(0x2160, 0x216F);  # Uppercase Roman numerals
        $Upper->add_range(0x24B6, 0x24CF);  # Circled Latin upper case letters
    }
    my $Posix_Upper = $perl->add_match_table("PosixUpper",
                            Initialize => $Upper & $ASCII,
                            );

    # Earliest releases didn't have title case.  Initialize it to empty if not
    # otherwise present
    my $Title = $perl->add_match_table('Title', Full_Name => 'Titlecase',
                                       Description => '(= \p{Gc=Lt})');
    my $lt = $gc->table('Lt');

    # Earlier versions of mktables had this related to $lt since they have
    # identical code points, but their caseless equivalents are not the same,
    # one being 'Cased' and the other being 'LC', and so now must be kept as
    # separate entities.
    if (defined $lt) {
        $Title += $lt;
    }
    else {
        push @tables_that_may_be_empty, $Title->complete_name;
    }

    my $Unicode_Cased = property_ref('Cased');
    if (defined $Unicode_Cased) {
        my $yes = $Unicode_Cased->table('Y');
        my $no = $Unicode_Cased->table('N');
        $Title->set_caseless_equivalent($yes);
        if (defined $Unicode_Upper) {
            $Unicode_Upper->table('Y')->set_caseless_equivalent($yes);
            $Unicode_Upper->table('N')->set_caseless_equivalent($no);
        }
        $Upper->set_caseless_equivalent($yes);
        if (defined $Unicode_Lower) {
            $Unicode_Lower->table('Y')->set_caseless_equivalent($yes);
            $Unicode_Lower->table('N')->set_caseless_equivalent($no);
        }
        $Lower->set_caseless_equivalent($yes);
    }
    else {
        # If this Unicode version doesn't have Cased, set up the Perl
        # extension from first principles.  From Unicode 5.1: Definition D120:
        # A character C is defined to be cased if and only if C has the
        # Lowercase or Uppercase property or has a General_Category value of
        # Titlecase_Letter.
        my $cased = $perl->add_match_table('Cased',
                        Initialize => $Lower + $Upper + $Title,
                        Description => 'Uppercase or Lowercase or Titlecase',
                        );
        # $notcased is purely for the caseless equivalents below
        my $notcased = $perl->add_match_table('_Not_Cased',
                                Initialize => ~ $cased,
                                Fate => $INTERNAL_ONLY,
                                Description => 'All not-cased code points');
        $Title->set_caseless_equivalent($cased);
        if (defined $Unicode_Upper) {
            $Unicode_Upper->table('Y')->set_caseless_equivalent($cased);
            $Unicode_Upper->table('N')->set_caseless_equivalent($notcased);
        }
        $Upper->set_caseless_equivalent($cased);
        if (defined $Unicode_Lower) {
            $Unicode_Lower->table('Y')->set_caseless_equivalent($cased);
            $Unicode_Lower->table('N')->set_caseless_equivalent($notcased);
        }
        $Lower->set_caseless_equivalent($cased);
    }

    # The remaining perl defined tables are mostly based on Unicode TR 18,
    # "Annex C: Compatibility Properties".  All of these have two versions,
    # one whose name generally begins with Posix that is posix-compliant, and
    # one that matches Unicode characters beyond the Posix, ASCII range

    my $Alpha = $perl->add_match_table('XPosixAlpha');

    # Alphabetic was not present in early releases
    my $Alphabetic = property_ref('Alphabetic');
    if (defined $Alphabetic && ! $Alphabetic->is_empty) {
        $Alpha->set_equivalent_to($Alphabetic->table('Y'), Related => 1);
    }
    else {

        # The Alphabetic property doesn't exist for early releases, so
        # generate it.  The actual definition, in 5.2 terms is:
        #
        # gc=L + gc=Nl + Other_Alphabetic
        #
        # Other_Alphabetic is also not defined in these early releases, but it
        # contains one gc=So range plus most of gc=Mn and gc=Mc, so we add
        # those last two as well, then subtract the relatively few of them that
        # shouldn't have been added.  (The gc=So range is the circled capital
        # Latin characters.  Early releases mistakenly didn't also include the
        # lower-case versions of these characters, and so we don't either, to
        # maintain consistency with those releases that first had this
        # property.
        $Alpha->initialize($gc->table('Letter')
                           + pre_3_dot_1_Nl()
                           + $gc->table('Mn')
                           + $gc->table('Mc')
                        );
        $Alpha->add_range(0x24D0, 0x24E9);  # gc=So
        foreach my $range (     [ 0x0300, 0x0344 ],
                                [ 0x0346, 0x034E ],
                                [ 0x0360, 0x0362 ],
                                [ 0x0483, 0x0486 ],
                                [ 0x0591, 0x05AF ],
                                [ 0x06DF, 0x06E0 ],
                                [ 0x06EA, 0x06EC ],
                                [ 0x0740, 0x074A ],
                                0x093C,
                                0x094D,
                                [ 0x0951, 0x0954 ],
                                0x09BC,
                                0x09CD,
                                0x0A3C,
                                0x0A4D,
                                0x0ABC,
                                0x0ACD,
                                0x0B3C,
                                0x0B4D,
                                0x0BCD,
                                0x0C4D,
                                0x0CCD,
                                0x0D4D,
                                0x0DCA,
                                [ 0x0E47, 0x0E4C ],
                                0x0E4E,
                                [ 0x0EC8, 0x0ECC ],
                                [ 0x0F18, 0x0F19 ],
                                0x0F35,
                                0x0F37,
                                0x0F39,
                                [ 0x0F3E, 0x0F3F ],
                                [ 0x0F82, 0x0F84 ],
                                [ 0x0F86, 0x0F87 ],
                                0x0FC6,
                                0x1037,
                                0x1039,
                                [ 0x17C9, 0x17D3 ],
                                [ 0x20D0, 0x20DC ],
                                0x20E1,
                                [ 0x302A, 0x302F ],
                                [ 0x3099, 0x309A ],
                                [ 0xFE20, 0xFE23 ],
                                [ 0x1D165, 0x1D169 ],
                                [ 0x1D16D, 0x1D172 ],
                                [ 0x1D17B, 0x1D182 ],
                                [ 0x1D185, 0x1D18B ],
                                [ 0x1D1AA, 0x1D1AD ],
        ) {
            if (ref $range) {
                $Alpha->delete_range($range->[0], $range->[1]);
            }
            else {
                $Alpha->delete_range($range, $range);
            }
        }
        $Alpha->add_description('Alphabetic');
        $Alpha->add_alias('Alphabetic');
    }
    my $Posix_Alpha = $perl->add_match_table("PosixAlpha",
                            Initialize => $Alpha & $ASCII,
                            );
    $Posix_Upper->set_caseless_equivalent($Posix_Alpha);
    $Posix_Lower->set_caseless_equivalent($Posix_Alpha);

    my $Alnum = $perl->add_match_table('Alnum', Full_Name => 'XPosixAlnum',
                        Description => 'Alphabetic and (decimal) Numeric',
                        Initialize => $Alpha + $gc->table('Decimal_Number'),
                        );
    $perl->add_match_table("PosixAlnum",
                            Initialize => $Alnum & $ASCII,
                            );

    my $Word = $perl->add_match_table('Word', Full_Name => 'XPosixWord',
                                Description => '\w, including beyond ASCII;'
                                            . ' = \p{Alnum} + \pM + \p{Pc}'
                                            . ' + \p{Join_Control}',
                                Initialize => $Alnum + $gc->table('Mark'),
                                );
    my $Pc = $gc->table('Connector_Punctuation'); # 'Pc' Not in release 1
    if (defined $Pc) {
        $Word += $Pc;
    }
    else {
        $Word += ord('_');  # Make sure this is a $Word
    }
    my $JC = property_ref('Join_Control');  # Wasn't in release 1
    if (defined $JC) {
        $Word += $JC->table('Y');
    }
    else {
        $Word += 0x200C + 0x200D;
    }

    # This is a Perl extension, so the name doesn't begin with Posix.
    my $PerlWord = $perl->add_match_table('PosixWord',
                    Description => '\w, restricted to ASCII',
                    Initialize => $Word & $ASCII,
                    );
    $PerlWord->add_alias('PerlWord');

    my $Blank = $perl->add_match_table('Blank', Full_Name => 'XPosixBlank',
                                Description => '\h, Horizontal white space',

                                # 200B is Zero Width Space which is for line
                                # break control, and was listed as
                                # Space_Separator in early releases
                                Initialize => $gc->table('Space_Separator')
                                            +   ord("\t")
                                            -   0x200B, # ZWSP
                                );
    $Blank->add_alias('HorizSpace');        # Another name for it.
    $perl->add_match_table("PosixBlank",
                            Initialize => $Blank & $ASCII,
                            );

    my $VertSpace = $perl->add_match_table('VertSpace',
                            Description => '\v',
                            Initialize =>
                               $gc->table('Line_Separator')
                             + $gc->table('Paragraph_Separator')
                             + utf8::unicode_to_native(0x0A)  # LINE FEED
                             + utf8::unicode_to_native(0x0B)  # VERTICAL TAB
                             + ord("\f")
                             + utf8::unicode_to_native(0x0D)  # CARRIAGE RETURN
                             + utf8::unicode_to_native(0x85)  # NEL
                    );
    # No Posix equivalent for vertical space

    my $Space = $perl->add_match_table('XPosixSpace',
                Description => '\s including beyond ASCII and vertical tab',
                Initialize => $Blank + $VertSpace,
    );
    $Space->add_alias('XPerlSpace');    # Pre-existing synonyms
    $Space->add_alias('SpacePerl');
    $Space->add_alias('Space') if $v_version lt v4.1.0;

    my $Posix_space = $perl->add_match_table("PosixSpace",
                            Initialize => $Space & $ASCII,
                            );
    $Posix_space->add_alias('PerlSpace'); # A pre-existing synonym

    my $Cntrl = $perl->add_match_table('Cntrl', Full_Name => 'XPosixCntrl',
                                        Description => 'Control characters');
    $Cntrl->set_equivalent_to($gc->table('Cc'), Related => 1);
    $perl->add_match_table("PosixCntrl",
                            Description => "ASCII control characters",
                            Definition =>  "ACK, BEL, BS, CAN, CR, DC1, DC2,"
                                         . " DC3, DC4, DEL, DLE, ENQ, EOM,"
                                         . " EOT, ESC, ETB, ETX, FF, FS, GS,"
                                         . " HT, LF, NAK, NUL, RS, SI, SO,"
                                         . " SOH, STX, SUB, SYN, US, VT",
                            Initialize => $Cntrl & $ASCII,
                            );

    my $perl_surrogate = $perl->add_match_table('_Perl_Surrogate');
    my $Cs = $gc->table('Cs');
    if (defined $Cs && ! $Cs->is_empty) {
        $perl_surrogate += $Cs;
    }
    else {
        push @tables_that_may_be_empty, '_Perl_Surrogate';
    }

    # $controls is a temporary used to construct Graph.
    my $controls = Range_List->new(Initialize => $gc->table('Unassigned')
                                                + $gc->table('Control')
                                                + $perl_surrogate);

    # Graph is  ~space &  ~(Cc|Cs|Cn) = ~(space + $controls)
    my $Graph = $perl->add_match_table('Graph', Full_Name => 'XPosixGraph',
                        Description => 'Characters that are graphical',
                        Initialize => ~ ($Space + $controls),
                        );
    $perl->add_match_table("PosixGraph",
                            Initialize => $Graph & $ASCII,
                            );

    $print = $perl->add_match_table('Print', Full_Name => 'XPosixPrint',
                        Description => 'Characters that are graphical plus space characters (but no controls)',
                        Initialize => $Blank + $Graph - $gc->table('Control'),
                        );
    $perl->add_match_table("PosixPrint",
                            Initialize => $print & $ASCII,
                            );

    my $Punct = $perl->add_match_table('Punct');
    $Punct->set_equivalent_to($gc->table('Punctuation'), Related => 1);

    # \p{punct} doesn't include the symbols, which posix does
    my $XPosixPunct = $perl->add_match_table('XPosixPunct',
                    Description => '\p{Punct} + ASCII-range \p{Symbol}',
                    Initialize => $gc->table('Punctuation')
                                + ($ASCII & $gc->table('Symbol')),
                                Perl_Extension => 1
        );
    $perl->add_match_table('PosixPunct', Perl_Extension => 1,
        Initialize => $ASCII & $XPosixPunct,
        );

    my $Digit = $perl->add_match_table('Digit', Full_Name => 'XPosixDigit',
                            Description => '[0-9] + all other decimal digits');
    $Digit->set_equivalent_to($gc->table('Decimal_Number'), Related => 1);
    my $PosixDigit = $perl->add_match_table("PosixDigit",
                                            Initialize => $Digit & $ASCII,
                                            );

    # Hex_Digit was not present in first release
    my $Xdigit = $perl->add_match_table('XDigit', Full_Name => 'XPosixXDigit');
    my $Hex = property_ref('Hex_Digit');
    if (defined $Hex && ! $Hex->is_empty) {
        $Xdigit->set_equivalent_to($Hex->table('Y'), Related => 1);
    }
    else {
        $Xdigit->initialize([ ord('0') .. ord('9'),
                              ord('A') .. ord('F'),
                              ord('a') .. ord('f'),
                              0xFF10..0xFF19, 0xFF21..0xFF26, 0xFF41..0xFF46]);
    }

    # AHex was not present in early releases
    my $PosixXDigit = $perl->add_match_table('PosixXDigit');
    my $AHex = property_ref('ASCII_Hex_Digit');
    if (defined $AHex && ! $AHex->is_empty) {
        $PosixXDigit->set_equivalent_to($AHex->table('Y'), Related => 1);
    }
    else {
        $PosixXDigit->initialize($Xdigit & $ASCII);
        $PosixXDigit->add_alias('AHex');
        $PosixXDigit->add_alias('Ascii_Hex_Digit');
    }

    my $any_folds = $perl->add_match_table("_Perl_Any_Folds",
                    Description => "Code points that particpate in some fold",
                    );
    my $loc_problem_folds = $perl->add_match_table(
               "_Perl_Problematic_Locale_Folds",
               Description =>
                   "Code points that are in some way problematic under locale",
    );

    # This allows regexec.c to skip some work when appropriate.  Some of the
    # entries in _Perl_Problematic_Locale_Folds are multi-character folds,
    my $loc_problem_folds_start = $perl->add_match_table(
               "_Perl_Problematic_Locale_Foldeds_Start",
               Description =>
                   "The first character of every sequence in _Perl_Problematic_Locale_Folds",
    );

    my $cf = property_ref('Case_Folding');

    # Every character 0-255 is problematic because what each folds to depends
    # on the current locale
    $loc_problem_folds->add_range(0, 255);
    $loc_problem_folds->add_range(0x130, 0x131);    # These are problematic in
                                                    # Turkic locales
    $loc_problem_folds_start += $loc_problem_folds;

    # Also problematic are anything these fold to outside the range.  Likely
    # forever the only thing folded to by these outside the 0-255 range is the
    # GREEK SMALL MU (from the MICRO SIGN), but it's easy to make the code
    # completely general, which should catch any unexpected changes or errors.
    # We look at each code point 0-255, and add its fold (including each part
    # of a multi-char fold) to the list.  See commit message
    # 31f05a37c4e9c37a7263491f2fc0237d836e1a80 for a more complete description
    # of the MU issue.
    foreach my $range ($loc_problem_folds->ranges) {
        foreach my $code_point ($range->start .. $range->end) {
            my $fold_range = $cf->containing_range($code_point);
            next unless defined $fold_range;

            # Skip if folds to itself
            next if $fold_range->value eq $CODE_POINT;

            my @hex_folds = split " ", $fold_range->value;
            my $start_cp = $hex_folds[0];
            next if $start_cp eq $CODE_POINT;
            $start_cp = hex $start_cp;
            foreach my $i (0 .. @hex_folds - 1) {
                my $cp = $hex_folds[$i];
                next if $cp eq $CODE_POINT;
                $cp = hex $cp;
                next unless $cp > 255;    # Already have the < 256 ones

                $loc_problem_folds->add_range($cp, $cp);
                $loc_problem_folds_start->add_range($start_cp, $start_cp);
            }
        }
    }

    my $folds_to_multi_char = $perl->add_match_table(
         "_Perl_Folds_To_Multi_Char",
         Description =>
              "Code points whose fold is a string of more than one character",
    );
    my $in_multi_fold = $perl->add_match_table(
               "_Perl_Is_In_Multi_Char_Fold",
               Description =>
                   "Code points that are in some multiple character fold",
    );
    if ($v_version lt v3.0.1) {
        push @tables_that_may_be_empty, '_Perl_Folds_To_Multi_Char',
                                        '_Perl_Is_In_Multi_Char_Fold',
                                        '_Perl_Non_Final_Folds';
    }

    # Look through all the known folds to populate these tables.
    foreach my $range ($cf->ranges) {
        next if $range->value eq $CODE_POINT;
        my $start = $range->start;
        my $end = $range->end;
        $any_folds->add_range($start, $end);

        my @hex_folds = split " ", $range->value;
        if (@hex_folds > 1) {   # Is multi-char fold
            $folds_to_multi_char->add_range($start, $end);
        }

        my $found_locale_problematic = 0;

        my $folded_count = @hex_folds;
        if ($folded_count > 3) {
            die Carp::my_carp("Maximum number of characters in a fold should be 3: Instead, it's  $folded_count for U+" . sprintf "%04X", $range->start);
        }

        # Look at each of the folded-to characters...
        foreach my $i (1 .. $folded_count) {
            my $cp = hex $hex_folds[$i-1];
            $any_folds->add_range($cp, $cp);

            # The fold is problematic if any of the folded-to characters is
            # already considered problematic.
            if ($loc_problem_folds->contains($cp)) {
                $loc_problem_folds->add_range($start, $end);
                $found_locale_problematic = 1;
            }

            if ($folded_count > 1) {
                $in_multi_fold->add_range($cp, $cp);
            }
        }

        # If this is a problematic fold, add to the start chars the
        # folding-from characters and first folded-to character.
        if ($found_locale_problematic) {
            $loc_problem_folds_start->add_range($start, $end);
            my $cp = hex $hex_folds[0];
            $loc_problem_folds_start->add_range($cp, $cp);
        }
    }

    my $dt = property_ref('Decomposition_Type');
    $dt->add_match_table('Non_Canon', Full_Name => 'Non_Canonical',
        Initialize => ~ ($dt->table('None') + $dt->table('Canonical')),
        Perl_Extension => 1,
        Note => 'Union of all non-canonical decompositions',
        );

    # For backward compatibility, Perl has its own definition for IDStart.
    # It is regular XID_Start plus the underscore, but all characters must be
    # Word characters as well
    my $XID_Start = property_ref('XID_Start');
    my $perl_xids = $perl->add_match_table('_Perl_IDStart',
                                            Perl_Extension => 1,
                                            Fate => $INTERNAL_ONLY,
                                            Initialize => ord('_')
                                            );
    if (defined $XID_Start
        || defined ($XID_Start = property_ref('ID_Start')))
    {
        $perl_xids += $XID_Start->table('Y');
    }
    else {
        # For Unicode versions that don't have the property, construct our own
        # from first principles.  The actual definition is:
        #     Letters
        #   + letter numbers (Nl)
        #   - Pattern_Syntax
        #   - Pattern_White_Space
        #   + stability extensions
        #   - NKFC modifications
        #
        # What we do in the code below is to include the identical code points
        # that are in the first release that had Unicode's version of this
        # property, essentially extrapolating backwards.  There were no
        # stability extensions until v4.1, so none are included; likewise in
        # no Unicode version so far do subtracting PatSyn and PatWS make any
        # difference, so those also are ignored.
        $perl_xids += $gc->table('Letter') + pre_3_dot_1_Nl();

        # We do subtract the NFKC modifications that are in the first version
        # that had this property.  We don't bother to test if they are in the
        # version in question, because if they aren't, the operation is a
        # no-op.  The NKFC modifications are discussed in
        # http://www.unicode.org/reports/tr31/#NFKC_Modifications
        foreach my $range ( 0x037A,
                            0x0E33,
                            0x0EB3,
                            [ 0xFC5E, 0xFC63 ],
                            [ 0xFDFA, 0xFE70 ],
                            [ 0xFE72, 0xFE76 ],
                            0xFE78,
                            0xFE7A,
                            0xFE7C,
                            0xFE7E,
                            [ 0xFF9E, 0xFF9F ],
        ) {
            if (ref $range) {
                $perl_xids->delete_range($range->[0], $range->[1]);
            }
            else {
                $perl_xids->delete_range($range, $range);
            }
        }
    }

    $perl_xids &= $Word;

    my $perl_xidc = $perl->add_match_table('_Perl_IDCont',
                                        Perl_Extension => 1,
                                        Fate => $INTERNAL_ONLY);
    my $XIDC = property_ref('XID_Continue');
    if (defined $XIDC
        || defined ($XIDC = property_ref('ID_Continue')))
    {
        $perl_xidc += $XIDC->table('Y');
    }
    else {
        # Similarly, we construct our own XIDC if necessary for early Unicode
        # versions.  The definition is:
        #     everything in XIDS
        #   + Gc=Mn
        #   + Gc=Mc
        #   + Gc=Nd
        #   + Gc=Pc
        #   - Pattern_Syntax
        #   - Pattern_White_Space
        #   + stability extensions
        #   - NFKC modifications
        #
        # The same thing applies to this as with XIDS for the PatSyn, PatWS,
        # and stability extensions.  There is a somewhat different set of NFKC
        # mods to remove (and add in this case).  The ones below make this
        # have identical code points as in the first release that defined it.
        $perl_xidc += $perl_xids
                    + $gc->table('L')
                    + $gc->table('Mn')
                    + $gc->table('Mc')
                    + $gc->table('Nd')
                    + utf8::unicode_to_native(0xB7)
                    ;
        if (defined (my $pc = $gc->table('Pc'))) {
            $perl_xidc += $pc;
        }
        else {  # 1.1.5 didn't have Pc, but these should have been in it
            $perl_xidc += 0xFF3F;
            $perl_xidc->add_range(0x203F, 0x2040);
            $perl_xidc->add_range(0xFE33, 0xFE34);
            $perl_xidc->add_range(0xFE4D, 0xFE4F);
        }

        # Subtract the NFKC mods
        foreach my $range ( 0x037A,
                            [ 0xFC5E, 0xFC63 ],
                            [ 0xFDFA, 0xFE1F ],
                            0xFE70,
                            [ 0xFE72, 0xFE76 ],
                            0xFE78,
                            0xFE7A,
                            0xFE7C,
                            0xFE7E,
        ) {
            if (ref $range) {
                $perl_xidc->delete_range($range->[0], $range->[1]);
            }
            else {
                $perl_xidc->delete_range($range, $range);
            }
        }
    }

    $perl_xidc &= $Word;

    my $charname_begin = $perl->add_match_table('_Perl_Charname_Begin',
                    Perl_Extension => 1,
                    Fate => $INTERNAL_ONLY,
                    Initialize => $gc->table('Letter') & $Alpha & $perl_xids,
                    );

    my $charname_continue = $perl->add_match_table('_Perl_Charname_Continue',
                        Perl_Extension => 1,
                        Fate => $INTERNAL_ONLY,
                        Initialize => $perl_xidc
                                    + ord(" ")
                                    + ord("(")
                                    + ord(")")
                                    + ord("-")
                        );

    my @composition = ('Name', 'Unicode_1_Name', '_Perl_Name_Alias');

    if (@named_sequences) {
        push @composition, 'Named_Sequence';
        foreach my $sequence (@named_sequences) {
            $perl_charname->add_anomalous_entry($sequence);
        }
    }

    my $alias_sentence = "";
    my %abbreviations;
    my $alias = property_ref('_Perl_Name_Alias');
    $perl_charname->set_proxy_for('_Perl_Name_Alias');

    # Add each entry in _Perl_Name_Alias to Perl_Charnames.  Where these go
    # with respect to any existing entry depends on the entry type.
    # Corrections go before said entry, as they should be returned in
    # preference over the existing entry.  (A correction to a correction
    # should be later in the _Perl_Name_Alias table, so it will correctly
    # precede the erroneous correction in Perl_Charnames.)
    #
    # Abbreviations go after everything else, so they are saved temporarily in
    # a hash for later.
    #
    # Everything else is added afterwards, which preserves the input
    # ordering

    foreach my $range ($alias->ranges) {
        next if $range->value eq "";
        my $code_point = $range->start;
        if ($code_point != $range->end) {
            Carp::my_carp_bug("Bad News.  Expecting only one code point in the range $range.  Just to keep going, using only the first code point;");
        }
        my ($value, $type) = split ': ', $range->value;
        my $replace_type;
        if ($type eq 'correction') {
            $replace_type = $MULTIPLE_BEFORE;
        }
        elsif ($type eq 'abbreviation') {

            # Save for later
            $abbreviations{$value} = $code_point;
            next;
        }
        else {
            $replace_type = $MULTIPLE_AFTER;
        }

        # Actually add; before or after current entry(ies) as determined
        # above.

        $perl_charname->add_duplicate($code_point, $value, Replace => $replace_type);
    }
    $alias_sentence = <<END;
The _Perl_Name_Alias property adds duplicate code point entries that are
alternatives to the original name.  If an addition is a corrected
name, it will be physically first in the table.  The original (less correct,
but still valid) name will be next; then any alternatives, in no particular
order; and finally any abbreviations, again in no particular order.
END

    # Now add the Unicode_1 names for the controls.  The Unicode_1 names had
    # precedence before 6.1, including the awful ones like "LINE FEED (LF)",
    # so should be first in the file; the other names have precedence starting
    # in 6.1,
    my $before_or_after = ($v_version lt v6.1.0)
                          ? $MULTIPLE_BEFORE
                          : $MULTIPLE_AFTER;

    foreach my $range (property_ref('Unicode_1_Name')->ranges) {
        my $code_point = $range->start;
        my $unicode_1_value = $range->value;
        next if $unicode_1_value eq "";     # Skip if name doesn't exist.

        if ($code_point != $range->end) {
            Carp::my_carp_bug("Bad News.  Expecting only one code point in the range $range.  Just to keep going, using only the first code point;");
        }

        # To handle EBCDIC, we don't hard code in the code points of the
        # controls; instead realizing that all of them are below 256.
        last if $code_point > 255;

        # We only add in the controls.
        next if $gc->value_of($code_point) ne 'Cc';

        # We reject this Unicode1 name for later Perls, as it is used for
        # another code point
        next if $unicode_1_value eq 'BELL' && $^V ge v5.17.0;

        # This won't add an exact duplicate.
        $perl_charname->add_duplicate($code_point, $unicode_1_value,
                                        Replace => $before_or_after);
    }

    # Now that have everything added, add in abbreviations after
    # everything else.  Sort so results don't change between runs of this
    # program
    foreach my $value (sort keys %abbreviations) {
        $perl_charname->add_duplicate($abbreviations{$value}, $value,
                                        Replace => $MULTIPLE_AFTER);
    }

    my $comment;
    if (@composition <= 2) { # Always at least 2
        $comment = join " and ", @composition;
    }
    else {
        $comment = join ", ", @composition[0 .. scalar @composition - 2];
        $comment .= ", and $composition[-1]";
    }

    $perl_charname->add_comment(join_lines( <<END
This file is for charnames.pm.  It is the union of the $comment properties.
Unicode_1_Name entries are used only for nameless code points in the Name
property.
$alias_sentence
This file doesn't include the algorithmically determinable names.  For those,
use 'unicore/Name.pm'
END
    ));
    property_ref('Name')->add_comment(join_lines( <<END
This file doesn't include the algorithmically determinable names.  For those,
use 'unicore/Name.pm'
END
    ));

    # Construct the Present_In property from the Age property.
    if (-e 'DAge.txt' && defined $age) {
        my $default_map = $age->default_map;
        my $in = Property->new('In',
                                Default_Map => $default_map,
                                Full_Name => "Present_In",
                                Perl_Extension => 1,
                                Type => $ENUM,
                                Initialize => $age,
                                );
        $in->add_comment(join_lines(<<END
THIS FILE SHOULD NOT BE USED FOR ANY PURPOSE.  The values in this file are the
same as for $age, and not for what $in really means.  This is because anything
defined in a given release should have multiple values: that release and all
higher ones.  But only one value per code point can be represented in a table
like this.
END
        ));

        # The Age tables are named like 1.5, 2.0, 2.1, ....  Sort so that the
        # lowest numbered (earliest) come first, with the non-numeric one
        # last.
        my ($first_age, @rest_ages) = sort { ($a->name !~ /^[\d.]*$/)
                                            ? 1
                                            : ($b->name !~ /^[\d.]*$/)
                                                ? -1
                                                : $a->name <=> $b->name
                                            } $age->tables;

        # The Present_In property is the cumulative age properties.  The first
        # one hence is identical to the first age one.
        my $first_in = $in->add_match_table($first_age->name);
        $first_in->set_equivalent_to($first_age, Related => 1);

        my $description_start = "Code point's usage introduced in version ";
        $first_age->add_description($description_start . $first_age->name);
        foreach my $alias ($first_age->aliases) {   # Include its aliases
            $first_in->add_alias($alias->name);
        }

        # To construct the accumulated values, for each of the age tables
        # starting with the 2nd earliest, merge the earliest with it, to get
        # all those code points existing in the 2nd earliest.  Repeat merging
        # the new 2nd earliest with the 3rd earliest to get all those existing
        # in the 3rd earliest, and so on.
        my $previous_in = $first_in;
        foreach my $current_age (@rest_ages) {
            next if $current_age->name !~ /^[\d.]*$/;   # Skip the non-numeric

            my $current_in = $in->add_match_table(
                                    $current_age->name,
                                    Initialize => $current_age + $previous_in,
                                    Description => $description_start
                                                    . $current_age->name
                                                    . ' or earlier',
                                    );
            foreach my $alias ($current_age->aliases) {
                $current_in->add_alias($alias->name);
            }
            $previous_in = $current_in;

            # Add clarifying material for the corresponding age file.  This is
            # in part because of the confusing and contradictory information
            # given in the Standard's documentation itself, as of 5.2.
            $current_age->add_description(
                            "Code point's usage was introduced in version "
                            . $current_age->name);
            $current_age->add_note("See also $in");

        }

        # And finally the code points whose usages have yet to be decided are
        # the same in both properties.  Note that permanently unassigned code
        # points actually have their usage assigned (as being permanently
        # unassigned), so that these tables are not the same as gc=cn.
        my $unassigned = $in->add_match_table($default_map);
        my $age_default = $age->table($default_map);
        $age_default->add_description(<<END
Code point's usage has not been assigned in any Unicode release thus far.
END
        );
        $unassigned->set_equivalent_to($age_default, Related => 1);
        foreach my $alias ($age_default->aliases) {
            $unassigned->add_alias($alias->name);
        }
    }

    my $patws = $perl->add_match_table('_Perl_PatWS',
                                       Perl_Extension => 1,
                                       Fate => $INTERNAL_ONLY);
    if (defined (my $off_patws = property_ref('Pattern_White_Space'))) {
        $patws->initialize($off_patws->table('Y'));
    }
    else {
        $patws->initialize([ ord("\t"),
                             ord("\n"),
                             utf8::unicode_to_native(0x0B), # VT
                             ord("\f"),
                             ord("\r"),
                             ord(" "),
                             utf8::unicode_to_native(0x85), # NEL
                             0x200E..0x200F,             # Left, Right marks
                             0x2028..0x2029              # Line, Paragraph seps
                           ] );
    }

    # See L<perlfunc/quotemeta>
    my $quotemeta = $perl->add_match_table('_Perl_Quotemeta',
                                           Perl_Extension => 1,
                                           Fate => $INTERNAL_ONLY,

                                           # Initialize to what's common in
                                           # all Unicode releases.
                                           Initialize =>
                                                  $gc->table('Control')
                                                + $Space
                                                + $patws
                                                + ((~ $Word) & $ASCII)
                           );

    if (defined (my $patsyn = property_ref('Pattern_Syntax'))) {
        $quotemeta += $patsyn->table('Y');
    }
    else {
        $quotemeta += ((~ $Word) & Range->new(0, 255))
                    - utf8::unicode_to_native(0xA8)
                    - utf8::unicode_to_native(0xAF)
                    - utf8::unicode_to_native(0xB2)
                    - utf8::unicode_to_native(0xB3)
                    - utf8::unicode_to_native(0xB4)
                    - utf8::unicode_to_native(0xB7)
                    - utf8::unicode_to_native(0xB8)
                    - utf8::unicode_to_native(0xB9)
                    - utf8::unicode_to_native(0xBC)
                    - utf8::unicode_to_native(0xBD)
                    - utf8::unicode_to_native(0xBE);
        $quotemeta += [ # These are above-Latin1 patsyn; hence should be the
                        # same in all releases
                        0x2010 .. 0x2027,
                        0x2030 .. 0x203E,
                        0x2041 .. 0x2053,
                        0x2055 .. 0x205E,
                        0x2190 .. 0x245F,
                        0x2500 .. 0x2775,
                        0x2794 .. 0x2BFF,
                        0x2E00 .. 0x2E7F,
                        0x3001 .. 0x3003,
                        0x3008 .. 0x3020,
                        0x3030 .. 0x3030,
                        0xFD3E .. 0xFD3F,
                        0xFE45 .. 0xFE46
                      ];
    }

    if (defined (my $di = property_ref('Default_Ignorable_Code_Point'))) {
        $quotemeta += $di->table('Y')
    }
    else {
        if ($v_version ge v2.0) {
            $quotemeta += $gc->table('Cf')
                       +  $gc->table('Cs');

            # These are above the Unicode version 1 max
            $quotemeta->add_range(0xE0000, 0xE0FFF);
        }
        $quotemeta += $gc->table('Cc')
                    - $Space;
        my $temp = Range_List->new(Initialize => [ 0x180B .. 0x180D,
                                                   0x2060 .. 0x206F,
                                                   0xFE00 .. 0xFE0F,
                                                   0xFFF0 .. 0xFFFB,
                                                  ]);
        $temp->add_range(0xE0000, 0xE0FFF) if $v_version ge v2.0;
        $quotemeta += $temp;
    }
    calculate_DI();
    $quotemeta += $DI;

    calculate_NChar();

    # Finished creating all the perl properties.  All non-internal non-string
    # ones have a synonym of 'Is_' prefixed.  (Internal properties begin with
    # an underscore.)  These do not get a separate entry in the pod file
    foreach my $table ($perl->tables) {
        foreach my $alias ($table->aliases) {
            next if $alias->name =~ /^_/;
            $table->add_alias('Is_' . $alias->name,
                               Re_Pod_Entry => 0,
                               UCD => 0,
                               Status => $alias->status,
                               OK_as_Filename => 0);
        }
    }

    # Perl tailors the WordBreak property so that \b{wb} doesn't split
    # adjacent spaces into separate words.  Unicode 11.0 moved in that
    # direction, but left TAB,  FIGURE SPACE (U+2007), and (ironically)
    # NO_BREAK SPACE as breaking, so we retained the original Perl
    # customization.  To do this, in the Perl copy of WB, simply replace the
    # mappings of horizontal space characters that otherwise would map to the
    # default or the 11.0 'WSegSpace' to instead map to our tailoring.
    my $perl_wb = property_ref('_Perl_WB');
    my $default = $perl_wb->default_map;
    for my $range ($Blank->ranges) {
        for my $i ($range->start .. $range->end) {
            my $value = $perl_wb->value_of($i);

            next unless $value eq $default || $value eq 'WSegSpace';
            $perl_wb->add_map($i, $i, 'Perl_Tailored_HSpace',
                              Replace => $UNCONDITIONALLY);
        }
    }

    # Also starting in Unicode 11.0, rules for some of the boundary types are
    # based on a non-UCD property (which we have read in if it exists).
    # Recall that these boundary properties partition the code points into
    # equivalence classes (represented as enums).
    #
    # The loop below goes through each code point that matches the non-UCD
    # property, and for each current equivalence class containing such a code
    # point, splits it so that those that are in both are now in a newly
    # created equivalence class whose name is a combination of the property
    # and the old class name, leaving unchanged everything that doesn't match
    # the non-UCD property.
    my $ep = property_ref('ExtPict');
    $ep = $ep->table('Y') if defined $ep;
    if (defined $ep) {
        foreach my $base_property (property_ref('GCB'),
                                   property_ref('WB'))
        {
            my $property = property_ref('_Perl_' . $base_property->name);
            foreach my $range ($ep->ranges) {
                foreach my $i ($range->start .. $range->end) {
                    my $current = $property->value_of($i);
                    $current = $property->table($current)->short_name;
                    $property->add_map($i, $i, 'ExtPict_' . $current,
                                       Replace => $UNCONDITIONALLY);
                }
            }
        }
    }

    # In Unicode 15.1, the InCB property was added, which causes us to have to
    # split GCB into subclasses that match various subclasses of InCB
    my $perl_gcb = property_ref('_Perl_GCB');
    my $incb = property_ref('InCB');
    if (defined $perl_gcb && defined $incb) {

        # For each class in GCB ...
        foreach my $gcb_table ($perl_gcb->tables) {
            my $gcb_name = $gcb_table->name;

            # ... we see if it has any code points that are in the three
            # classes of interest in INCB.
            foreach my $incb_table ($incb->table('Consonant'),
                                    $incb->table('Extend'),
                                    $incb->table('Linker'))
            {
                my $intersection = $gcb_table & $incb_table;

                # If the intersection is empty, then nothing need be done.
                next unless $intersection->ranges;

                # Likewise if the intersection doesn't subtract anything,
                # nothing need be done.
                next if $gcb_table->matches_identically_to($intersection);

                # Otherwise, construct a new table consisting of the
                # intersection, removing its entries from the existing GCB
                # table.  The name of the new table is the combination of the
                # GCB and InCB table names
                my $incb_name = $incb_table->name;
                my $combined_name = "${gcb_name}_$incb_name";

                foreach my $range ($intersection->ranges) {
                    $perl_gcb->replace_map($range->start, $range->end,
                                           $combined_name);
                }
            }
        }
    }

    # Create a version of the LineBreak property with the mappings that are
    # omitted in the default algorithm remapped to what
    # http://www.unicode.org/reports/tr14 says they should be.
    #
    # First, create a plain copy, but with all property values written out in
    # their long form, as regen/mk_invlist.pl expects that, and also fix
    # occurrences of the typo in early Unicode versions: 'inseperable'.
    my $perl_lb = property_ref('_Perl_LB');
    if (! defined $perl_lb) {
        $perl_lb = Property->new('_Perl_LB',
                                 Fate => $INTERNAL_ONLY,
                                 Perl_Extension => 1,
                                 Directory => $map_directory,
                                 Type => $STRING);
        my $lb = property_ref('Line_Break');

        # Populate from $lb, but use full name and fix typo.
        foreach my $range ($lb->ranges) {
            my $full_name = $lb->table($range->value)->full_name;
            $full_name = 'Inseparable'
                                if standardize($full_name) eq 'inseperable';
            $perl_lb->add_map($range->start, $range->end, $full_name);
        }
    }

    # What tr14 says is this:

    # Original 	   Resolved  General_Category
    # AI, SG, XX      AL      Any
    # SA              CM      Only Mn or Mc
    # SA              AL      Any except Mn and Mc
    # CJ              NS      Any

    $perl_lb->set_default_map('Alphabetic', 'full_name');    # XX -> AL

    my $ea = property_ref('East_Asian_Width');
    my $Cn_EP;
    $Cn_EP = $ep & $gc->table('Unassigned') if defined $ep;

    for my $range ($perl_lb->ranges) {
        my $value = standardize($range->value);
        if (   $value eq standardize('Unknown')
            || $value eq standardize('Ambiguous')
            || $value eq standardize('Surrogate'))
        {
            $perl_lb->add_map($range->start, $range->end, 'Alphabetic',
                              Replace => $UNCONDITIONALLY);
        }
        elsif ($value eq standardize('Conditional_Japanese_Starter')) {
            $perl_lb->add_map($range->start, $range->end, 'Nonstarter',
                              Replace => $UNCONDITIONALLY);
        }
        elsif ($value eq standardize('Complex_Context')) {
            for my $i ($range->start .. $range->end) {
                my $gc_val = $gc->value_of($i);
                if ($gc_val eq 'Mn' || $gc_val eq 'Mc') {
                    $perl_lb->add_map($i, $i, 'Combining_Mark',
                                      Replace => $UNCONDITIONALLY);
                }
                else {
                    $perl_lb->add_map($i, $i, 'Alphabetic',
                                      Replace => $UNCONDITIONALLY);
                }
            }
        }
        elsif (defined $ep && $value eq standardize('Ideographic')) {

            # Unicode 14 adds a rule to not break lines before any potential
            # EBase,  They say that any unassigned code point that is ExtPict,
            # is potentially an EBase.  In 14.0, all such ones are in the
            # ExtPict=ID category.  We must split that category for the
            # pairwise rule table to work.
            for my $i ($range->start .. $range->end) {
                if ($Cn_EP->contains($i)) {
                    $perl_lb->add_map($i, $i,
                                'Unassigned_Extended_Pictographic_Ideographic',
                                Replace => $UNCONDITIONALLY);
                }
            }
        }
        elsif (    defined $ea
               && (   $value eq standardize('Close_Parenthesis')
                   || $value eq standardize('Open_Punctuation')))
        {
            # Unicode 13 splits the OP and CP properties each into East Asian,
            # and non-.  We retain the (now somewhat misleading) names OP and
            # CP for the non-East Asian variety, as there are very few East
            # Asian ones.
            my $replace = ($value eq standardize('Open_Punctuation'))
                          ? 'East_Asian_OP'
                          : 'East_Asian_CP';
            for my $i ($range->start .. $range->end) {
                my $ea_val = $ea->value_of($i);
                if ($ea_val eq 'F' || $ea_val eq 'W' || $ea_val eq 'H') {
                    $perl_lb->add_map($i, $i, $replace,
                                                Replace => $UNCONDITIONALLY);
                }
            }
        }
        elsif ($v_version ge 15.1.0 && $value eq standardize('Quotation')) {

            # Unicode 15.1 splits LB=QU initial quotes and final quotes, and
            # regular quotes
            for my $i ($range->start .. $range->end) {
                my $gc_val = $gc->value_of($i);
                if ($gc_val eq 'Pi') {
                    $perl_lb->replace_map($i, $i, "Initial_Quote");
                }
                elsif ($gc_val eq 'Pf') {
                    $perl_lb->replace_map($i, $i, "Final_Quote");
                }
            }
        }
    }

    # This is an Alphabetic, but it doesn't need to be split off, because no
    # current rule involving Alphabetics requires not including this.
    $perl_lb->replace_map(0x25CC, 0x25CC, "Dotted_Circle")
                                                      if $v_version ge 15.1.0;

    # This property is a modification of the scx property
    my $perl_scx = Property->new('_Perl_SCX',
                                 Fate => $INTERNAL_ONLY,
                                 Perl_Extension => 1,
                                 Directory => $map_directory,
                                 Type => $ENUM);
    my $source;

    # Use scx if available; otherwise sc;  if neither is there (a very old
    # Unicode version, just say that everything is 'Common'
    if (defined $scx) {
        $source = $scx;
        $perl_scx->set_default_map('Unknown');
    }
    elsif (defined $script) {
        $source = $script;

        # Early versions of 'sc', had everything be 'Common'
        if (defined $script->table('Unknown')) {
            $perl_scx->set_default_map('Unknown');
        }
        else {
            $perl_scx->set_default_map('Common');
        }
    } else {
        $perl_scx->add_match_table('Common');
        $perl_scx->add_map(0, $MAX_UNICODE_CODEPOINT, 'Common');

        $perl_scx->add_match_table('Unknown');
        $perl_scx->set_default_map('Unknown');
    }

    $perl_scx->_set_format($STRING_WHITE_SPACE_LIST);
    $perl_scx->set_pre_declared_maps(0); # PropValueAliases doesn't list these

    if (defined $source) {
        $perl_scx->initialize($source);

        # UTS 39 says that the scx property should be modified for these
        # countries where certain mixed scripts are commonly used.
        for my $range ($perl_scx->ranges) {
            my $value = $range->value;
            my $changed = $value =~ s/ ( \b Han i? \b ) /$1 Hanb Jpan Kore/xi;
             $changed |=  $value =~ s/ ( \b Hira (gana)? \b ) /$1 Jpan/xi;
             $changed |=  $value =~ s/ ( \b Kata (kana)? \b ) /$1 Jpan/xi;
             $changed |=  $value =~ s{ ( \b Katakana_or_Hiragana \b ) }
                                     {$1 Katakana Hiragana Jpan}xi;
             $changed |=  $value =~ s/ ( \b Hang (ul)? \b ) /$1 Kore/xi;
             $changed |=  $value =~ s/ ( \b Bopo (mofo)? \b ) /$1 Hanb/xi;

            if ($changed) {
                $value = join " ", uniques split " ", $value;
                $range->set_value($value)
            }
        }

        foreach my $table ($source->tables) {
            my $scx_table = $perl_scx->add_match_table($table->name,
                                    Full_Name => $table->full_name);
            foreach my $alias ($table->aliases) {
                $scx_table->add_alias($alias->name);
            }
        }
    }

    # Here done with all the basic stuff.  Ready to populate the information
    # about each character if annotating them.
    if ($annotate) {

        # See comments at its declaration
        $annotate_ranges = Range_Map->new;

        # This separates out the non-characters from the other unassigneds, so
        # can give different annotations for each.
        $unassigned_sans_noncharacters = Range_List->new(
                                    Initialize => $gc->table('Unassigned'));
        $unassigned_sans_noncharacters &= (~ $NChar);

        for (my $i = 0; $i <= $MAX_UNICODE_CODEPOINT + 1; $i++ ) {
            $i = populate_char_info($i);    # Note sets $i so may cause skips

        }
    }

    return;
}

sub add_perl_synonyms() {
    # A number of Unicode tables have Perl synonyms that are expressed in
    # the single-form, \p{name}.  These are:
    #   All the binary property Y tables, so that \p{Name=Y} gets \p{Name} and
    #       \p{Is_Name} as synonyms
    #   \p{Script_Extensions=Value} gets \p{Value}, \p{Is_Value} as synonyms
    #   \p{General_Category=Value} gets \p{Value}, \p{Is_Value} as synonyms
    #   \p{Block=Value} gets \p{In_Value} as a synonym, and, if there is no
    #       conflict, \p{Value} and \p{Is_Value} as well
    #
    # This routine generates these synonyms, warning of any unexpected
    # conflicts.

    # Construct the list of tables to get synonyms for.  Start with all the
    # binary and the General_Category ones.
    my @tables = grep { $_->type == $BINARY || $_->type == $FORCED_BINARY }
                                                            property_ref('*');
    push @tables, $gc->tables;

    # If the version of Unicode includes the Script Extensions (preferably),
    # or Script property, add its tables
    if (defined $scx) {
        push @tables, $scx->tables;
    }
    else {
        push @tables, $script->tables if defined $script;
    }

    # The Block tables are kept separate because they are treated differently.
    # And the earliest versions of Unicode didn't include them, so add only if
    # there are some.
    my @blocks;
    push @blocks, $block->tables if defined $block;

    # Here, have the lists of tables constructed.  Process blocks last so that
    # if there are name collisions with them, blocks have lowest priority.
    # Should there ever be other collisions, manual intervention would be
    # required.  See the comments at the beginning of the program for a
    # possible way to handle those semi-automatically.
    foreach my $table (@tables,  @blocks) {

        # For non-binary properties, the synonym is just the name of the
        # table, like Greek, but for binary properties the synonym is the name
        # of the property, and means the code points in its 'Y' table.
        my $nominal = $table;
        my $nominal_property = $nominal->property;
        my $actual;
        if (! $nominal->isa('Property')) {
            $actual = $table;
        }
        else {

            # Here is a binary property.  Use the 'Y' table.  Verify that is
            # there
            my $yes = $nominal->table('Y');
            unless (defined $yes) {  # Must be defined, but is permissible to
                                     # be empty.
                Carp::my_carp_bug("Undefined $nominal, 'Y'.  Skipping.");
                next;
            }
            $actual = $yes;
        }

        foreach my $alias ($nominal->aliases) {

            # Attempt to create a table in the perl directory for the
            # candidate table, using whatever aliases in it that don't
            # conflict.  Also add non-conflicting aliases for all these
            # prefixed by 'Is_' (and/or 'In_' for Block property tables)
            PREFIX:
            foreach my $prefix ("", 'Is_', 'In_') {

                # Only Block properties can have added 'In_' aliases.
                next if $prefix eq 'In_' and $nominal_property != $block;

                my $proposed_name = $prefix . $alias->name;

                # No Is_Is, In_In, nor combinations thereof
                trace "$proposed_name is a no-no" if main::DEBUG && $to_trace && $proposed_name =~ /^ I [ns] _I [ns] _/x;
                next if $proposed_name =~ /^ I [ns] _I [ns] _/x;

                trace "Seeing if can add alias or table: 'perl=$proposed_name' based on $nominal" if main::DEBUG && $to_trace;

                # Get a reference to any existing table in the perl
                # directory with the desired name.
                my $pre_existing = $perl->table($proposed_name);

                if (! defined $pre_existing) {

                    # No name collision, so OK to add the perl synonym.

                    my $make_re_pod_entry;
                    my $ok_as_filename;
                    my $status = $alias->status;
                    if ($nominal_property == $block) {

                        # For block properties, only the compound form is
                        # preferred for external use; the others are
                        # discouraged.  The pod file contains wild cards for
                        # the 'In' and 'Is' forms so no entries for those; and
                        # we don't want people using the name without any
                        # prefix, so discourage that.
                        if ($prefix eq "") {
                            $make_re_pod_entry = 1;
                            $status = $status || $DISCOURAGED;
                            $ok_as_filename = 0;
                        }
                        elsif ($prefix eq 'In_') {
                            $make_re_pod_entry = 0;
                            $status = $status || $DISCOURAGED;
                            $ok_as_filename = 1;
                        }
                        else {
                            $make_re_pod_entry = 0;
                            $status = $status || $DISCOURAGED;
                            $ok_as_filename = 0;
                        }
                    }
                    elsif ($prefix ne "") {

                        # The 'Is' prefix is handled in the pod by a wild
                        # card, and we won't use it for an external name
                        $make_re_pod_entry = 0;
                        $status = $status || $NORMAL;
                        $ok_as_filename = 0;
                    }
                    else {

                        # Here, is an empty prefix, non block.  This gets its
                        # own pod entry and can be used for an external name.
                        $make_re_pod_entry = 1;
                        $status = $status || $NORMAL;
                        $ok_as_filename = 1;
                    }

                    # Here, there isn't a perl pre-existing table with the
                    # name.  Look through the list of equivalents of this
                    # table to see if one is a perl table.
                    foreach my $equivalent ($actual->leader->equivalents) {
                        next if $equivalent->property != $perl;

                        # Here, have found a table for $perl.  Add this alias
                        # to it, and are done with this prefix.
                        $equivalent->add_alias($proposed_name,
                                        Re_Pod_Entry => $make_re_pod_entry,

                                        # Currently don't output these in the
                                        # ucd pod, as are strongly discouraged
                                        # from being used
                                        UCD => 0,

                                        Status => $status,
                                        OK_as_Filename => $ok_as_filename);
                        trace "adding alias perl=$proposed_name to $equivalent" if main::DEBUG && $to_trace;
                        next PREFIX;
                    }

                    # Here, $perl doesn't already have a table that is a
                    # synonym for this property, add one.
                    my $added_table = $perl->add_match_table($proposed_name,
                                            Re_Pod_Entry => $make_re_pod_entry,

                                            # See UCD comment just above
                                            UCD => 0,

                                            Status => $status,
                                            OK_as_Filename => $ok_as_filename);
                    # And it will be related to the actual table, since it is
                    # based on it.
                    $added_table->set_equivalent_to($actual, Related => 1);
                    trace "added ", $perl->table($proposed_name) if main::DEBUG && $to_trace;
                    next;
                } # End of no pre-existing.

                # Here, there is a pre-existing table that has the proposed
                # name.  We could be in trouble, but not if this is just a
                # synonym for another table that we have already made a child
                # of the pre-existing one.
                if ($pre_existing->is_set_equivalent_to($actual)) {
                    trace "$pre_existing is already equivalent to $actual; adding alias perl=$proposed_name to it" if main::DEBUG && $to_trace;
                    $pre_existing->add_alias($proposed_name);
                    next;
                }

                # Here, there is a name collision, but it still could be OK if
                # the tables match the identical set of code points, in which
                # case, we can combine the names.  Compare each table's code
                # point list to see if they are identical.
                trace "Potential name conflict with $pre_existing having ", $pre_existing->count, " code points" if main::DEBUG && $to_trace;
                if ($pre_existing->matches_identically_to($actual)) {

                    # Here, they do match identically.  Not a real conflict.
                    # Make the perl version a child of the Unicode one, except
                    # in the non-obvious case of where the perl name is
                    # already a synonym of another Unicode property.  (This is
                    # excluded by the test for it being its own parent.)  The
                    # reason for this exclusion is that then the two Unicode
                    # properties become related; and we don't really know if
                    # they are or not.  We generate documentation based on
                    # relatedness, and this would be misleading.  Code
                    # later executed in the process will cause the tables to
                    # be represented by a single file anyway, without making
                    # it look in the pod like they are necessarily related.
                    if ($pre_existing->parent == $pre_existing
                        && ($pre_existing->property == $perl
                            || $actual->property == $perl))
                    {
                        trace "Setting $pre_existing equivalent to $actual since one is \$perl, and match identical sets" if main::DEBUG && $to_trace;
                        $pre_existing->set_equivalent_to($actual, Related => 1);
                    }
                    elsif (main::DEBUG && $to_trace) {
                        trace "$pre_existing is equivalent to $actual since match identical sets, but not setting them equivalent, to preserve the separateness of the perl aliases";
                        trace $pre_existing->parent;
                    }
                    next PREFIX;
                }

                # Here they didn't match identically, there is a real conflict
                # between our new name and a pre-existing property.
                $actual->add_conflicting($proposed_name, 'p', $pre_existing);
                $pre_existing->add_conflicting($nominal->full_name,
                                               'p',
                                               $actual);

                # Don't output a warning for aliases for the block
                # properties (unless they start with 'In_') as it is
                # expected that there will be conflicts and the block
                # form loses.
                if ($verbosity >= $NORMAL_VERBOSITY
                    && ($actual->property != $block || $prefix eq 'In_'))
                {
                    print simple_fold(join_lines(<<END
There is already an alias named $proposed_name (from $pre_existing),
so not creating this alias for $actual
END
                    ), "", 4);
                }

                # Keep track for documentation purposes.
                $has_In_conflicts++ if $prefix eq 'In_';
                $has_Is_conflicts++ if $prefix eq 'Is_';
            }
        }
    }

    # There are some properties which have No and Yes (and N and Y) as
    # property values, but aren't binary, and could possibly be confused with
    # binary ones.  So create caveats for them.  There are tables that are
    # named 'No', and tables that are named 'N', but confusion is not likely
    # unless they are the same table.  For example, N meaning Number or
    # Neutral is not likely to cause confusion, so don't add caveats to things
    # like them.
    foreach my $property (grep { $_->type != $BINARY
                                 && $_->type != $FORCED_BINARY }
                                                            property_ref('*'))
    {
        my $yes = $property->table('Yes');
        if (defined $yes) {
            my $y = $property->table('Y');
            if (defined $y && $yes == $y) {
                foreach my $alias ($property->aliases) {
                    $yes->add_conflicting($alias->name);
                }
            }
        }
        my $no = $property->table('No');
        if (defined $no) {
            my $n = $property->table('N');
            if (defined $n && $no == $n) {
                foreach my $alias ($property->aliases) {
                    $no->add_conflicting($alias->name, 'P');
                }
            }
        }
    }

    return;
}

sub register_file_for_name($table, $directory_ref, $file) {
    # Given info about a table and a datafile that it should be associated
    # with, register that association

    # $directory_ref    # Array of the directory path for the file
    # $file             # The file name in the final directory.

    trace "table=$table, file=$file, directory=@$directory_ref, fate=", $table->fate if main::DEBUG && $to_trace;

    if ($table->isa('Property')) {
        $table->set_file_path(@$directory_ref, $file);
        push @map_properties, $table;

        # No swash means don't do the rest of this.
        return if $table->fate != $ORDINARY
                  && ! ($table->name =~ /^_/ && $table->fate == $INTERNAL_ONLY);

        # Get the path to the file
        my @path = $table->file_path;

        # Use just the file name if no subdirectory.
        shift @path if $path[0] eq File::Spec->curdir();

        my $file = join '/', @path;

        # Create a hash entry for Unicode::UCD to get the file that stores this
        # property's map table
        foreach my $alias ($table->aliases) {
            my $name = $alias->name;
            if ($name =~ /^_/) {
                $strict_property_to_file_of{lc $name} = $file;
            }
            else {
                $loose_property_to_file_of{standardize($name)} = $file;
            }
        }

        # And a way for Unicode::UCD to find the proper key in the SwashInfo
        # hash for this property.
        $file_to_swash_name{$file} = "To" . $table->swash_name;
        return;
    }

    # Do all of the work for all equivalent tables when called with the leader
    # table, so skip if isn't the leader.
    return if $table->leader != $table;

    # If this is a complement of another file, use that other file instead,
    # with a ! prepended to it.
    my $complement;
    if (($complement = $table->complement) != 0) {
        my @directories = $complement->file_path;

        # This assumes that the 0th element is something like 'lib',
        # the 1th element the property name (in its own directory), like
        # 'AHex', and the 2th element the file like 'Y' which will have a .pl
        # appended to it later.
        $directories[1] =~ s/^/!/;
        $file = pop @directories;
        $directory_ref =\@directories;
    }

    # Join all the file path components together, using slashes.
    my $full_filename = join('/', @$directory_ref, $file);

    # All go in the same subdirectory of unicore, or the special
    # pseudo-directory '#'
    if ($directory_ref->[0] !~ / ^ $matches_directory | \# $ /x) {
        Carp::my_carp("Unexpected directory in "
                .  join('/', @{$directory_ref}, $file));
    }

    # For this table and all its equivalents ...
    foreach my $table ($table, $table->equivalents) {

        # Associate it with its file internally.  Don't include the
        # $matches_directory first component
        $table->set_file_path(@$directory_ref, $file);

        # No swash means don't do the rest of this.
        next if $table->isa('Map_Table') && $table->fate != $ORDINARY;

        my $sub_filename = join('/', $directory_ref->[1, -1], $file);

        my $property = $table->property;
        my $property_name = ($property == $perl)
                             ? ""  # 'perl' is never explicitly stated
                             : standardize($property->name) . '=';

        my $is_default = 0; # Is this table the default one for the property?

        # To calculate $is_default, we find if this table is the same as the
        # default one for the property.  But this is complicated by the
        # possibility that there is a master table for this one, and the
        # information is stored there instead of here.
        my $parent = $table->parent;
        my $leader_prop = $parent->property;
        my $default_map = $leader_prop->default_map;
        if (defined $default_map) {
            my $default_table = $leader_prop->table($default_map);
            $is_default = 1 if defined $default_table && $parent == $default_table;
        }

        # Calculate the loose name for this table.  Mostly it's just its name,
        # standardized.  But in the case of Perl tables that are single-form
        # equivalents to Unicode properties, it is the latter's name.
        my $loose_table_name =
                        ($property != $perl || $leader_prop == $perl)
                        ? standardize($table->name)
                        : standardize($parent->name);

        my $deprecated = ($table->status eq $DEPRECATED)
                         ? $table->status_info
                         : "";
        my $caseless_equivalent = $table->caseless_equivalent;

        # And for each of the table's aliases...  This inner loop eventually
        # goes through all aliases in the UCD that we generate regex match
        # files for
        foreach my $alias ($table->aliases) {
            my $standard = UCD_name($table, $alias);

            # Generate an entry in either the loose or strict hashes, which
            # will translate the property and alias names combination into the
            # file where the table for them is stored.
            if ($alias->loose_match) {
                if (exists $loose_to_file_of{$standard}) {
                    Carp::my_carp("Can't change file registered to $loose_to_file_of{$standard} to '$sub_filename'.");
                }
                else {
                    $loose_to_file_of{$standard} = $sub_filename;
                }
            }
            else {
                if (exists $stricter_to_file_of{$standard}) {
                    Carp::my_carp("Can't change file registered to $stricter_to_file_of{$standard} to '$sub_filename'.");
                }
                else {
                    $stricter_to_file_of{$standard} = $sub_filename;

                    # Tightly coupled with how Unicode::UCD works, for a
                    # floating point number that is a whole number, get rid of
                    # the trailing decimal point and 0's, so that Unicode::UCD
                    # will work.  Also note that this assumes that such a
                    # number is matched strictly; so if that were to change,
                    # this would be wrong.
                    if ((my $integer_name = $alias->name)
                            =~ s/^ ( -? \d+ ) \.0+ $ /$1/x)
                    {
                        $stricter_to_file_of{$property_name . $integer_name}
                                                            = $sub_filename;
                    }
                }
            }

            # For Unicode::UCD, create a mapping of the prop=value to the
            # canonical =value for that property.
            if ($standard =~ /=/) {

                # This could happen if a strict name mapped into an existing
                # loose name.  In that event, the strict names would have to
                # be moved to a new hash.
                if (exists($loose_to_standard_value{$standard})) {
                    Carp::my_carp_bug("'$standard' conflicts with a pre-existing use.  Bad News.  Continuing anyway");
                }
                $loose_to_standard_value{$standard} = $loose_table_name;
            }

            # Keep a list of the deprecated properties and their filenames
            if ($deprecated && $complement == 0) {
                $Unicode::UCD::why_deprecated{$sub_filename} = $deprecated;
            }

            # And a substitute table, if any, for case-insensitive matching
            if ($caseless_equivalent != 0) {
                $caseless_equivalent_to{$standard} = $caseless_equivalent;
            }

            # Add to defaults list if the table this alias belongs to is the
            # default one
            $loose_defaults{$standard} = 1 if $is_default;
        }
    }

    return;
}

{   # Closure
    my %base_names;  # Names already used for avoiding DOS 8.3 filesystem
                     # conflicts
    my %full_dir_name_of;   # Full length names of directories used.

    sub construct_filename($name, $mutable, $directories_ref) {
        # Return a file name for a table, based on the table name, but perhaps
        # changed to get rid of non-portable characters in it, and to make
        # sure that it is unique on a file system that allows the names before
        # any period to be at most 8 characters (DOS).  While we're at it
        # check and complain if there are any directory conflicts.

        # $name                 # The name to start with
        # $mutable              # Boolean: can it be changed?  If no, but
                                # yet it must be to work properly, a warning
                                # is given
        # $directories_ref      # A reference to an array containing the
                                # path to the file, with each element one path
                                # component.  This is used because the same
                                # name can be used in different directories.

        my $warn = ! defined wantarray;  # If true, then if the name is
                                # changed, a warning is issued as well.

        if (! defined $name) {
            Carp::my_carp("Undefined name in directory "
                          . File::Spec->join(@$directories_ref)
                          . ". '_' used");
            return '_';
        }

        # Make sure that no directory names conflict with each other.  Look at
        # each directory in the input file's path.  If it is already in use,
        # assume it is correct, and is merely being re-used, but if we
        # truncate it to 8 characters, and find that there are two directories
        # that are the same for the first 8 characters, but differ after that,
        # then that is a problem.
        foreach my $directory (@$directories_ref) {
            my $short_dir = substr($directory, 0, 8);
            if (defined $full_dir_name_of{$short_dir}) {
                next if $full_dir_name_of{$short_dir} eq $directory;
                Carp::my_carp("Directory $directory conflicts with directory $full_dir_name_of{$short_dir}.  Bad News.  Continuing anyway");
            }
            else {
                $full_dir_name_of{$short_dir} = $directory;
            }
        }

        my $path = join '/', @$directories_ref;
        $path .= '/' if $path;

        # Remove interior underscores.
        (my $filename = $name) =~ s/ (?<=.) _ (?=.) //xg;

        # Convert the dot in floating point numbers to an underscore
        $filename =~ s/\./_/ if $filename =~ / ^ \d+ \. \d+ $ /x;

        my $suffix = "";

        # Extract any suffix, delete any non-word character, and truncate to 3
        # after the dot
        if ($filename =~ m/ ( .*? ) ( \. .* ) /x) {
            $filename = $1;
            $suffix = $2;
            $suffix =~ s/\W+//g;
            substr($suffix, 4) = "" if length($suffix) > 4;
        }

        # Change any non-word character outside the suffix into an underscore,
        # and truncate to 8.
        $filename =~ s/\W+/_/g;   # eg., "L&" -> "L_"
        substr($filename, 8) = "" if length($filename) > 8;

        # Make sure the basename doesn't conflict with something we
        # might have already written. If we have, say,
        #     InGreekExtended1
        #     InGreekExtended2
        # they become
        #     InGreekE
        #     InGreek2
        my $warned = 0;
        while (my $num = $base_names{$path}{lc "$filename$suffix"}++) {
            $num++; # so basenames with numbers start with '2', which
                    # just looks more natural.

            # Want to append $num, but if it'll make the basename longer
            # than 8 characters, pre-truncate $filename so that the result
            # is acceptable.
            my $delta = length($filename) + length($num) - 8;
            if ($delta > 0) {
                substr($filename, -$delta) = $num;
            }
            else {
                $filename .= $num;
            }
            if ($warn && ! $warned) {
                $warned = 1;
                Carp::my_carp("'$path$name' conflicts with another name on a filesystem with 8 significant characters (like DOS).  Proceeding anyway.");
            }
        }

        return $filename if $mutable;

        # If not changeable, must return the input name, but warn if needed to
        # change it beyond shortening it.
        if ($name ne $filename
            && substr($name, 0, length($filename)) ne $filename) {
            Carp::my_carp("'$path$name' had to be changed into '$filename'.  Bad News.  Proceeding anyway.");
        }
        return $name;
    }
}

# The pod file contains a very large table.  Many of the lines in that table
# would exceed a typical output window's size, and so need to be wrapped with
# a hanging indent to make them look good.  The pod language is really
# insufficient here.  There is no general construct to do that in pod, so it
# is done here by beginning each such line with a space to cause the result to
# be output without formatting, and doing all the formatting here.  This leads
# to the result that if the eventual display window is too narrow it won't
# look good, and if the window is too wide, no advantage is taken of that
# extra width.  A further complication is that the output may be indented by
# the formatter so that there is less space than expected.  What I (khw) have
# done is to assume that that indent is a particular number of spaces based on
# what it is in my Linux system;  people can always resize their windows if
# necessary, but this is obviously less than desirable, but the best that can
# be expected.
my $automatic_pod_indent = 8;

# Try to format so that uses fewest lines, but few long left column entries
# slide into the right column.  An experiment on 5.1 data yielded the
# following percentages that didn't cut into the other side along with the
# associated first-column widths
# 69% = 24
# 80% not too bad except for a few blocks
# 90% = 33; # , cuts 353/3053 lines from 37 = 12%
# 95% = 37;
my $indent_info_column = 27;    # 75% of lines didn't have overlap

my $FILLER = 3;     # Length of initial boiler-plate columns in a pod line
                    # The 3 is because of:
                    #   1   for the leading space to tell the pod formatter to
                    #       output as-is
                    #   1   for the flag
                    #   1   for the space between the flag and the main data

sub format_pod_line($first_column_width, $entry, $info, $status = "", $loose_match = 1 ) {
    # Take a pod line and return it, formatted properly

    # $entry Contents of left column
    # $info Contents of right column

    my $flags = "";
    $flags .= $STRICTER if ! $loose_match;

    $flags .= $status if $status;

    # There is a blank in the left column to cause the pod formatter to
    # output the line as-is.
    return sprintf " %-*s%-*s %s\n",
                    # The first * in the format is replaced by this, the -1 is
                    # to account for the leading blank.  There isn't a
                    # hard-coded blank after this to separate the flags from
                    # the rest of the line, so that in the unlikely event that
                    # multiple flags are shown on the same line, they both
                    # will get displayed at the expense of that separation,
                    # but since they are left justified, a blank will be
                    # inserted in the normal case.
                    $FILLER - 1,
                    $flags,

                    # The other * in the format is replaced by this number to
                    # cause the first main column to right fill with blanks.
                    # The -1 is for the guaranteed blank following it.
                    $first_column_width - $FILLER - 1,
                    $entry,
                    $info;
}

my @zero_match_tables;  # List of tables that have no matches in this release

sub make_re_pod_entries($input_table) {
    # This generates the entries for the pod file for a given table.
    # Also done at this time are any children tables.  The output looks like:
    # \p{Common}              \p{Script=Common} (Short: \p{Zyyy}) (5178)

    # Generate parent and all its children at the same time.
    return if $input_table->parent != $input_table;

    my $property = $input_table->property;
    my $type = $property->type;
    my $full_name = $property->full_name;

    my $count = $input_table->count;
    my $unicode_count;
    my $non_unicode_string;
    if ($count > $MAX_UNICODE_CODEPOINTS) {
        $unicode_count = $count - ($MAX_WORKING_CODEPOINT
                                    - $MAX_UNICODE_CODEPOINT);
        $non_unicode_string = " plus all above-Unicode code points";
    }
    else {
        $unicode_count = $count;
        $non_unicode_string = "";
    }

    my $string_count = clarify_number($unicode_count) . $non_unicode_string;

    my $definition = $input_table->calculate_table_definition;
    if ($definition) {

        # Save the definition for later use.
        $input_table->set_definition($definition);

        $definition = ": $definition";
    }

    my $status = $input_table->status;
    my $status_info = $input_table->status_info;
    my $caseless_equivalent = $input_table->caseless_equivalent;

    # Don't mention a placeholder equivalent as it isn't to be listed in the
    # pod
    $caseless_equivalent = 0 if $caseless_equivalent != 0
                                && $caseless_equivalent->fate > $ORDINARY;

    my $entry_for_first_table; # The entry for the first table output.
                           # Almost certainly, it is the parent.

    # For each related table (including itself), we will generate a pod entry
    # for each name each table goes by
    foreach my $table ($input_table, $input_table->children) {

        # Unicode::UCD cannot deal with null string property values, so skip
        # any tables that have no non-null names.
        next if ! grep { $_->name ne "" } $table->aliases;

        # First, gather all the info that applies to this table as a whole.

        push @zero_match_tables, $table if $count == 0
                                            # Don't mention special tables
                                            # as being zero length
                                           && $table->fate == $ORDINARY;

        my $table_property = $table->property;

        # The short name has all the underscores removed, while the full name
        # retains them.  Later, we decide whether to output a short synonym
        # for the full one, we need to compare apples to apples, so we use the
        # short name's length including underscores.
        my $table_property_short_name_length;
        my $table_property_short_name
            = $table_property->short_name(\$table_property_short_name_length);
        my $table_property_full_name = $table_property->full_name;

        # Get how much savings there is in the short name over the full one
        # (delta will always be <= 0)
        my $table_property_short_delta = $table_property_short_name_length
                                         - length($table_property_full_name);
        my @table_description = $table->description;
        my @table_note = $table->note;

        # Generate an entry for each alias in this table.
        my $entry_for_first_alias;  # saves the first one encountered.
        foreach my $alias ($table->aliases) {

            # Skip if not to go in pod.
            next unless $alias->make_re_pod_entry;

            # Start gathering all the components for the entry
            my $name = $alias->name;

            # Skip if name is empty, as can't be accessed by regexes.
            next if $name eq "";

            my $entry;      # Holds the left column, may include extras
            my $entry_ref;  # To refer to the left column's contents from
                            # another entry; has no extras

            # First the left column of the pod entry.  Tables for the $perl
            # property always use the single form.
            if ($table_property == $perl) {
                $entry = "\\p{$name}";
                $entry .= " \\p$name" if length $name == 1; # Show non-braced
                                                            # form too
                $entry_ref = "\\p{$name}";
            }
            else {    # Compound form.

                # Only generate one entry for all the aliases that mean true
                # or false in binary properties.  Append a '*' to indicate
                # some are missing.  (The heading comment notes this.)
                my $rhs;
                if ($type == $BINARY) {
                    next if $name ne 'N' && $name ne 'Y';
                    $rhs = "$name*";
                }
                elsif ($type != $FORCED_BINARY) {
                    $rhs = $name;
                }
                else {

                    # Forced binary properties require special handling.  It
                    # has two sets of tables, one set is true/false; and the
                    # other set is everything else.  Entries are generated for
                    # each set.  Use the Bidi_Mirrored property (which appears
                    # in all Unicode versions) to get a list of the aliases
                    # for the true/false tables.  Of these, only output the N
                    # and Y ones, the same as, a regular binary property.  And
                    # output all the rest, same as a non-binary property.
                    my $bm = property_ref("Bidi_Mirrored");
                    if ($name eq 'N' || $name eq 'Y') {
                        $rhs = "$name*";
                    } elsif (grep { $name eq $_->name } $bm->table("Y")->aliases,
                                                        $bm->table("N")->aliases)
                    {
                        next;
                    }
                    else {
                        $rhs = $name;
                    }
                }

                # Colon-space is used to give a little more space to be easier
                # to read;
                $entry = "\\p{"
                        . $table_property_full_name
                        . ": $rhs}";

                # But for the reference to this entry, which will go in the
                # right column, where space is at a premium, use equals
                # without a space
                $entry_ref = "\\p{" . $table_property_full_name . "=$name}";
            }

            # Then the right (info) column.  This is stored as components of
            # an array for the moment, then joined into a string later.  For
            # non-internal only properties, begin the info with the entry for
            # the first table we encountered (if any), as things are ordered
            # so that that one is the most descriptive.  This leads to the
            # info column of an entry being a more descriptive version of the
            # name column
            my @info;
            if ($name =~ /^_/) {
                push @info,
                        '(For internal use by Perl, not necessarily stable)';
            }
            elsif ($entry_for_first_alias) {
                push @info, $entry_for_first_alias;
            }

            # If this entry is equivalent to another, add that to the info,
            # using the first such table we encountered
            if ($entry_for_first_table) {
                if (@info) {
                    push @info, "(= $entry_for_first_table)";
                }
                else {
                    push @info, $entry_for_first_table;
                }
            }

            # If the name is a large integer, add an equivalent with an
            # exponent for better readability
            if ($name =~ /^[+-]?[\d]+$/ && $name >= 10_000) {
                push @info, sprintf "(= %.1e)", $name
            }

            my $parenthesized = "";
            if (! $entry_for_first_alias) {

                # This is the first alias for the current table.  The alias
                # array is ordered so that this is the fullest, most
                # descriptive alias, so it gets the fullest info.  The other
                # aliases are mostly merely pointers to this one, using the
                # information already added above.

                # Display any status message, but only on the parent table
                if ($status && ! $entry_for_first_table) {
                    push @info, $status_info;
                }

                # Put out any descriptive info
                if (@table_description || @table_note) {
                    push @info, join "; ", @table_description, @table_note;
                }

                # Look to see if there is a shorter name we can point people
                # at
                my $standard_name = standardize($name);
                my $short_name;
                my $proposed_short = $table->short_name;
                if (defined $proposed_short) {
                    my $standard_short = standardize($proposed_short);

                    # If the short name is shorter than the standard one, or
                    # even if it's not, but the combination of it and its
                    # short property name (as in \p{prop=short} ($perl doesn't
                    # have this form)) saves at least two characters, then,
                    # cause it to be listed as a shorter synonym.
                    if (length $standard_short < length $standard_name
                        || ($table_property != $perl
                            && (length($standard_short)
                                - length($standard_name)
                                + $table_property_short_delta)  # (<= 0)
                                < -2))
                    {
                        $short_name = $proposed_short;
                        if ($table_property != $perl) {
                            $short_name = $table_property_short_name
                                          . "=$short_name";
                        }
                        $short_name = "\\p{$short_name}";
                    }
                }

                # And if this is a compound form name, see if there is a
                # single form equivalent
                my $single_form;
                if ($table_property != $perl && $table_property != $block) {

                    # Special case the binary N tables, so that will print
                    # \P{single}, but use the Y table values to populate
                    # 'single', as we haven't likewise populated the N table.
                    # For forced binary tables, we can't just look at the N
                    # table, but must see if this table is equivalent to the N
                    # one, as there are two equivalent beasts in these
                    # properties.
                    my $test_table;
                    my $p;
                    if (   ($type == $BINARY
                            && $input_table == $property->table('No'))
                        || ($type == $FORCED_BINARY
                            && $property->table('No')->
                                        is_set_equivalent_to($input_table)))
                    {
                        $test_table = $property->table('Yes');
                        $p = 'P';
                    }
                    else {
                        $test_table = $input_table;
                        $p = 'p';
                    }

                    # Look for a single form amongst all the children.
                    foreach my $table ($test_table->children) {
                        next if $table->property != $perl;
                        my $proposed_name = $table->short_name;
                        next if ! defined $proposed_name;

                        # Don't mention internal-only properties as a possible
                        # single form synonym
                        next if substr($proposed_name, 0, 1) eq '_';

                        $proposed_name = "\\$p\{$proposed_name}";
                        if (! defined $single_form
                            || length($proposed_name) < length $single_form)
                        {
                            $single_form = $proposed_name;

                            # The goal here is to find a single form; not the
                            # shortest possible one.  We've already found a
                            # short name.  So, stop at the first single form
                            # found, which is likely to be closer to the
                            # original.
                            last;
                        }
                    }
                }

                # Output both short and single in the same parenthesized
                # expression, but with only one of 'Single', 'Short' if there
                # are both items.
                if ($short_name || $single_form || $table->conflicting) {
                    $parenthesized .= "Short: $short_name" if $short_name;
                    if ($short_name && $single_form) {
                        $parenthesized .= ', ';
                    }
                    elsif ($single_form) {
                        $parenthesized .= 'Single: ';
                    }
                    $parenthesized .= $single_form if $single_form;
                }
            }

            if ($caseless_equivalent != 0) {
                $parenthesized .=  '; ' if $parenthesized ne "";
                $parenthesized .= "/i= " . $caseless_equivalent->complete_name;
            }


            # Warn if this property isn't the same as one that a
            # semi-casual user might expect.  The other components of this
            # parenthesized structure are calculated only for the first entry
            # for this table, but the conflicting is deemed important enough
            # to go on every entry.
            my $conflicting = join " NOR ", $table->conflicting;
            if ($conflicting) {
                $parenthesized .=  '; ' if $parenthesized ne "";
                $parenthesized .= "NOT $conflicting";
            }

            push @info, "($parenthesized)" if $parenthesized;

            if ($name =~ /_$/ && $alias->loose_match) {
                push @info, "Note the trailing '_' matters in spite of loose matching rules.";
            }

            if ($table_property != $perl && $table->perl_extension) {
                push @info, '(Perl extension)';
            }
            my $definition = $table->definition // "";
            $definition = "" if $entry_for_first_alias;
            $definition = ": $definition" if $definition;
            push @info, "($string_count$definition)";

            # Now, we have both the entry and info so add them to the
            # list of all the properties.
            push @match_properties,
                format_pod_line($indent_info_column,
                                $entry,
                                join( " ", @info),
                                $alias->status,
                                $alias->loose_match);

            $entry_for_first_alias = $entry_ref unless $entry_for_first_alias;
        } # End of looping through the aliases for this table.

        if (! $entry_for_first_table) {
            $entry_for_first_table = $entry_for_first_alias;
        }
    } # End of looping through all the related tables
    return;
}

sub make_ucd_table_pod_entries($table) {
    # Generate the entries for the UCD section of the pod for $table.  This
    # also calculates if names are ambiguous, so has to be called even if the
    # pod is not being output

    my $short_name = $table->name;
    my $standard_short_name = standardize($short_name);
    my $full_name = $table->full_name;
    my $standard_full_name = standardize($full_name);

    my $full_info = "";     # Text of info column for full-name entries
    my $other_info = "";    # Text of info column for short-name entries
    my $short_info = "";    # Text of info column for other entries
    my $meaning = "";       # Synonym of this table

    my $property = ($table->isa('Property'))
                   ? $table
                   : $table->parent->property;

    my $perl_extension = $table->perl_extension;
    my $is_perl_extension_match_table_but_not_dollar_perl
                                                        = $property != $perl
                                                       && $perl_extension
                                                       && $property != $table;

    # Get the more official name for perl extensions that aren't
    # stand-alone properties
    if ($is_perl_extension_match_table_but_not_dollar_perl) {
        if ($property->type == $BINARY) {
            $meaning = $property->full_name;
        }
        else {
            $meaning = $table->parent->complete_name;
        }
    }

    # There are three types of info column.  One for the short name, one for
    # the full name, and one for everything else.  They mostly are the same,
    # so initialize in the same loop.

    foreach my $info_ref (\$full_info, \$short_info, \$other_info) {
        if ($info_ref != \$full_info) {

            # The non-full name columns include the full name
            $$info_ref .= $full_name;
        }


        if ($is_perl_extension_match_table_but_not_dollar_perl) {

            # Add the synonymous name for the non-full name entries; and to
            # the full-name entry if it adds extra information
            if (   standardize($meaning) ne $standard_full_name
                || $info_ref == \$other_info
                || $info_ref == \$short_info)
            {
                my $parenthesized =  $info_ref != \$full_info;
                $$info_ref .= " " if $$info_ref && $parenthesized;
                $$info_ref .= "(=" if $parenthesized;
                $$info_ref .= "$meaning";
                $$info_ref .= ")" if $parenthesized;
                $$info_ref .= ".";
            }
        }

        # And the full-name entry includes the short name, if shorter
        if ($info_ref == \$full_info
            && length $standard_short_name < length $standard_full_name)
        {
            $full_info =~ s/\.\Z//;
            $full_info .= "  " if $full_info;
            $full_info .= "(Short: $short_name)";
        }

        if ($table->perl_extension) {
            $$info_ref =~ s/\.\Z//;
            $$info_ref .= ".  " if $$info_ref;
            $$info_ref .= "(Perl extension)";
        }
    }

    my $definition;
    my $definition_table;
    my $type = $table->property->type;
    if ($type == $BINARY || $type == $FORCED_BINARY) {
        $definition_table = $table->property->table('Y');
    }
    elsif ($table->isa('Match_Table')) {
        $definition_table = $table;
    }

    $definition = $definition_table->calculate_table_definition
                                            if defined $definition_table
                                                    && $definition_table != 0;

    # Add any extra annotations to the full name entry
    foreach my $more_info ($table->description,
                            $definition,
                            $table->note,
                            $table->status_info)
    {
        next unless $more_info;
        $full_info =~ s/\.\Z//;
        $full_info .= ".  " if $full_info;
        $full_info .= $more_info;
    }
    if ($table->property->type == $FORCED_BINARY) {
        if ($full_info) {
            $full_info =~ s/\.\Z//;
            $full_info .= ".  ";
        }
        $full_info .= "This is a combination property which has both:"
                    . " 1) a map to various string values; and"
                    . " 2) a map to boolean Y/N, where 'Y' means the"
                    . " string value is non-empty.  Add the prefix 'is'"
                    . " to the prop_invmap() call to get the latter";
    }

    # These keep track if have created full and short name pod entries for the
    # property
    my $done_full = 0;
    my $done_short = 0;

    # Every possible name is kept track of, even those that aren't going to be
    # output.  This way we can be sure to find the ambiguities.
    foreach my $alias ($table->aliases) {
        my $name = $alias->name;
        my $standard = standardize($name);
        my $info;
        my $output_this = $alias->ucd;

        # If the full and short names are the same, we want to output the full
        # one's entry, so it has priority.
        if ($standard eq $standard_full_name) {
            next if $done_full;
            $done_full = 1;
            $info = $full_info;
        }
        elsif ($standard eq $standard_short_name) {
            next if $done_short;
            $done_short = 1;
            next if $standard_short_name eq $standard_full_name;
            $info = $short_info;
        }
        else {
            $info = $other_info;
        }

        $combination_property{$standard} = 1
                                  if $table->property->type == $FORCED_BINARY;

        # Here, we have set up the two columns for this entry.  But if an
        # entry already exists for this name, we have to decide which one
        # we're going to later output.
        if (exists $ucd_pod{$standard}) {

            # If the two entries refer to the same property, it's not going to
            # be ambiguous.  (Likely it's because the names when standardized
            # are the same.)  But that means if they are different properties,
            # there is ambiguity.
            if ($ucd_pod{$standard}->{'property'} != $property) {

                # Here, we have an ambiguity.  This code assumes that one is
                # scheduled to be output and one not and that one is a perl
                # extension (which is not to be output) and the other isn't.
                # If those assumptions are wrong, things have to be rethought.
                if ($ucd_pod{$standard}{'output_this'} == $output_this
                    || $ucd_pod{$standard}{'perl_extension'} == $perl_extension
                    || $output_this == $perl_extension)
                {
                    Carp::my_carp("Bad news.  $property and $ucd_pod{$standard}->{'property'} have unexpected output status and perl-extension combinations.  Proceeding anyway.");
                }

                # We modify the info column of the one being output to
                # indicate the ambiguity.  Set $which to point to that one's
                # info.
                my $which;
                if ($ucd_pod{$standard}{'output_this'}) {
                    $which = \$ucd_pod{$standard}->{'info'};
                }
                else {
                    $which = \$info;
                    $meaning = $ucd_pod{$standard}{'meaning'};
                }

                chomp $$which;
                $$which =~ s/\.\Z//;
                $$which .= "; NOT '$standard' meaning '$meaning'";

                $ambiguous_names{$standard} = 1;
            }

            # Use the non-perl-extension variant
            next unless $ucd_pod{$standard}{'perl_extension'};
        }

        # Store enough information about this entry that we can later look for
        # ambiguities, and output it properly.
        $ucd_pod{$standard} = { 'name' => $name,
                                'info' => $info,
                                'meaning' => $meaning,
                                'output_this' => $output_this,
                                'perl_extension' => $perl_extension,
                                'property' => $property,
                                'status' => $alias->status,
        };
    } # End of looping through all this table's aliases

    return;
}

sub pod_alphanumeric_sort {
    # Sort pod entries alphanumerically.

    # The first few character columns are filler, plus the '\p{'; and get rid
    # of all the trailing stuff, starting with the trailing '}', so as to sort
    # on just 'Name=Value'
    (my $a = lc $a) =~ s/^ .*? \{ //x;
    $a =~ s/}.*//;
    (my $b = lc $b) =~ s/^ .*? \{ //x;
    $b =~ s/}.*//;

    # Determine if the two operands are both internal only or both not.
    # Character 0 should be a '\'; 1 should be a p; 2 should be '{', so 3
    # should be the underscore that begins internal only
    my $a_is_internal = (substr($a, 0, 1) eq '_');
    my $b_is_internal = (substr($b, 0, 1) eq '_');

    # Sort so the internals come last in the table instead of first (which the
    # leading underscore would otherwise indicate).
    if ($a_is_internal != $b_is_internal) {
        return 1 if $a_is_internal;
        return -1
    }

    # Determine if the two operands are compound or not, and if so if are
    # "numeric" property values or not, like \p{Age: 3.0}.  But there are also
    # things like \p{Canonical_Combining_Class: CCC133} and \p{Age: V10_0},
    # all of which this considers numeric, and for sorting, looks just at the
    # numeric parts.  It can also be a rational like \p{Numeric Value=-1/2}.
    my $split_re = qr/
        ^ ( [^:=]+ ) # $1 is undef if not a compound form, otherwise is the
                     # property name
        [:=] \s*     # The syntax for the compound form
        (?:          # followed by ...
            (        # $2 gets defined if what follows is a "numeric"
                     # expression, which is ...
              ( -? \d+ (?: [.\/] \d+)?  # An integer, float, or rational
                                        # number, optionally signed
               | [[:alpha:]]{2,} \d+ $ ) # or something like CCC131.  Either
                                         # of these go into $3
             | ( V \d+ _ \d+ )           # or a Unicode's Age property version
                                         # number, into $4
            )
            | .* $    # If not "numeric", accept anything so that $1 gets
                      # defined if it is any compound form
        ) /ix;
    my ($a_initial, $a_numeric, $a_number, $a_version) = ($a =~ $split_re);
    my ($b_initial, $b_numeric, $b_number, $b_version) = ($b =~ $split_re);

    # Sort alphabeticlly on the whole property name if either operand isn't
    # compound, or they differ.
    return $a cmp $b if   ! defined $a_initial
                       || ! defined $b_initial
                       || $a_initial ne $b_initial;

    if (! defined $a_numeric) {

        # If neither is numeric, use alpha sort
        return $a cmp $b if ! defined $b_numeric;
        return 1;  # Sort numeric ahead of alpha
    }

    # Here $a is numeric
    return -1 if ! defined $b_numeric;  # Numeric sorts before alpha

    # Here they are both numeric in the same property.
    # Convert version numbers into regular numbers
    if (defined $a_version) {
        ($a_number = $a_version) =~ s/^V//i;
        $a_number =~ s/_/./;
    }
    else {  # Otherwise get rid of the, e.g., CCC in CCC9 */
        $a_number =~ s/ ^ [[:alpha:]]+ //x;
    }
    if (defined $b_version) {
        ($b_number = $b_version) =~ s/^V//i;
        $b_number =~ s/_/./;
    }
    else {
        $b_number =~ s/ ^ [[:alpha:]]+ //x;
    }

    # Convert rationals to floating for the comparison.
    $a_number = eval $a_number if $a_number =~ qr{/};
    $b_number = eval $b_number if $b_number =~ qr{/};

    return $a_number <=> $b_number || $a cmp $b;
}

sub make_pod () {
    # Create the .pod file.  This generates the various subsections and then
    # combines them in one big HERE document.

    my $Is_flags_text = "If an entry has flag(s) at its beginning, like \"$DEPRECATED\", the \"Is_\" form has the same flag(s)";

    return unless defined $pod_directory;
    print "Making pod file\n" if $verbosity >= $PROGRESS;

    my $exception_message =
    '(Any exceptions are individually noted beginning with the word NOT.)';
    my @block_warning;
    if (-e 'Blocks.txt') {

        # Add the line: '\p{In_*}    \p{Block: *}', with the warning message
        # if the global $has_In_conflicts indicates we have them.
        push @match_properties, format_pod_line($indent_info_column,
                                                '\p{In_*}',
                                                '\p{Block: *}'
                                                    . (($has_In_conflicts)
                                                      ? " $exception_message"
                                                      : ""),
                                                 $DISCOURAGED);
        @block_warning = << "END";

In particular, matches in the Block property have single forms
defined by Perl that begin with C<"In_">, C<"Is_>, or even with no prefix at
all,  Like all B<DISCOURAGED> forms, these are not stable.  For example,
C<\\p{Block=Deseret}> can currently be written as C<\\p{In_Deseret}>,
C<\\p{Is_Deseret}>, or C<\\p{Deseret}>.  But, a new Unicode version may
come along that would force Perl to change the meaning of one or more of
these, and your program would no longer be correct.  Currently there are no
such conflicts with the form that begins C<"In_">, but there are many with the
other two shortcuts, and Unicode continues to define new properties that begin
with C<"In">, so it's quite possible that a conflict will occur in the future.
The compound form is guaranteed to not become obsolete, and its meaning is
clearer anyway.  See L<perlunicode/"Blocks"> for more information about this.

User-defined properties must begin with "In" or "Is".  These override any
Unicode property of the same name.
END
    }
    my $text = $Is_flags_text;
    $text = "$exception_message $text" if $has_Is_conflicts;

    # And the 'Is_ line';
    push @match_properties, format_pod_line($indent_info_column,
                                            '\p{Is_*}',
                                            "\\p{*} $text");
    push @match_properties, format_pod_line($indent_info_column,
            '\p{Name=*}',
            "Combination of Name and Name_Alias properties; has special"
          . " loose matching rules, for which see Unicode UAX #44");
    push @match_properties, format_pod_line($indent_info_column,
                                            '\p{Na=*}',
                                            '\p{Name=*}');

    # Sort the properties array for output.  It is sorted alphabetically
    # except numerically for numeric properties, and only output unique lines.
    @match_properties = sort pod_alphanumeric_sort uniques @match_properties;

    my $formatted_properties = simple_fold(\@match_properties,
                                        "",
                                        # indent succeeding lines by two extra
                                        # which looks better
                                        $indent_info_column + 2,

                                        # shorten the line length by how much
                                        # the formatter indents, so the folded
                                        # line will fit in the space
                                        # presumably available
                                        $automatic_pod_indent);
    # Add column headings, indented to be a little more centered, but not
    # exactly
    $formatted_properties =  format_pod_line($indent_info_column,
                                                    '    NAME',
                                                    '           INFO')
                                    . "\n"
                                    . $formatted_properties;

    # Generate pod documentation lines for the tables that match nothing
    my $zero_matches = "";
    if (@zero_match_tables) {
        @zero_match_tables = uniques(@zero_match_tables);
        $zero_matches = join "\n\n",
                        map { $_ = '=item \p{' . $_->complete_name . "}" }
                            sort { $a->complete_name cmp $b->complete_name }
                            @zero_match_tables;

        $zero_matches = <<END;

=head2 Legal C<\\p{}> and C<\\P{}> constructs that match no characters

Unicode has some property-value pairs that currently don't match anything.
This happens generally either because they are obsolete, or they exist for
symmetry with other forms, but no language has yet been encoded that uses
them.  In this version of Unicode, the following match zero code points:

=over 4

$zero_matches

=back

END
    }

    # Generate list of properties that we don't accept, grouped by the reasons
    # why.  This is so only put out the 'why' once, and then list all the
    # properties that have that reason under it.

    my %why_list;   # The keys are the reasons; the values are lists of
                    # properties that have the key as their reason

    # For each property, add it to the list that are suppressed for its reason
    # The sort will cause the alphabetically first properties to be added to
    # each list first, so each list will be sorted.
    foreach my $property (sort keys %why_suppressed) {
        next unless $why_suppressed{$property};
        push @{$why_list{$why_suppressed{$property}}}, $property;
    }

    # For each reason (sorted by the first property that has that reason)...
    my @bad_re_properties;
    foreach my $why (sort { $why_list{$a}->[0] cmp $why_list{$b}->[0] }
                     keys %why_list)
    {
        # Add to the output, all the properties that have that reason.
        my $has_item = 0;   # Flag if actually output anything.
        foreach my $name (@{$why_list{$why}}) {

            # Split compound names into $property and $table components
            my $property = $name;
            my $table;
            if ($property =~ / (.*) = (.*) /x) {
                $property = $1;
                $table = $2;
            }

            # This release of Unicode may not have a property that is
            # suppressed, so don't reference a non-existent one.
            $property = property_ref($property);
            next if ! defined $property;

            # And since this list is only for match tables, don't list the
            # ones that don't have match tables.
            next if ! $property->to_create_match_tables;

            # Find any abbreviation, and turn it into a compound name if this
            # is a property=value pair.
            my $short_name = $property->name;
            $short_name .= '=' . $property->table($table)->name if $table;

            # Start with an empty line.
            push @bad_re_properties, "\n\n" unless $has_item;

            # And add the property as an item for the reason.
            push @bad_re_properties, "\n=item I<$name> ($short_name)\n";
            $has_item = 1;
        }

        # And add the reason under the list of properties, if such a list
        # actually got generated.  Note that the header got added
        # unconditionally before.  But pod ignores extra blank lines, so no
        # harm.
        push @bad_re_properties, "\n$why\n" if $has_item;

    } # End of looping through each reason.

    if (! @bad_re_properties) {
        push @bad_re_properties,
                "*** This installation accepts ALL non-Unihan properties ***";
    }
    else {
        # Add =over only if non-empty to avoid an empty =over/=back section,
        # which is considered bad form.
        unshift @bad_re_properties, "\n=over 4\n";
        push @bad_re_properties, "\n=back\n";
    }

    # Similarly, generate a list of files that we don't use, grouped by the
    # reasons why (Don't output if the reason is empty).  First, create a hash
    # whose keys are the reasons, and whose values are anonymous arrays of all
    # the files that share that reason.
    my %grouped_by_reason;
    foreach my $file (keys %skipped_files) {
        next unless $skipped_files{$file};
        push @{$grouped_by_reason{$skipped_files{$file}}}, $file;
    }

    # Then, sort each group.
    foreach my $group (keys %grouped_by_reason) {
        @{$grouped_by_reason{$group}} = sort { lc $a cmp lc $b }
                                        @{$grouped_by_reason{$group}} ;
    }

    # Finally, create the output text.  For each reason (sorted by the
    # alphabetically first file that has that reason)...
    my @unused_files;
    foreach my $reason (sort { lc $grouped_by_reason{$a}->[0]
                               cmp lc $grouped_by_reason{$b}->[0]
                              }
                         keys %grouped_by_reason)
    {
        # Add all the files that have that reason to the output.  Start
        # with an empty line.
        push @unused_files, "\n\n";
        push @unused_files, map { "\n=item F<$_> \n" }
                            @{$grouped_by_reason{$reason}};
        # And add the reason under the list of files
        push @unused_files, "\n$reason\n";
    }

    # Similarly, create the output text for the UCD section of the pod
    my @ucd_pod;
    foreach my $key (keys %ucd_pod) {
        next unless $ucd_pod{$key}->{'output_this'};
        push @ucd_pod, format_pod_line($indent_info_column,
                                       $ucd_pod{$key}->{'name'},
                                       $ucd_pod{$key}->{'info'},
                                       $ucd_pod{$key}->{'status'},
                                      );
    }

    # Sort alphabetically, and fold for output
    @ucd_pod = sort { lc substr($a, 2) cmp lc substr($b, 2) } @ucd_pod;
    my $ucd_pod = simple_fold(\@ucd_pod,
                           ' ',
                           $indent_info_column,
                           $automatic_pod_indent);
    $ucd_pod =  format_pod_line($indent_info_column, 'NAME', '  INFO')
                . "\n"
                . $ucd_pod;
    my $space_hex = sprintf("%02x", ord " ");
    local $" = "";

    # Everything is ready to assemble.
    my @OUT = << "END";
=begin comment

$HEADER

To change this file, edit $0 instead.

=end comment

=head1 NAME

$pod_file - Index of Unicode Version $unicode_version character properties in Perl

=head1 DESCRIPTION

This document provides information about the portion of the Unicode database
that deals with character properties, that is the portion that is defined on
single code points.  (L</Other information in the Unicode data base>
below briefly mentions other data that Unicode provides.)

Perl can provide access to all non-provisional Unicode character properties,
though not all are enabled by default.  The omitted ones are the Unihan
properties and certain
deprecated or Unicode-internal properties.  (An installation may choose to
recompile Perl's tables to change this.  See L</Unicode character
properties that are NOT accepted by Perl>.)

For most purposes, access to Unicode properties from the Perl core is through
regular expression matches, as described in the next section.
For some special purposes, and to access the properties that are not suitable
for regular expression matching, all the Unicode character properties that
Perl handles are accessible via the standard L<Unicode::UCD> module, as
described in the section L</Properties accessible through Unicode::UCD>.

Perl also provides some additional extensions and short-cut synonyms
for Unicode properties.

This document merely lists all available properties and does not attempt to
explain what each property really means.  There is a brief description of each
Perl extension; see L<perlunicode/Other Properties> for more information on
these.  There is some detail about Blocks, Scripts, General_Category,
and Bidi_Class in L<perlunicode>, but to find out about the intricacies of the
official Unicode properties, refer to the Unicode standard.  A good starting
place is L<$unicode_reference_url>.

Note that you can define your own properties; see
L<perlunicode/"User-Defined Character Properties">.

=head1 Properties accessible through C<\\p{}> and C<\\P{}>

The Perl regular expression C<\\p{}> and C<\\P{}> constructs give access to
most of the Unicode character properties.  The table below shows all these
constructs, both single and compound forms.

B<Compound forms> consist of two components, separated by an equals sign or a
colon.  The first component is the property name, and the second component is
the particular value of the property to match against, for example,
C<\\p{Script_Extensions: Greek}> and C<\\p{Script_Extensions=Greek}> both mean
to match characters whose Script_Extensions property value is Greek.
(C<Script_Extensions> is an improved version of the C<Script> property.)

B<Single forms>, like C<\\p{Greek}>, are mostly Perl-defined shortcuts for
their equivalent compound forms.  The table shows these equivalences.  (In our
example, C<\\p{Greek}> is a just a shortcut for
C<\\p{Script_Extensions=Greek}>).  There are also a few Perl-defined single
forms that are not shortcuts for a compound form.  One such is C<\\p{Word}>.
These are also listed in the table.

In parsing these constructs, Perl always ignores Upper/lower case differences
everywhere within the {braces}.  Thus C<\\p{Greek}> means the same thing as
C<\\p{greek}>.  But note that changing the case of the C<"p"> or C<"P"> before
the left brace completely changes the meaning of the construct, from "match"
(for C<\\p{}>) to "doesn't match" (for C<\\P{}>).  Casing in this document is
for improved legibility.

Also, white space, hyphens, and underscores are normally ignored
everywhere between the {braces}, and hence can be freely added or removed
even if the C</x> modifier hasn't been specified on the regular expression.
But in the table below $a_bold_stricter at the beginning of an entry
means that tighter (stricter) rules are used for that entry:

=over 4

=over 4

=item Single form (C<\\p{name}>) tighter rules:

White space, hyphens, and underscores ARE significant
except for:

=over 4

=item * white space adjacent to a non-word character

=item * underscores separating digits in numbers

=back

That means, for example, that you can freely add or remove white space
adjacent to (but within) the braces without affecting the meaning.

=item Compound form (C<\\p{name=value}> or C<\\p{name:value}>) tighter rules:

The tighter rules given above for the single form apply to everything to the
right of the colon or equals; the looser rules still apply to everything to
the left.

That means, for example, that you can freely add or remove white space
adjacent to (but within) the braces and the colon or equal sign.

=back

=back

Some properties are considered obsolete by Unicode, but still available.
There are several varieties of obsolescence:

=over 4

=over 4

=item Stabilized

A property may be stabilized.  Such a determination does not indicate
that the property should or should not be used; instead it is a declaration
that the property will not be maintained nor extended for newly encoded
characters.  Such properties are marked with $a_bold_stabilized in the
table.

=item Deprecated

A property may be deprecated, perhaps because its original intent
has been replaced by another property, or because its specification was
somehow defective.  This means that its use is strongly
discouraged, so much so that a warning will be issued if used, unless the
regular expression is in the scope of a C<S<no warnings 'deprecated'>>
statement.  $A_bold_deprecated flags each such entry in the table, and
the entry there for the longest, most descriptive version of the property will
give the reason it is deprecated, and perhaps advice.  Perl may issue such a
warning, even for properties that aren't officially deprecated by Unicode,
when there used to be characters or code points that were matched by them, but
no longer.  This is to warn you that your program may not work like it did on
earlier Unicode releases.

A deprecated property may be made unavailable in a future Perl version, so it
is best to move away from them.

A deprecated property may also be stabilized, but this fact is not shown.

=item Obsolete

Properties marked with $a_bold_obsolete in the table are considered (plain)
obsolete.  Generally this designation is given to properties that Unicode once
used for internal purposes (but not any longer).

=item Discouraged

This is not actually a Unicode-specified obsolescence, but applies to certain
Perl extensions that are present for backwards compatibility, but are
discouraged from being used.  These are not obsolete, but their meanings are
not stable.  Future Unicode versions could force any of these extensions to be
removed without warning, replaced by another property with the same name that
means something different.  $A_bold_discouraged flags each such entry in the
table.  Use the equivalent shown instead.

@block_warning

=back

=back

The table below has two columns.  The left column contains the C<\\p{}>
constructs to look up, possibly preceded by the flags mentioned above; and
the right column contains information about them, like a description, or
synonyms.  The table shows both the single and compound forms for each
property that has them.  If the left column is a short name for a property,
the right column will give its longer, more descriptive name; and if the left
column is the longest name, the right column will show any equivalent shortest
name, in both single and compound forms if applicable.

If braces are not needed to specify a property (e.g., C<\\pL>), the left
column contains both forms, with and without braces.

The right column will also caution you if a property means something different
than what might normally be expected.

All single forms are Perl extensions; a few compound forms are as well, and
are noted as such.

Numbers in (parentheses) indicate the total number of Unicode code points
matched by the property.  For the entries that give the longest, most
descriptive version of the property, the count is followed by a list of some
of the code points matched by it.  The list includes all the matched
characters in the 0-255 range, enclosed in the familiar [brackets] the same as
a regular expression bracketed character class.  Following that, the next few
higher matching ranges are also given.  To avoid visual ambiguity, the SPACE
character is represented as C<\\x$space_hex>.

For emphasis, those properties that match no code points at all are listed as
well in a separate section following the table.

Most properties match the same code points regardless of whether C<"/i">
case-insensitive matching is specified or not.  But a few properties are
affected.  These are shown with the notation S<C<(/i= I<other_property>)>>
in the second column.  Under case-insensitive matching they match the
same code pode points as the property I<other_property>.

There is no description given for most non-Perl defined properties (See
L<$unicode_reference_url> for that).

For compactness, 'B<*>' is used as a wildcard instead of showing all possible
combinations.  For example, entries like:

 \\p{Gc: *}                                  \\p{General_Category: *}

mean that 'Gc' is a synonym for 'General_Category', and anything that is valid
for the latter is also valid for the former.  Similarly,

 \\p{Is_*}                                   \\p{*}

means that if and only if, for example, C<\\p{Foo}> exists, then
C<\\p{Is_Foo}> and C<\\p{IsFoo}> are also valid and all mean the same thing.
And similarly, C<\\p{Foo=Bar}> means the same as C<\\p{Is_Foo=Bar}> and
C<\\p{IsFoo=Bar}>.  "*" here is restricted to something not beginning with an
underscore.

Also, in binary properties, 'Yes', 'T', and 'True' are all synonyms for 'Y'.
And 'No', 'F', and 'False' are all synonyms for 'N'.  The table shows 'Y*' and
'N*' to indicate this, and doesn't have separate entries for the other
possibilities.  Note that not all properties which have values 'Yes' and 'No'
are binary, and they have all their values spelled out without using this wild
card, and a C<NOT> clause in their description that highlights their not being
binary.  These also require the compound form to match them, whereas true
binary properties have both single and compound forms available.

Note that all non-essential underscores are removed in the display of the
short names below.

B<Legend summary:>

=over 4

=item Z<>B<*> is a wild-card

=item B<(\\d+)> in the info column gives the number of Unicode code points matched
by this property.

=item B<$DEPRECATED> means this is deprecated.

=item B<$OBSOLETE> means this is obsolete.

=item B<$STABILIZED> means this is stabilized.

=item B<$STRICTER> means tighter (stricter) name matching applies.

=item B<$DISCOURAGED> means use of this form is discouraged, and may not be
stable.

=back

$formatted_properties

$zero_matches

=head1 Properties accessible through Unicode::UCD

The value of any Unicode (not including Perl extensions) character
property mentioned above for any single code point is available through
L<Unicode::UCD/charprop()>.  L<Unicode::UCD/charprops_all()> returns the
values of all the Unicode properties for a given code point.

Besides these, all the Unicode character properties mentioned above
(except for those marked as for internal use by Perl) are also
accessible by L<Unicode::UCD/prop_invlist()>.

Due to their nature, not all Unicode character properties are suitable for
regular expression matches, nor C<prop_invlist()>.  The remaining
non-provisional, non-internal ones are accessible via
L<Unicode::UCD/prop_invmap()> (except for those that this Perl installation
hasn't included; see L<below for which those are|/Unicode character properties
that are NOT accepted by Perl>).

For compatibility with other parts of Perl, all the single forms given in the
table in the L<section above|/Properties accessible through \\p{} and \\P{}>
are recognized.  BUT, there are some ambiguities between some Perl extensions
and the Unicode properties, all of which are silently resolved in favor of the
official Unicode property.  To avoid surprises, you should only use
C<prop_invmap()> for forms listed in the table below, which omits the
non-recommended ones.  The affected forms are the Perl single form equivalents
of Unicode properties, such as C<\\p{sc}> being a single-form equivalent of
C<\\p{gc=sc}>, which is treated by C<prop_invmap()> as the C<Script> property,
whose short name is C<sc>.  The table indicates the current ambiguities in the
INFO column, beginning with the word C<"NOT">.

The standard Unicode properties listed below are documented in
L<$unicode_reference_url>; Perl_Decimal_Digit is documented in
L<Unicode::UCD/prop_invmap()>.  The other Perl extensions are in
L<perlunicode/Other Properties>;

The first column in the table is a name for the property; the second column is
an alternative name, if any, plus possibly some annotations.  The alternative
name is the property's full name, unless that would simply repeat the first
column, in which case the second column indicates the property's short name
(if different).  The annotations are given only in the entry for the full
name.  The annotations for binary properties include a list of the first few
ranges that the property matches.  To avoid any ambiguity, the SPACE character
is represented as C<\\x$space_hex>.

If a property is obsolete, etc, the entry will be flagged with the same
characters used in the table in the L<section above|/Properties accessible
through \\p{} and \\P{}>, like B<$DEPRECATED> or B<$STABILIZED>.

$ucd_pod

=head1 Properties accessible through other means

Certain properties are accessible also via core function calls.  These are:

 Lowercase_Mapping          lc() and lcfirst()
 Titlecase_Mapping          ucfirst()
 Uppercase_Mapping          uc()

Also, Case_Folding is accessible through the C</i> modifier in regular
expressions, the C<\\F> transliteration escape, and the C<L<fc|perlfunc/fc>>
operator.

Besides being able to say C<\\p{Name=...}>, the Name and Name_Aliases
properties are accessible through the C<\\N{}> interpolation in double-quoted
strings and regular expressions; and functions C<charnames::viacode()>,
C<charnames::vianame()>, and C<charnames::string_vianame()> (which require a
C<use charnames ();> to be specified.

Finally, most properties related to decomposition are accessible via
L<Unicode::Normalize>.

=head1 Unicode character properties that are NOT accepted by Perl

Perl will generate an error for a few character properties in Unicode when
used in a regular expression.  The non-Unihan ones are listed below, with the
reasons they are not accepted, perhaps with work-arounds.  The short names for
the properties are listed enclosed in (parentheses).
As described after the list, an installation can change the defaults and choose
to accept any of these.  The list is machine generated based on the
choices made for the installation that generated this document.

@bad_re_properties

An installation can choose to allow any of these to be matched by downloading
the Unicode database from L<http://www.unicode.org/Public/> to
C<\$Config{privlib}>/F<unicore/> in the Perl source tree, changing the
controlling lists contained in the program
C<\$Config{privlib}>/F<unicore/mktables> and then re-compiling and installing.
(C<\%Config> is available from the Config module).

Also, perl can be recompiled to operate on an earlier version of the Unicode
standard.  Further information is at
C<\$Config{privlib}>/F<unicore/README.perl>.

=head1 Other information in the Unicode data base

The Unicode data base is delivered in two different formats.  The XML version
is valid for more modern Unicode releases.  The other version is a collection
of files.  The two are intended to give equivalent information.  Perl uses the
older form; this allows you to recompile Perl to use early Unicode releases.

The only non-character property that Perl currently supports is Named
Sequences, in which a sequence of code points
is given a name and generally treated as a single entity.  (Perl supports
these via the C<\\N{...}> double-quotish construct,
L<charnames/charnames::string_vianame(name)>, and L<Unicode::UCD/namedseq()>.

Below is a list of the files in the Unicode data base that Perl doesn't
currently use, along with very brief descriptions of their purposes.
Some of the names of the files have been shortened from those that Unicode
uses, in order to allow them to be distinguishable from similarly named files
on file systems for which only the first 8 characters of a name are
significant.

=over 4

@unused_files

=back

=head1 SEE ALSO

L<$unicode_reference_url>

L<perlrecharclass>

L<perlunicode>

END

    # And write it.  The 0 means no utf8.
    main::write([ $pod_directory, "$pod_file.pod" ], 0, \@OUT);
    return;
}

sub make_Name_pm () {
    # Create and write Name.pm, which contains subroutines and data to use in
    # conjunction with Name.pl

    # Maybe there's nothing to do.
    return unless $has_hangul_syllables || @code_points_ending_in_code_point;

    my @name = <<END;
$HEADER
$INTERNAL_ONLY_HEADER

END

    # Convert these structures to output format.
    my $code_points_ending_in_code_point =
        main::simple_dumper(\@code_points_ending_in_code_point,
                            ' ' x 8);
    my $names = main::simple_dumper(\%names_ending_in_code_point,
                                    ' ' x 8);
    my $loose_names = main::simple_dumper(\%loose_names_ending_in_code_point,
                                    ' ' x 8);

    # Do the same with the Hangul names,
    my $jamo;
    my $jamo_l;
    my $jamo_v;
    my $jamo_t;
    my $jamo_re;
    if ($has_hangul_syllables) {

        # Construct a regular expression of all the possible
        # combinations of the Hangul syllables.
        my @L_re;   # Leading consonants
        for my $i ($LBase .. $LBase + $LCount - 1) {
            push @L_re, $Jamo{$i}
        }
        my @V_re;   # Middle vowels
        for my $i ($VBase .. $VBase + $VCount - 1) {
            push @V_re, $Jamo{$i}
        }
        my @T_re;   # Trailing consonants
        for my $i ($TBase + 1 .. $TBase + $TCount - 1) {
            push @T_re, $Jamo{$i}
        }

        # The whole re is made up of the L V T combination.
        $jamo_re = '('
                    . join ('|', sort @L_re)
                    . ')('
                    . join ('|', sort @V_re)
                    . ')('
                    . join ('|', sort @T_re)
                    . ')?';

        # These hashes needed by the algorithm were generated
        # during reading of the Jamo.txt file
        $jamo = main::simple_dumper(\%Jamo, ' ' x 8);
        $jamo_l = main::simple_dumper(\%Jamo_L, ' ' x 8);
        $jamo_v = main::simple_dumper(\%Jamo_V, ' ' x 8);
        $jamo_t = main::simple_dumper(\%Jamo_T, ' ' x 8);
    }

    push @name, <<END;

package charnames;

# This module contains machine-generated tables and code for the
# algorithmically-determinable Unicode character names.  The following
# routines can be used to translate between name and code point and vice versa

{ # Closure

    # Matches legal code point.  4-6 hex numbers, If there are 6, the first
    # two must be 10; if there are 5, the first must not be a 0.  Written this
    # way to decrease backtracking.  The first regex allows the code point to
    # be at the end of a word, but to work properly, the word shouldn't end
    # with a valid hex character.  The second one won't match a code point at
    # the end of a word, and doesn't have the run-on issue
    my \$run_on_code_point_re = qr/$run_on_code_point_re/;
    my \$code_point_re = qr/$code_point_re/;

    # In the following hash, the keys are the bases of names which include
    # the code point in the name, like CJK UNIFIED IDEOGRAPH-4E01.  The value
    # of each key is another hash which is used to get the low and high ends
    # for each range of code points that apply to the name.
    my %names_ending_in_code_point = (
$names
    );

    # The following hash is a copy of the previous one, except is for loose
    # matching, so each name has blanks and dashes squeezed out
    my %loose_names_ending_in_code_point = (
$loose_names
    );

    # And the following array gives the inverse mapping from code points to
    # names.  Lowest code points are first
    \@code_points_ending_in_code_point = (
$code_points_ending_in_code_point
    );

    # Is exportable, make read-only
    Internals::SvREADONLY(\@code_points_ending_in_code_point, 1);
END
    # Earlier releases didn't have Jamos.  No sense outputting
    # them unless will be used.
    if ($has_hangul_syllables) {
        push @name, <<END;

    # Convert from code point to Jamo short name for use in composing Hangul
    # syllable names
    my %Jamo = (
$jamo
    );

    # Leading consonant (can be null)
    my %Jamo_L = (
$jamo_l
    );

    # Vowel
    my %Jamo_V = (
$jamo_v
    );

    # Optional trailing consonant
    my %Jamo_T = (
$jamo_t
    );

    # Computed re that splits up a Hangul name into LVT or LV syllables
    my \$syllable_re = qr/$jamo_re/;

    my \$HANGUL_SYLLABLE = "HANGUL SYLLABLE ";
    my \$loose_HANGUL_SYLLABLE = "HANGULSYLLABLE";

    # These constants names and values were taken from the Unicode standard,
    # version 5.1, section 3.12.  They are used in conjunction with Hangul
    # syllables
    my \$SBase = $SBase_string;
    my \$LBase = $LBase_string;
    my \$VBase = $VBase_string;
    my \$TBase = $TBase_string;
    my \$SCount = $SCount;
    my \$LCount = $LCount;
    my \$VCount = $VCount;
    my \$TCount = $TCount;
    my \$NCount = \$VCount * \$TCount;
END
    } # End of has Jamos

    push @name, << 'END';

    sub name_to_code_point_special {
        my ($name, $loose) = @_;

        # Returns undef if not one of the specially handled names; otherwise
        # returns the code point equivalent to the input name
        # $loose is non-zero if to use loose matching, 'name' in that case
        # must be input as upper case with all blanks and dashes squeezed out.
END
    if ($has_hangul_syllables) {
        push @name, << 'END';

        if ((! $loose && $name =~ s/$HANGUL_SYLLABLE//)
            || ($loose && $name =~ s/$loose_HANGUL_SYLLABLE//))
        {
            return if $name !~ qr/^$syllable_re$/;
            my $L = $Jamo_L{$1};
            my $V = $Jamo_V{$2};
            my $T = (defined $3) ? $Jamo_T{$3} : 0;
            return ($L * $VCount + $V) * $TCount + $T + $SBase;
        }
END
    }
    push @name, << 'END';

        # Name must end in 'code_point' for this to handle.
        return if (($loose && $name !~ /^ (.*?) ($run_on_code_point_re) $/x)
                   || (! $loose && $name !~ /^ (.*) ($code_point_re) $/x));

        my $base = $1;
        my $code_point = CORE::hex $2;
        my $names_ref;

        if ($loose) {
            $names_ref = \%loose_names_ending_in_code_point;
        }
        else {
            return if $base !~ s/-$//;
            $names_ref = \%names_ending_in_code_point;
        }

        # Name must be one of the ones which has the code point in it.
        return if ! $names_ref->{$base};

        # Look through the list of ranges that apply to this name to see if
        # the code point is in one of them.
        for (my $i = 0; $i < scalar @{$names_ref->{$base}{'low'}}; $i++) {
            return if $names_ref->{$base}{'low'}->[$i] > $code_point;
            next if $names_ref->{$base}{'high'}->[$i] < $code_point;

            # Here, the code point is in the range.
            return $code_point;
        }

        # Here, looked like the name had a code point number in it, but
        # did not match one of the valid ones.
        return;
    }

    sub code_point_to_name_special {
        my $code_point = shift;

        # Returns the name of a code point if algorithmically determinable;
        # undef if not
END
    if ($has_hangul_syllables) {
        push @name, << 'END';

        # If in the Hangul range, calculate the name based on Unicode's
        # algorithm
        if ($code_point >= $SBase && $code_point <= $SBase + $SCount -1) {
            use integer;
            my $SIndex = $code_point - $SBase;
            my $L = $LBase + $SIndex / $NCount;
            my $V = $VBase + ($SIndex % $NCount) / $TCount;
            my $T = $TBase + $SIndex % $TCount;
            $name = "$HANGUL_SYLLABLE$Jamo{$L}$Jamo{$V}";
            $name .= $Jamo{$T} if $T != $TBase;
            return $name;
        }
END
    }
    push @name, << 'END';

        # Look through list of these code points for one in range.
        foreach my $hash (@code_points_ending_in_code_point) {
            return if $code_point < $hash->{'low'};
            if ($code_point <= $hash->{'high'}) {
                return sprintf("%s-%04X", $hash->{'name'}, $code_point);
            }
        }
        return;            # None found
    }
} # End closure

1;
END

    main::write("Name.pm", 0, \@name);  # The 0 means no utf8.
    return;
}

sub make_UCD () {
    # Create and write UCD.pl, which passes info about the tables to
    # Unicode::UCD

    # Stringify structures for output
    my $loose_property_name_of
                           = simple_dumper(\%loose_property_name_of, ' ' x 4);
    chomp $loose_property_name_of;

    my $strict_property_name_of
                           = simple_dumper(\%strict_property_name_of, ' ' x 4);
    chomp $strict_property_name_of;

    my $stricter_to_file_of = simple_dumper(\%stricter_to_file_of, ' ' x 4);
    chomp $stricter_to_file_of;

    my $inline_definitions = simple_dumper(\@inline_definitions, " " x 4);
    chomp $inline_definitions;

    my $loose_to_file_of = simple_dumper(\%loose_to_file_of, ' ' x 4);
    chomp $loose_to_file_of;

    my $nv_floating_to_rational
                           = simple_dumper(\%nv_floating_to_rational, ' ' x 4);
    chomp $nv_floating_to_rational;

    my $why_deprecated = simple_dumper(\%Unicode::UCD::why_deprecated, ' ' x 4);
    chomp $why_deprecated;

    # We set the key to the file when we associated files with tables, but we
    # couldn't do the same for the value then, as we might not have the file
    # for the alternate table figured out at that time.
    foreach my $cased (keys %caseless_equivalent_to) {
        my @path = $caseless_equivalent_to{$cased}->file_path;
        my $path;
        if ($path[0] eq "#") {  # Pseudo-directory '#'
            $path = join '/', @path;
        }
        else {  # Gets rid of lib/
            $path = join '/', @path[1, -1];
        }
        $caseless_equivalent_to{$cased} = $path;
    }
    my $caseless_equivalent_to
                           = simple_dumper(\%caseless_equivalent_to, ' ' x 4);
    chomp $caseless_equivalent_to;

    my $loose_property_to_file_of
                        = simple_dumper(\%loose_property_to_file_of, ' ' x 4);
    chomp $loose_property_to_file_of;

    my $strict_property_to_file_of
                        = simple_dumper(\%strict_property_to_file_of, ' ' x 4);
    chomp $strict_property_to_file_of;

    my $file_to_swash_name = simple_dumper(\%file_to_swash_name, ' ' x 4);
    chomp $file_to_swash_name;

    # Create a mapping from each alias of Perl single-form extensions to all
    # its equivalent aliases, for quick look-up.
    my %perlprop_to_aliases;
    foreach my $table ($perl->tables) {

        # First create the list of the aliases of each extension
        my @aliases_list;    # List of legal aliases for this extension

        my $table_name = $table->name;
        my $standard_table_name = standardize($table_name);
        my $table_full_name = $table->full_name;
        my $standard_table_full_name = standardize($table_full_name);

        # Make sure that the list has both the short and full names
        push @aliases_list, $table_name, $table_full_name;

        my $found_ucd = 0;  # ? Did we actually get an alias that should be
                            # output for this table

        # Go through all the aliases (including the two just added), and add
        # any new unique ones to the list
        foreach my $alias ($table->aliases) {

            # Skip non-legal names
            next unless $alias->ok_as_filename;
            next unless $alias->ucd;

            $found_ucd = 1;     # have at least one legal name

            my $name = $alias->name;
            my $standard = standardize($name);

            # Don't repeat a name that is equivalent to one already on the
            # list
            next if $standard eq $standard_table_name;
            next if $standard eq $standard_table_full_name;

            push @aliases_list, $name;
        }

        # If there were no legal names, don't output anything.
        next unless $found_ucd;

        # To conserve memory in the program reading these in, omit full names
        # that are identical to the short name, when those are the only two
        # aliases for the property.
        if (@aliases_list == 2 && $aliases_list[0] eq $aliases_list[1]) {
            pop @aliases_list;
        }

        # Here, @aliases_list is the list of all the aliases that this
        # extension legally has.  Now can create a map to it from each legal
        # standardized alias
        foreach my $alias ($table->aliases) {
            next unless $alias->ucd;
            next unless $alias->ok_as_filename;
            push @{$perlprop_to_aliases{standardize($alias->name)}},
                 uniques @aliases_list;
        }
    }

    # Make a list of all combinations of properties/values that are suppressed.
    my @suppressed;
    if (! $debug_skip) {    # This tends to fail in this debug mode
        foreach my $property_name (keys %why_suppressed) {

            # Just the value
            my $value_name = $1 if $property_name =~ s/ = ( .* ) //x;

            # The hash may contain properties not in this release of Unicode
            next unless defined (my $property = property_ref($property_name));

            # Find all combinations
            foreach my $prop_alias ($property->aliases) {
                my $prop_alias_name = standardize($prop_alias->name);

                # If no =value, there's just one combination possible for this
                if (! $value_name) {

                    # The property may be suppressed, but there may be a proxy
                    # for it, so it shouldn't be listed as suppressed
                    next if $prop_alias->ucd;
                    push @suppressed, $prop_alias_name;
                }
                else {  # Otherwise
                    foreach my $value_alias
                                    ($property->table($value_name)->aliases)
                    {
                        next if $value_alias->ucd;

                        push @suppressed, "$prop_alias_name="
                                        .  standardize($value_alias->name);
                    }
                }
            }
        }
    }
    @suppressed = sort @suppressed; # So doesn't change between runs of this
                                    # program

    # Convert the structure below (designed for Name.pm) to a form that UCD
    # wants, so it doesn't have to modify it at all; i.e. so that it includes
    # an element for the Hangul syllables in the appropriate place, and
    # otherwise changes the name to include the "-<code point>" suffix.
    my @algorithm_names;
    my $done_hangul = $v_version lt v2.0.0;  # Hanguls as we know them came
                                             # along in this version
    # Copy it linearly.
    for my $i (0 .. @code_points_ending_in_code_point - 1) {

        # Insert the hanguls in the correct place.
        if (! $done_hangul
            && $code_points_ending_in_code_point[$i]->{'low'} > $SBase)
        {
            $done_hangul = 1;
            push @algorithm_names, { low => $SBase,
                                     high => $SBase + $SCount - 1,
                                     name => '<hangul syllable>',
                                    };
        }

        # Copy the current entry, modified.
        push @algorithm_names, {
            low => $code_points_ending_in_code_point[$i]->{'low'},
            high => $code_points_ending_in_code_point[$i]->{'high'},
            name =>
               "$code_points_ending_in_code_point[$i]->{'name'}-<code point>",
        };
    }

    # Serialize these structures for output.
    my $loose_to_standard_value
                          = simple_dumper(\%loose_to_standard_value, ' ' x 4);
    chomp $loose_to_standard_value;

    my $string_property_loose_to_name
                    = simple_dumper(\%string_property_loose_to_name, ' ' x 4);
    chomp $string_property_loose_to_name;

    my $perlprop_to_aliases = simple_dumper(\%perlprop_to_aliases, ' ' x 4);
    chomp $perlprop_to_aliases;

    my $prop_aliases = simple_dumper(\%prop_aliases, ' ' x 4);
    chomp $prop_aliases;

    my $prop_value_aliases = simple_dumper(\%prop_value_aliases, ' ' x 4);
    chomp $prop_value_aliases;

    my $suppressed = (@suppressed) ? simple_dumper(\@suppressed, ' ' x 4) : "";
    chomp $suppressed;

    my $algorithm_names = simple_dumper(\@algorithm_names, ' ' x 4);
    chomp $algorithm_names;

    my $ambiguous_names = simple_dumper(\%ambiguous_names, ' ' x 4);
    chomp $ambiguous_names;

    my $combination_property = simple_dumper(\%combination_property, ' ' x 4);
    chomp $combination_property;

    my $loose_defaults = simple_dumper(\%loose_defaults, ' ' x 4);
    chomp $loose_defaults;

    my @ucd = <<END;
$HEADER
$INTERNAL_ONLY_HEADER

# This file is for the use of Unicode::UCD

# Highest legal Unicode code point
\$Unicode::UCD::MAX_UNICODE_CODEPOINT = 0x$MAX_UNICODE_CODEPOINT_STRING;

# Hangul syllables
\$Unicode::UCD::HANGUL_BEGIN = $SBase_string;
\$Unicode::UCD::HANGUL_COUNT = $SCount;

# Maps Unicode (not Perl single-form extensions) property names in loose
# standard form to their corresponding standard names
\%Unicode::UCD::loose_property_name_of = (
$loose_property_name_of
);

# Same, but strict names
\%Unicode::UCD::strict_property_name_of = (
$strict_property_name_of
);

# Gives the definitions (in the form of inversion lists) for those properties
# whose definitions aren't kept in files
\@Unicode::UCD::inline_definitions = (
$inline_definitions
);

# Maps property, table to file for those using stricter matching.  For paths
# whose directory is '#', the file is in the form of a numeric index into
# \@inline_definitions
\%Unicode::UCD::stricter_to_file_of = (
$stricter_to_file_of
);

# Maps property, table to file for those using loose matching.  For paths
# whose directory is '#', the file is in the form of a numeric index into
# \@inline_definitions
\%Unicode::UCD::loose_to_file_of = (
$loose_to_file_of
);

# Maps floating point to fractional form
\%Unicode::UCD::nv_floating_to_rational = (
$nv_floating_to_rational
);

# If a %e floating point number doesn't have this number of digits in it after
# the decimal point to get this close to a fraction, it isn't considered to be
# that fraction even if all the digits it does have match.
\$Unicode::UCD::e_precision = $E_FLOAT_PRECISION;

# Deprecated tables to generate a warning for.  The key is the file containing
# the table, so as to avoid duplication, as many property names can map to the
# file, but we only need one entry for all of them.
\%Unicode::UCD::why_deprecated = (
$why_deprecated
);

# A few properties have different behavior under /i matching.  This maps
# those to substitute files to use under /i.
\%Unicode::UCD::caseless_equivalent = (
$caseless_equivalent_to
);

# Property names to mapping files
\%Unicode::UCD::loose_property_to_file_of = (
$loose_property_to_file_of
);

# Property names to mapping files
\%Unicode::UCD::strict_property_to_file_of = (
$strict_property_to_file_of
);

# Files to the swash names within them.
\%Unicode::UCD::file_to_swash_name = (
$file_to_swash_name
);

# Keys are all the possible "prop=value" combinations, in loose form; values
# are the standard loose name for the 'value' part of the key
\%Unicode::UCD::loose_to_standard_value = (
$loose_to_standard_value
);

# String property loose names to standard loose name
\%Unicode::UCD::string_property_loose_to_name = (
$string_property_loose_to_name
);

# Keys are Perl extensions in loose form; values are each one's list of
# aliases
\%Unicode::UCD::loose_perlprop_to_name = (
$perlprop_to_aliases
);

# Keys are standard property name; values are each one's aliases
\%Unicode::UCD::prop_aliases = (
$prop_aliases
);

# Keys of top level are standard property name; values are keys to another
# hash,  Each one is one of the property's values, in standard form.  The
# values are that prop-val's aliases.  If only one specified, the short and
# long alias are identical.
\%Unicode::UCD::prop_value_aliases = (
$prop_value_aliases
);

# Ordered (by code point ordinal) list of the ranges of code points whose
# names are algorithmically determined.  Each range entry is an anonymous hash
# of the start and end points and a template for the names within it.
\@Unicode::UCD::algorithmic_named_code_points = (
$algorithm_names
);

# The properties that as-is have two meanings, and which must be disambiguated
\%Unicode::UCD::ambiguous_names = (
$ambiguous_names
);

# Keys are the prop-val combinations which are the default values for the
# given property, expressed in standard loose form
\%Unicode::UCD::loose_defaults = (
$loose_defaults
);

# The properties that are combinations, in that they have both a map table and
# a match table.  This is actually for UCD.t, so it knows how to test for
# these.
\%Unicode::UCD::combination_property = (
$combination_property
);

# All combinations of names that are suppressed.
# This is actually for UCD.t, so it knows which properties shouldn't have
# entries.  If it got any bigger, would probably want to put it in its own
# file to use memory only when it was needed, in testing.
\@Unicode::UCD::suppressed_properties = (
$suppressed
);

1;
END

    main::write("UCD.pl", 0, \@ucd);  # The 0 means no utf8.
    return;
}

sub write_all_tables() {
    # Write out all the tables generated by this program to files, as well as
    # the supporting data structures, pod file, and .t file.

    my @writables;              # List of tables that actually get written
    my %match_tables_to_write;  # Used to collapse identical match tables
                                # into one file.  Each key is a hash function
                                # result to partition tables into buckets.
                                # Each value is an array of the tables that
                                # fit in the bucket.

    # For each property ...
    # (sort so that if there is an immutable file name, it has precedence, so
    # some other property can't come in and take over its file name.  (We
    # don't care if both defined, as they had better be different anyway.)
    # The property named 'Perl' needs to be first (it doesn't have any
    # immutable file name) because empty properties are defined in terms of
    # its table named 'All' under the -annotate option.)   We also sort by
    # the property's name.  This is just for repeatability of the outputs
    # between runs of this program, but does not affect correctness.
    PROPERTY:
    foreach my $property ($perl,
                          sort { return -1 if defined $a->file;
                                 return 1 if defined $b->file;
                                 return $a->name cmp $b->name;
                                } grep { $_ != $perl } property_ref('*'))
    {
        my $type = $property->type;

        # And for each table for that property, starting with the mapping
        # table for it ...
        TABLE:
        foreach my $table($property,

                        # and all the match tables for it (if any), sorted so
                        # the ones with the shortest associated file name come
                        # first.  The length sorting prevents problems of a
                        # longer file taking a name that might have to be used
                        # by a shorter one.  The alphabetic sorting prevents
                        # differences between releases
                        sort {  my $ext_a = $a->external_name;
                                return 1 if ! defined $ext_a;
                                my $ext_b = $b->external_name;
                                return -1 if ! defined $ext_b;

                                # But return the non-complement table before
                                # the complement one, as the latter is defined
                                # in terms of the former, and needs to have
                                # the information for the former available.
                                return 1 if $a->complement != 0;
                                return -1 if $b->complement != 0;

                                # Similarly, return a subservient table after
                                # a leader
                                return 1 if $a->leader != $a;
                                return -1 if $b->leader != $b;

                                my $cmp = length $ext_a <=> length $ext_b;

                                # Return result if lengths not equal
                                return $cmp if $cmp;

                                # Alphabetic if lengths equal
                                return $ext_a cmp $ext_b
                        } $property->tables
                    )
        {

            # Here we have a table associated with a property.  It could be
            # the map table (done first for each property), or one of the
            # other tables.  Determine which type.
            my $is_property = $table->isa('Property');

            my $name = $table->name;
            my $complete_name = $table->complete_name;

            # See if should suppress the table if is empty, but warn if it
            # contains something.
            my $suppress_if_empty_warn_if_not
                    = $why_suppress_if_empty_warn_if_not{$complete_name} || 0;

            # Calculate if this table should have any code points associated
            # with it or not.
            my $expected_empty =

                # $perl should be empty
                ($is_property && ($table == $perl))

                # Match tables in properties we skipped populating should be
                # empty
                || (! $is_property && ! $property->to_create_match_tables)

                # Tables and properties that are expected to have no code
                # points should be empty
                || $suppress_if_empty_warn_if_not
            ;

            # Set a boolean if this table is the complement of an empty binary
            # table
            my $is_complement_of_empty_binary =
                $type == $BINARY &&
                (($table == $property->table('Y')
                    && $property->table('N')->is_empty)
                || ($table == $property->table('N')
                    && $property->table('Y')->is_empty));

            if ($table->is_empty) {

                if ($suppress_if_empty_warn_if_not) {
                    $table->set_fate($SUPPRESSED,
                                     $suppress_if_empty_warn_if_not);
                }

                # Suppress (by skipping them) expected empty tables.
                next TABLE if $expected_empty;

                # And setup to later output a warning for those that aren't
                # known to be allowed to be empty.  Don't do the warning if
                # this table is a child of another one to avoid duplicating
                # the warning that should come from the parent one.
                if (($table == $property || $table->parent == $table)
                    && $table->fate != $SUPPRESSED
                    && $table->fate != $MAP_PROXIED
                    && ! grep { $complete_name =~ /^$_$/ }
                                                    @tables_that_may_be_empty)
                {
                    push @unhandled_properties, "$table";
                }

                # The old way of expressing an empty match list was to
                # complement the list that matches everything.  The new way is
                # to create an empty inversion list, but this doesn't work for
                # annotating, so use the old way then.
                $table->set_complement($All) if $annotate
                                                && $table != $property;
            }
            elsif ($expected_empty) {
                my $because = "";
                if ($suppress_if_empty_warn_if_not) {
                    $because = " because $suppress_if_empty_warn_if_not";
                }

                Carp::my_carp("Not expecting property $table$because.  Generating file for it anyway.");
            }

            # Some tables should match everything
            my $expected_full =
                ($table->fate == $SUPPRESSED)
                ? 0
                : ($is_property)
                  ? # All these types of map tables will be full because
                    # they will have been populated with defaults
                    ($type == $ENUM)

                  : # A match table should match everything if its method
                    # shows it should
                    ($table->matches_all

                    # The complement of an empty binary table will match
                    # everything
                    || $is_complement_of_empty_binary
                    )
            ;

            my $count = $table->count;
            if ($expected_full) {
                if ($count != $MAX_WORKING_CODEPOINTS) {
                    Carp::my_carp("$table matches only "
                    . clarify_number($count)
                    . " Unicode code points but should match "
                    . clarify_number($MAX_WORKING_CODEPOINTS)
                    . " (off by "
                    .  clarify_number(abs($MAX_WORKING_CODEPOINTS - $count))
                    . ").  Proceeding anyway.");
                }

                # Here is expected to be full.  If it is because it is the
                # complement of an (empty) binary table that is to be
                # suppressed, then suppress this one as well.
                if ($is_complement_of_empty_binary) {
                    my $opposing_name = ($name eq 'Y') ? 'N' : 'Y';
                    my $opposing = $property->table($opposing_name);
                    my $opposing_status = $opposing->status;
                    if ($opposing_status) {
                        $table->set_status($opposing_status,
                                           $opposing->status_info);
                    }
                }
            }
            elsif ($count == $MAX_UNICODE_CODEPOINTS
                   && $name ne "Any"
                   && ($table == $property || $table->leader == $table)
                   && $table->property->status ne $NORMAL)
            {
                    Carp::my_carp("$table unexpectedly matches all Unicode code points.  Proceeding anyway.");
            }

            if ($table->fate >= $SUPPRESSED) {
                if (! $is_property) {
                    my @children = $table->children;
                    foreach my $child (@children) {
                        if ($child->fate < $SUPPRESSED) {
                            Carp::my_carp_bug("'$table' is suppressed and has a child '$child' which isn't");
                        }
                    }
                }
                next TABLE;

            }

            if (! $is_property) {

                make_ucd_table_pod_entries($table) if $table->property == $perl;

                # Several things need to be done just once for each related
                # group of match tables.  Do them on the parent.
                if ($table->parent == $table) {

                    # Add an entry in the pod file for the table; it also does
                    # the children.
                    make_re_pod_entries($table) if defined $pod_directory;

                    # See if the table matches identical code points with
                    # something that has already been processed and is ready
                    # for output.  In that case, no need to have two files
                    # with the same code points in them.  We use the table's
                    # hash() method to store these in buckets, so that it is
                    # quite likely that if two tables are in the same bucket
                    # they will be identical, so don't have to compare tables
                    # frequently.  The tables have to have the same status to
                    # share a file, so add this to the bucket hash.  (The
                    # reason for this latter is that UCD.pm associates a
                    # status with a file.) We don't check tables that are
                    # inverses of others, as it would lead to some coding
                    # complications, and checking all the regular ones should
                    # find everything.
                    if ($table->complement == 0) {
                        my $hash = $table->hash . ';' . $table->status;

                        # Look at each table that is in the same bucket as
                        # this one would be.
                        foreach my $comparison
                                            (@{$match_tables_to_write{$hash}})
                        {
                            # If the table doesn't point back to this one, we
                            # see if it matches identically
                            if (   $comparison->leader != $table
                                && $table->matches_identically_to($comparison))
                            {
                                $table->set_equivalent_to($comparison,
                                                                Related => 0);
                                next TABLE;
                            }
                        }

                        # Here, not equivalent, add this table to the bucket.
                        push @{$match_tables_to_write{$hash}}, $table;
                    }
                }
            }
            else {

                # Here is the property itself.
                # Don't write out or make references to the $perl property
                next if $table == $perl;

                make_ucd_table_pod_entries($table);

                # There is a mapping stored of the various synonyms to the
                # standardized name of the property for Unicode::UCD.
                # Also, the pod file contains entries of the form:
                # \p{alias: *}         \p{full: *}
                # rather than show every possible combination of things.

                my @property_aliases = $property->aliases;

                my $full_property_name = $property->full_name;
                my $property_name = $property->name;
                my $standard_property_name = standardize($property_name);
                my $standard_property_full_name
                                        = standardize($full_property_name);

                # We also create for Unicode::UCD a list of aliases for
                # the property.  The list starts with the property name;
                # then its full name.
                my @property_list;
                my @standard_list;
                if ( $property->fate <= $MAP_PROXIED) {
                    @property_list = ($property_name, $full_property_name);
                    @standard_list = ($standard_property_name,
                                        $standard_property_full_name);
                }

                # For each synonym ...
                for my $i (0 .. @property_aliases - 1)  {
                    my $alias = $property_aliases[$i];
                    my $alias_name = $alias->name;
                    my $alias_standard = standardize($alias_name);


                    # Add other aliases to the list of property aliases
                    if ($property->fate <= $MAP_PROXIED
                        && ! grep { $alias_standard eq $_ } @standard_list)
                    {
                        push @property_list, $alias_name;
                        push @standard_list, $alias_standard;
                    }

                    # For Unicode::UCD, set the mapping of the alias to the
                    # property
                    if ($type == $STRING) {
                        if ($property->fate <= $MAP_PROXIED) {
                            $string_property_loose_to_name{$alias_standard}
                                            = $standard_property_name;
                        }
                    }
                    else {
                        my $hash_ref = ($alias_standard =~ /^_/)
                                       ? \%strict_property_name_of
                                       : \%loose_property_name_of;
                        if (exists $hash_ref->{$alias_standard}) {
                            Carp::my_carp("There already is a property with the same standard name as $alias_name: $hash_ref->{$alias_standard}.  Old name is retained");
                        }
                        else {
                            $hash_ref->{$alias_standard}
                                                = $standard_property_name;
                        }

                        # Now for the re pod entry for this alias.  Skip if not
                        # outputting a pod; skip the first one, which is the
                        # full name so won't have an entry like: '\p{full: *}
                        # \p{full: *}', and skip if don't want an entry for
                        # this one.
                        next if $i == 0
                                || ! defined $pod_directory
                                || ! $alias->make_re_pod_entry;

                        my $rhs = "\\p{$full_property_name: *}";
                        if ($property != $perl && $table->perl_extension) {
                            $rhs .= ' (Perl extension)';
                        }
                        push @match_properties,
                            format_pod_line($indent_info_column,
                                        '\p{' . $alias->name . ': *}',
                                        $rhs,
                                        $alias->status);
                    }
                }

                # The list of all possible names is attached to each alias, so
                # lookup is easy
                if (@property_list) {
                    push @{$prop_aliases{$standard_list[0]}}, @property_list;
                }

                if ($property->fate <= $MAP_PROXIED) {

                    # Similarly, we create for Unicode::UCD a list of
                    # property-value aliases.

                    # Look at each table in the property...
                    foreach my $table ($property->tables) {
                        my @values_list;
                        my $table_full_name = $table->full_name;
                        my $standard_table_full_name
                                              = standardize($table_full_name);
                        my $table_name = $table->name;
                        my $standard_table_name = standardize($table_name);

                        # The list starts with the table name and its full
                        # name.
                        push @values_list, $table_name, $table_full_name;

                        # We add to the table each unique alias that isn't
                        # discouraged from use.
                        foreach my $alias ($table->aliases) {
                            next if $alias->status
                                 && $alias->status eq $DISCOURAGED;
                            my $name = $alias->name;
                            my $standard = standardize($name);
                            next if $standard eq $standard_table_name;
                            next if $standard eq $standard_table_full_name;
                            push @values_list, $name;
                        }

                        # Here @values_list is a list of all the aliases for
                        # the table.  That is, all the property-values given
                        # by this table.  By agreement with Unicode::UCD,
                        # if the name and full name are identical, and there
                        # are no other names, drop the duplicate entry to save
                        # memory.
                        if (@values_list == 2
                            && $values_list[0] eq $values_list[1])
                        {
                            pop @values_list
                        }

                        # To save memory, unlike the similar list for property
                        # aliases above, only the standard forms have the list.
                        # This forces an extra step of converting from input
                        # name to standard name, but the savings are
                        # considerable.  (There is only marginal savings if we
                        # did this with the property aliases.)
                        push @{$prop_value_aliases{$standard_property_name}{$standard_table_name}}, @values_list;
                    }
                }

                # Don't write out a mapping file if not desired.
                next if ! $property->to_output_map;
            }

            # Here, we know we want to write out the table, but don't do it
            # yet because there may be other tables that come along and will
            # want to share the file, and the file's comments will change to
            # mention them.  So save for later.
            push @writables, $table;

        } # End of looping through the property and all its tables.
    } # End of looping through all properties.

    # Now have all the tables that will have files written for them.  Do it.
    foreach my $table (@writables) {
        my @directory;
        my $filename;
        my $property = $table->property;
        my $is_property = ($table == $property);

        # For very short tables, instead of writing them out to actual files,
        # we in-line their inversion list definitions into UCD.pm.  The
        # definition replaces the file name, and the special pseudo-directory
        # '#' is used to signal this.  This significantly cuts down the number
        # of files written at little extra cost to the hashes in UCD.pm.
        # And it means, no run-time files to read to get the definitions.
        if (! $is_property
            && ! $annotate  # For annotation, we want to explicitly show
                            # everything, so keep in files
            && $table->ranges <= 3)
        {
            my @ranges = $table->ranges;
            my $count = @ranges;
            if ($count == 0) {  # 0th index reserved for 0-length lists
                $filename = 0;
            }
            elsif ($table->leader != $table) {

                # Here, is a table that is equivalent to another; code
                # in register_file_for_name() causes its leader's definition
                # to be used

                next;
            }
            else {  # No equivalent table so far.

                # Build up its definition range-by-range.
                my $definition = "";
                while (defined (my $range = shift @ranges)) {
                    my $end = $range->end;
                    if ($end < $MAX_WORKING_CODEPOINT) {
                        $count++;
                        $end = "\n" . ($end + 1);
                    }
                    else {  # Extends to infinity, hence no 'end'
                        $end = "";
                    }
                    $definition .= "\n" . $range->start . $end;
                }
                $definition = "V$count" . $definition;
                $filename = @inline_definitions;
                push @inline_definitions, $definition;
            }
            @directory = "#";
            register_file_for_name($table, \@directory, $filename);
            next;
        }

        if (! $is_property) {
            # Match tables for the property go in lib/$subdirectory, which is
            # the property's name.  Don't use the standard file name for this,
            # as may get an unfamiliar alias
            @directory = ($matches_directory, ($property->match_subdir)
                                              ? $property->match_subdir
                                              : $property->external_name);
        }
        else {

            @directory = $table->directory;
            $filename = $table->file;
        }

        # Use specified filename if available, or default to property's
        # shortest name.  We need an 8.3 safe filename (which means "an 8
        # safe" filename, since after the dot is only 'pl', which is < 3)
        # The 2nd parameter is if the filename shouldn't be changed, and
        # it shouldn't iff there is a hard-coded name for this table.
        $filename = construct_filename(
                                $filename || $table->external_name,
                                ! $filename,    # mutable if no filename
                                \@directory);

        register_file_for_name($table, \@directory, $filename);

        # Only need to write one file when shared by more than one
        # property
        next if ! $is_property
                && ($table->leader != $table || $table->complement != 0);

        # Construct a nice comment to add to the file
        $table->set_final_comment;

        $table->write;
    }


    # Write out the pod file
    make_pod;

    # And Name.pm, UCD.pl
    make_Name_pm;
    make_UCD;

    make_property_test_script() if $make_test_script;
    make_normalization_test_script() if $make_norm_test_script;
    return;
}

my @white_space_separators = ( # This used only for making the test script.
                            "",
                            ' ',
                            "\t",
                            '   '
                        );

sub generate_separator($lhs) {
    # This used only for making the test script.  It generates the colon or
    # equal separator between the property and property value, with random
    # white space surrounding the separator

    return "" if $lhs eq "";  # No separator if there's only one (the r) side

    # Choose space before and after randomly
    my $spaces_before =$white_space_separators[rand(@white_space_separators)];
    my $spaces_after = $white_space_separators[rand(@white_space_separators)];

    # And return the whole complex, half the time using a colon, half the
    # equals
    return $spaces_before
            . (rand() < 0.5) ? '=' : ':'
            . $spaces_after;
}

sub generate_tests($lhs, $rhs, $valid_code, $invalid_code, $warning) {
    # This used only for making the test script.  It generates test cases that
    # are expected to compile successfully in perl.  Note that the LHS and
    # RHS are assumed to already be as randomized as the caller wants.

    # $lhs          # The property: what's to the left of the colon
                    #  or equals separator
    # $rhs          # The property value; what's to the right
    # $valid_code   # A code point that's known to be in the
                        # table given by LHS=RHS; undef if table is
                        # empty
    # $invalid_code # A code point known to not be in the table;
                    # undef if the table is all code points
    # $warning

    # Get the colon or equal
    my $separator = generate_separator($lhs);

    # The whole 'property=value'
    my $name = "$lhs$separator$rhs";

    my @output;
    # Create a complete set of tests, with complements.
    if (defined $valid_code) {
        push @output, <<"EOC"
Expect(1, $valid_code, '\\p{$name}', $warning);
Expect(0, $valid_code, '\\p{^$name}', $warning);
Expect(0, $valid_code, '\\P{$name}', $warning);
Expect(1, $valid_code, '\\P{^$name}', $warning);
EOC
    }
    if (defined $invalid_code) {
        push @output, <<"EOC"
Expect(0, $invalid_code, '\\p{$name}', $warning);
Expect(1, $invalid_code, '\\p{^$name}', $warning);
Expect(1, $invalid_code, '\\P{$name}', $warning);
Expect(0, $invalid_code, '\\P{^$name}', $warning);
EOC
    }
    return @output;
}

sub generate_wildcard_tests($lhs, $rhs, $valid_code, $invalid_code, $warning) {
    # This used only for making the test script.  It generates wildcardl
    # matching test cases that are expected to compile successfully in perl.

    # $lhs           # The property: what's to the left of the
                     # or equals separator
    # $rhs           # The property value; what's to the right
    # $valid_code    # A code point that's known to be in the
                     # table given by LHS=RHS; undef if table is
                     # empty
    # $invalid_code  # A code point known to not be in the table;
                     # undef if the table is all code points
    # $warning

    return if $lhs eq "";
    return if $lhs =~ / ^ Is_ /x;   # These are not currently supported

    # Generate a standardized pattern, with colon being the delimitter
    my $wildcard = "$lhs=:\\A$rhs\\z:";

    my @output;
    push @output, "Expect(1, $valid_code, '\\p{$wildcard}', $warning);"
                                                        if defined $valid_code;
    push @output, "Expect(0, $invalid_code, '\\p{$wildcard}', $warning);"
                                                      if defined $invalid_code;
    return @output;
}

sub generate_error($lhs, $rhs, $already_in_error=0) {
    # This used only for making the test script.  It generates test cases that
    # are expected to not only not match, but to be syntax or similar errors

    # $lhs                # The property: what's to the left of the
                          # colon or equals separator
    # $rhs                # The property value; what's to the right
    # $already_in_error   # Boolean; if true it's known that the
                          # unmodified LHS and RHS will cause an error.
                          # This routine should not force another one
    # Get the colon or equal
    my $separator = generate_separator($lhs);

    # Since this is an error only, don't bother to randomly decide whether to
    # put the error on the left or right side; and assume that the RHS is
    # loosely matched, again for convenience rather than rigor.
    $rhs = randomize_loose_name($rhs, 'ERROR') unless $already_in_error;

    my $property = $lhs . $separator . $rhs;

    return <<"EOC";
Error('\\p{$property}');
Error('\\P{$property}');
EOC
}

# These are used only for making the test script
# XXX Maybe should also have a bad strict seps, which includes underscore.

my @good_loose_seps = (
            " ",
            "-",
            "\t",
            "",
            "_",
           );
my @bad_loose_seps = (
           "/a/",
           ':=',
          );

sub randomize_stricter_name($name) {
    # This used only for making the test script.  Take the input name and
    # return a randomized, but valid version of it under the stricter matching
    # rules.

    # If the name looks like a number (integer, floating, or rational), do
    # some extra work
    if ($name =~ qr{ ^ ( -? ) (\d+ ( ( [./] ) \d+ )? ) $ }x) {
        my $sign = $1;
        my $number = $2;
        my $separator = $3;

        # If there isn't a sign, part of the time add a plus
        # Note: Not testing having any denominator having a minus sign
        if (! $sign) {
            $sign = '+' if rand() <= .3;
        }

        # And add 0 or more leading zeros.
        $name = $sign . ('0' x int rand(10)) . $number;

        if (defined $separator) {
            my $extra_zeros = '0' x int rand(10);

            if ($separator eq '.') {

                # Similarly, add 0 or more trailing zeros after a decimal
                # point
                $name .= $extra_zeros;
            }
            else {

                # Or, leading zeros before the denominator
                $name =~ s,/,/$extra_zeros,;
            }
        }
    }

    # For legibility of the test, only change the case of whole sections at a
    # time.  To do this, first split into sections.  The split returns the
    # delimiters
    my @sections;
    for my $section (split / ( [ - + \s _ . ]+ ) /x, $name) {
        trace $section if main::DEBUG && $to_trace;

        if (length $section > 1 && $section !~ /\D/) {

            # If the section is a sequence of digits, about half the time
            # randomly add underscores between some of them.
            if (rand() > .5) {

                # Figure out how many underscores to add.  max is 1 less than
                # the number of digits.  (But add 1 at the end to make sure
                # result isn't 0, and compensate earlier by subtracting 2
                # instead of 1)
                my $num_underscores = int rand(length($section) - 2) + 1;

                # And add them evenly throughout, for convenience, not rigor
                use integer;
                my $spacing = (length($section) - 1)/ $num_underscores;
                my $temp = $section;
                $section = "";
                for my $i (1 .. $num_underscores) {
                    $section .= substr($temp, 0, $spacing, "") . '_';
                }
                $section .= $temp;
            }
            push @sections, $section;
        }
        else {

            # Here not a sequence of digits.  Change the case of the section
            # randomly
            my $switch = int rand(4);
            if ($switch == 0) {
                push @sections, uc $section;
            }
            elsif ($switch == 1) {
                push @sections, lc $section;
            }
            elsif ($switch == 2) {
                push @sections, ucfirst $section;
            }
            else {
                push @sections, $section;
            }
        }
    }
    trace "returning", join "", @sections if main::DEBUG && $to_trace;
    return join "", @sections;
}

sub randomize_loose_name($name, $want_error=0) {
    # This used only for making the test script

    $name = randomize_stricter_name($name);

    my @parts;
    push @parts, $good_loose_seps[rand(@good_loose_seps)];

    # Preserve trailing ones for the sake of not stripping the underscore from
    # 'L_'
    for my $part (split /[-\s_]+ (?= . )/, $name) {
        if (@parts) {
            if ($want_error and rand() < 0.3) {
                push @parts, $bad_loose_seps[rand(@bad_loose_seps)];
                $want_error = 0;
            }
            else {
                push @parts, $good_loose_seps[rand(@good_loose_seps)];
            }
        }
        push @parts, $part;
    }
    my $new = join("", @parts);
    trace "$name => $new" if main::DEBUG && $to_trace;

    if ($want_error) {
        if (rand() >= 0.5) {
            $new .= $bad_loose_seps[rand(@bad_loose_seps)];
        }
        else {
            $new = $bad_loose_seps[rand(@bad_loose_seps)] . $new;
        }
    }
    return $new;
}

# Used to make sure don't generate duplicate test cases.
my %test_generated;

sub make_property_test_script() {
    # This used only for making the test script
    # this written directly -- it's huge.

    print "Making test script\n" if $verbosity >= $PROGRESS;

    # This uses randomness to test different possibilities without testing all
    # possibilities.  To ensure repeatability, set the seed to 0.  But if
    # tests are added, it will perturb all later ones in the .t file
    srand 0;

    $t_path = 'TestProp.pl' unless defined $t_path; # the traditional name

    # Create a list of what the %f representation is for each rational number.
    # This will be used below.
    my @valid_base_floats = '0.0';
    foreach my $e_representation (keys %nv_floating_to_rational) {
        push @valid_base_floats,
                            eval $nv_floating_to_rational{$e_representation};
    }

    # It doesn't matter whether the elements of this array contain single lines
    # or multiple lines. main::write doesn't count the lines.
    my @output;

    push @output, <<'EOF_CODE';
Error('\p{Script=InGreek}');    # Bug #69018
Test_GCB("1100 $nobreak 1161");  # Bug #70940
Expect(0, 0x2028, '\p{Print}', ""); # Bug # 71722
Expect(0, 0x2029, '\p{Print}', ""); # Bug # 71722
Expect(1, 0xFF10, '\p{XDigit}', ""); # Bug # 71726
Error('\p{InKana}');    # 'Kana' is not a block so InKana shouldn't compile
Expect(1, 0xB6, '\p{In=V1_1}', ""); # Didn't use to work
Expect(1, 0x3A2,'\p{In=NA}', "");   # Didn't use to work

# Make sure this gets tested; it was not part of the official test suite at
# the time this was added.  Note that this is as it would appear in the
# official suite, and gets modified to check for the perl tailoring by
# Test_WB()
Test_WB("$breakable 0020 $breakable 0020 $breakable 0308 $breakable");
Test_LB("$nobreak 200B $nobreak 0020 $nobreak 0020 $breakable 2060 $breakable");
Expect(1, ord(" "), '\p{gc=:(?aa)s:}', "");     # /aa is valid
Expect(1, ord(" "), '\p{gc=:(?-s)s:}', "");     # /-s is valid
EOF_CODE

    # Sort these so get results in same order on different runs of this
    # program
    foreach my $property (sort { $a->has_dependency <=> $b->has_dependency
                                    or
                                 lc $a->name cmp lc $b->name
                               } property_ref('*'))
    {
        # Non-binary properties should not match \p{};  Test all for that.
        if ($property->type != $BINARY && $property->type != $FORCED_BINARY) {
            my @property_aliases = grep { $_->status ne $INTERNAL_ALIAS }
                                                            $property->aliases;
            foreach my $property_alias ($property->aliases) {
                my $name = standardize($property_alias->name);

                # But some names are ambiguous, meaning a binary property with
                # the same name when used in \p{}, and a different
                # (non-binary) property in other contexts.
                next if grep { $name eq $_ } keys %ambiguous_names;

                push @output, <<"EOF_CODE";
Error('\\p{$name}');
Error('\\P{$name}');
EOF_CODE
            }
        }
        foreach my $table (sort { $a->has_dependency <=> $b->has_dependency
                                    or
                                  lc $a->name cmp lc $b->name
                                } $property->tables)
        {

            # Find code points that match, and don't match this table.
            my $valid = $table->get_valid_code_point;
            my $invalid = $table->get_invalid_code_point;
            my $warning = ($table->status eq $DEPRECATED)
                            ? "'deprecated'"
                            : '""';

            # Test each possible combination of the property's aliases with
            # the table's.  If this gets to be too many, could do what is done
            # in the set_final_comment() for Tables
            my @table_aliases = grep { $_->status ne $INTERNAL_ALIAS } $table->aliases;
            next unless @table_aliases;
            my @property_aliases = grep { $_->status ne $INTERNAL_ALIAS } $table->property->aliases;
            next unless @property_aliases;

            # Every property can be optionally be prefixed by 'Is_', so test
            # that those work, by creating such a new alias for each
            # pre-existing one.
            push @property_aliases, map { Alias->new("Is_" . $_->name,
                                                    $_->loose_match,
                                                    $_->make_re_pod_entry,
                                                    $_->ok_as_filename,
                                                    $_->status,
                                                    $_->ucd,
                                                    )
                                         } @property_aliases;
            my $max = max(scalar @table_aliases, scalar @property_aliases);
            for my $j (0 .. $max - 1) {

                # The current alias for property is the next one on the list,
                # or if beyond the end, start over.  Similarly for table
                my $property_name
                            = $property_aliases[$j % @property_aliases]->name;

                $property_name = "" if $table->property == $perl;
                my $table_alias = $table_aliases[$j % @table_aliases];
                my $table_name = $table_alias->name;
                my $loose_match = $table_alias->loose_match;

                # If the table doesn't have a file, any test for it is
                # already guaranteed to be in error
                my $already_error = ! $table->file_path;

                # A table that begins with these could actually be a
                # user-defined property, so won't be compile time errors, as
                # the definitions of those can be deferred until runtime
                next if $already_error && $table_name =~ / ^ I[ns] /x;

                # Generate error cases for this alias.
                push @output, generate_error($property_name,
                                             $table_name,
                                             $already_error);

                # If the table is guaranteed to always generate an error,
                # quit now without generating success cases.
                next if $already_error;

                # Now for the success cases.  First, wildcard matching, as it
                # shouldn't have any randomization.
                if ($table_alias->status eq $NORMAL) {
                    push @output, generate_wildcard_tests($property_name,
                                                          $table_name,
                                                          $valid,
                                                          $invalid,
                                                          $warning,
                                                         );
                }
                my $random;
                if ($loose_match) {

                    # For loose matching, create an extra test case for the
                    # standard name.
                    my $standard = standardize($table_name);

                    # $test_name should be a unique combination for each test
                    # case; used just to avoid duplicate tests
                    my $test_name = "$property_name=$standard";

                    # Don't output duplicate test cases.
                    if (! exists $test_generated{$test_name}) {
                        $test_generated{$test_name} = 1;
                        push @output, generate_tests($property_name,
                                                     $standard,
                                                     $valid,
                                                     $invalid,
                                                     $warning,
                                                 );
                        if ($table_alias->status eq $NORMAL) {
                            push @output, generate_wildcard_tests(
                                                     $property_name,
                                                     $standard,
                                                     $valid,
                                                     $invalid,
                                                     $warning,
                                                 );
                        }
                    }
                    $random = randomize_loose_name($table_name)
                }
                else { # Stricter match
                    $random = randomize_stricter_name($table_name);
                }

                # Now for the main test case for this alias.
                my $test_name = "$property_name=$random";
                if (! exists $test_generated{$test_name}) {
                    $test_generated{$test_name} = 1;
                    push @output, generate_tests($property_name,
                                                 $random,
                                                 $valid,
                                                 $invalid,
                                                 $warning,
                                             );

                    if ($property->name eq 'nv') {
                        if ($table_name !~ qr{/}) {
                            push @output, generate_tests($property_name,
                                                sprintf("%.15e", $table_name),
                                                $valid,
                                                $invalid,
                                                $warning,
                                            );
                    }
                    else {
                        # If the name is a rational number, add tests for a
                        # non-reduced form, and for a floating point equivalent.

                        # 60 is a number divisible by a bunch of things
                        my ($numerator, $denominator) = $table_name
                                                        =~ m! (.+) / (.+) !x;
                        $numerator *= 60;
                        $denominator *= 60;
                        push @output, generate_tests($property_name,
                                                    "$numerator/$denominator",
                                                    $valid,
                                                    $invalid,
                                                    $warning,
                                    );

                        # Calculate the float, and the %e representation
                        my $float = eval $table_name;
                        my $e_representation = sprintf("%.*e",
                                                $E_FLOAT_PRECISION, $float);
                        # Parse that
                        my ($non_zeros, $zeros, $exponent_sign, $exponent)
                           = $e_representation
                               =~ / -? [1-9] \. (\d*?) (0*) e ([+-]) (\d+) /x;
                        my $min_e_precision;
                        my $min_f_precision;

                        if ($exponent_sign eq '+' && $exponent != 0) {
                            Carp::my_carp_bug("Not yet equipped to handle"
                                            . " positive exponents");
                            return;
                        }
                        else {
                            # We're trying to find the minimum precision that
                            # is needed to indicate this particular rational
                            # for the given $E_FLOAT_PRECISION.  For %e, any
                            # trailing zeros, like 1.500e-02 aren't needed, so
                            # the correct value is how many non-trailing zeros
                            # there are after the decimal point.
                            $min_e_precision = length $non_zeros;

                            # For %f, like .01500, we want at least
                            # $E_FLOAT_PRECISION digits, but any trailing
                            # zeros aren't needed, so we can subtract the
                            # length of those.  But we also need to include
                            # the zeros after the decimal point, but before
                            # the first significant digit.
                            $min_f_precision = $E_FLOAT_PRECISION
                                             + $exponent
                                             - length $zeros;
                        }

                        # Make tests for each possible precision from 1 to
                        # just past the worst case.
                        my $upper_limit = ($min_e_precision > $min_f_precision)
                                           ? $min_e_precision
                                           : $min_f_precision;

                        for my $i (1 .. $upper_limit + 1) {
                            for my $format ("e", "f") {
                                my $this_table
                                          = sprintf("%.*$format", $i, $float);

                                # If we don't have enough precision digits,
                                # make a fail test; otherwise a pass test.
                                my $pass = ($format eq "e")
                                            ? $i >= $min_e_precision
                                            : $i >= $min_f_precision;
                                if ($pass) {
                                    push @output, generate_tests($property_name,
                                                                $this_table,
                                                                $valid,
                                                                $invalid,
                                                                $warning,
                                                );
                                }
                                elsif (   $format eq "e"

                                          # Here we would fail, but in the %f
                                          # case, the representation at this
                                          # precision could actually be a
                                          # valid one for some other rational
                                       || ! grep { $this_table
                                                            =~ / ^ $_ 0* $ /x }
                                                            @valid_base_floats)
                                {
                                    push @output,
                                        generate_error($property_name,
                                                       $this_table,
                                                       1   # 1 => already an
                                                           # error
                                                );
                                }
                            }
                        }
                    }
                    }
                }
            }
            $table->DESTROY();
        }
        $property->DESTROY();
    }

    # Make any test of the boundary (break) properties TODO if the code
    # doesn't match the version being compiled
    my $TODO_FAILING_BREAKS = ($version_of_mk_invlist_bounds ne $v_version)
                             ? "\nsub TODO_FAILING_BREAKS { 1 }\n"
                             : "\nsub TODO_FAILING_BREAKS { 0 }\n";

    @output= map {
        map s/^/    /mgr,
        map "$_;\n",
        split /;\n/, $_
    } @output;

    # Cause there to be 'if' statements to only execute a portion of this
    # long-running test each time, so that we can have a bunch of .t's running
    # in parallel
    my $chunks = 10     # Number of test files
               - 1      # For GCB & SB
               - 1      # For WB
               - 4;     # LB split into this many files
    my @output_chunked;
    my $chunk_count=0;
    my $chunk_size= int(@output / $chunks) + 1;
    while (@output) {
        $chunk_count++;
        my @chunk= splice @output, 0, $chunk_size;
        push @output_chunked,
            "if (!\$::TESTCHUNK or \$::TESTCHUNK == $chunk_count) {\n",
                @chunk,
            "}\n";
    }

    $chunk_count++;
    push @output_chunked,
        "if (!\$::TESTCHUNK or \$::TESTCHUNK == $chunk_count) {\n",
            (map {"    Test_GCB('$_');\n"} @backslash_X_tests),
            (map {"    Test_SB('$_');\n"} @SB_tests),
        "}\n";


    $chunk_size= int(@LB_tests / 4) + 1;
    @LB_tests = map {"    Test_LB('$_');\n"} @LB_tests;
    while (@LB_tests) {
        $chunk_count++;
        my @chunk= splice @LB_tests, 0, $chunk_size;
        push @output_chunked,
            "if (!\$::TESTCHUNK or \$::TESTCHUNK == $chunk_count) {\n",
                @chunk,
            "}\n";
    }

    $chunk_count++;
    push @output_chunked,
        "if (!\$::TESTCHUNK or \$::TESTCHUNK == $chunk_count) {\n",
            (map {"    Test_WB('$_');\n"} @WB_tests),
        "}\n";

    &write($t_path,
           0,           # Not utf8;
           [$HEADER,
            $TODO_FAILING_BREAKS,
            <DATA>,
            @output_chunked,
            "Finished();\n",
           ]);

    return;
}

sub make_normalization_test_script() {
    print "Making normalization test script\n" if $verbosity >= $PROGRESS;

    my $n_path = 'TestNorm.pl';

    unshift @normalization_tests, <<'END';
use utf8;
use Test::More;

sub ord_string {    # Convert packed ords to printable string
    use charnames ();
    return "'" . join("", map { '\N{' . charnames::viacode($_) . '}' }
                                                unpack "U*", shift) .  "'";
    #return "'" . join(" ", map { sprintf "%04X", $_ } unpack "U*", shift) .  "'";
}

sub Test_N {
    my ($source, $nfc, $nfd, $nfkc, $nfkd) = @_;
    my $display_source = ord_string($source);
    my $display_nfc = ord_string($nfc);
    my $display_nfd = ord_string($nfd);
    my $display_nfkc = ord_string($nfkc);
    my $display_nfkd = ord_string($nfkd);

    use Unicode::Normalize;
    #    NFC
    #      nfc ==  toNFC(source) ==  toNFC(nfc) ==  toNFC(nfd)
    #      nfkc ==  toNFC(nfkc) ==  toNFC(nfkd)
    #
    #    NFD
    #      nfd ==  toNFD(source) ==  toNFD(nfc) ==  toNFD(nfd)
    #      nfkd ==  toNFD(nfkc) ==  toNFD(nfkd)
    #
    #    NFKC
    #      nfkc == toNFKC(source) == toNFKC(nfc) == toNFKC(nfd) ==
    #      toNFKC(nfkc) == toNFKC(nfkd)
    #
    #    NFKD
    #      nfkd == toNFKD(source) == toNFKD(nfc) == toNFKD(nfd) ==
    #      toNFKD(nfkc) == toNFKD(nfkd)

    is(NFC($source), $nfc, "NFC($display_source) eq $display_nfc");
    is(NFC($nfc), $nfc, "NFC($display_nfc) eq $display_nfc");
    is(NFC($nfd), $nfc, "NFC($display_nfd) eq $display_nfc");
    is(NFC($nfkc), $nfkc, "NFC($display_nfkc) eq $display_nfkc");
    is(NFC($nfkd), $nfkc, "NFC($display_nfkd) eq $display_nfkc");

    is(NFD($source), $nfd, "NFD($display_source) eq $display_nfd");
    is(NFD($nfc), $nfd, "NFD($display_nfc) eq $display_nfd");
    is(NFD($nfd), $nfd, "NFD($display_nfd) eq $display_nfd");
    is(NFD($nfkc), $nfkd, "NFD($display_nfkc) eq $display_nfkd");
    is(NFD($nfkd), $nfkd, "NFD($display_nfkd) eq $display_nfkd");

    is(NFKC($source), $nfkc, "NFKC($display_source) eq $display_nfkc");
    is(NFKC($nfc), $nfkc, "NFKC($display_nfc) eq $display_nfkc");
    is(NFKC($nfd), $nfkc, "NFKC($display_nfd) eq $display_nfkc");
    is(NFKC($nfkc), $nfkc, "NFKC($display_nfkc) eq $display_nfkc");
    is(NFKC($nfkd), $nfkc, "NFKC($display_nfkd) eq $display_nfkc");

    is(NFKD($source), $nfkd, "NFKD($display_source) eq $display_nfkd");
    is(NFKD($nfc), $nfkd, "NFKD($display_nfc) eq $display_nfkd");
    is(NFKD($nfd), $nfkd, "NFKD($display_nfd) eq $display_nfkd");
    is(NFKD($nfkc), $nfkd, "NFKD($display_nfkc) eq $display_nfkd");
    is(NFKD($nfkd), $nfkd, "NFKD($display_nfkd) eq $display_nfkd");
}
END

    &write($n_path,
           1,           # Is utf8;
           [
            @normalization_tests,
            'done_testing();'
            ]);
    return;
}

# Skip reasons, so will be exact same text and hence the files with each
# reason will get grouped together in perluniprops.
my $Documentation = "Documentation";
my $Indic_Skip
            = "Provisional; for the analysis and processing of Indic scripts";
my $Validation = "Validation Tests";
my $Validation_Documentation = "Documentation of validation Tests";
my $Unused_Skip = "Currently unused by Perl";

# This is a list of the input files and how to handle them.  The files are
# processed in their order in this list.  Some reordering is possible if
# desired, but the PropertyAliases and PropValueAliases files should be first,
# and the extracted before the others (as data in an extracted file can be
# over-ridden by the non-extracted.  Some other files depend on data derived
# from an earlier file, like UnicodeData requires data from Jamo, and the case
# changing and folding requires data from Unicode.  Mostly, it is safest to
# order by first version releases in (except the Jamo).
#
# The version strings allow the program to know whether to expect a file or
# not, but if a file exists in the directory, it will be processed, even if it
# is in a version earlier than expected, so you can copy files from a later
# release into an earlier release's directory.
my @input_file_objects = (
    Input_file->new('PropertyAliases.txt', v3.2,
                    Handler => \&process_PropertyAliases,
                    Early => [ \&substitute_PropertyAliases ],
                    Required_Even_in_Debug_Skip => 1,
                   ),
    Input_file->new(undef, v0,  # No file associated with this
                    Progress_Message => 'Finishing property setup',
                    Handler => \&finish_property_setup,
                   ),
    Input_file->new('PropValueAliases.txt', v3.2,
                     Handler => \&process_PropValueAliases,
                     Early => [ \&substitute_PropValueAliases ],
                     Has_Missings_Defaults => $NOT_IGNORED,
                     Required_Even_in_Debug_Skip => 1,
                    ),
    Input_file->new("${EXTRACTED}DGeneralCategory.txt", v3.1.0,
                    Property => 'General_Category',
                   ),
    Input_file->new("${EXTRACTED}DCombiningClass.txt", v3.1.0,
                    Property => 'Canonical_Combining_Class',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DNumType.txt", v3.1.0,
                    Property => 'Numeric_Type',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DEastAsianWidth.txt", v3.1.0,
                    Property => 'East_Asian_Width',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DLineBreak.txt", v3.1.0,
                    Property => 'Line_Break',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DBidiClass.txt", v3.1.1,
                    Property => 'Bidi_Class',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DDecompositionType.txt", v3.1.0,
                    Property => 'Decomposition_Type',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DBinaryProperties.txt", v3.1.0),
    Input_file->new("${EXTRACTED}DNumValues.txt", v3.1.0,
                    Property => 'Numeric_Value',
                    Each_Line_Handler => \&filter_numeric_value_line,
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DJoinGroup.txt", v3.1.0,
                    Property => 'Joining_Group',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),

    Input_file->new("${EXTRACTED}DJoinType.txt", v3.1.0,
                    Property => 'Joining_Type',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("${EXTRACTED}DName.txt", v10.0.0,
                    Skip => 'This file adds no new information not already'
                          . ' present in other files',
                    # And it's unnecessary programmer work to handle this new
                    # format.  Previous Derived files actually had bug fixes
                    # in them that were useful, but that should not be the
                    # case here.
                   ),
    Input_file->new('Jamo.txt', v2.0.0,
                    Property => 'Jamo_Short_Name',
                    Each_Line_Handler => \&filter_jamo_line,
                   ),
    Input_file->new('UnicodeData.txt', v1.1.5,
                    Pre_Handler => \&setup_UnicodeData,

                    # We clean up this file for some early versions.
                    Each_Line_Handler => [ (($v_version lt v2.0.0 )
                                            ? \&filter_v1_ucd
                                            : ($v_version eq v2.1.5)
                                                ? \&filter_v2_1_5_ucd

                                                # And for 5.14 Perls with 6.0,
                                                # have to also make changes
                                                : ($v_version ge v6.0.0
                                                   && $^V lt v5.17.0)
                                                    ? \&filter_v6_ucd
                                                    : undef),

                                            # Early versions did not have the
                                            # proper Unicode_1 names for the
                                            # controls
                                            (($v_version lt v3.0.0)
                                            ? \&filter_early_U1_names
                                            : undef),

                                            # Early versions did not correctly
                                            # use the later method for giving
                                            # decimal digit values
                                            (($v_version le v3.2.0)
                                            ? \&filter_bad_Nd_ucd
                                            : undef),

                                            # And the main filter
                                            \&filter_UnicodeData_line,
                                         ],
                    EOF_Handler => \&EOF_UnicodeData,
                   ),
    Input_file->new('CJKXREF.TXT', v1.1.5,
                    Withdrawn => v2.0.0,
                    Skip => 'Gives the mapping of CJK code points '
                          . 'between Unicode and various other standards',
                   ),
    Input_file->new('ArabicShaping.txt', v2.0.0,
                    Each_Line_Handler =>
                        ($v_version lt 4.1.0)
                                    ? \&filter_old_style_arabic_shaping
                                    : undef,
                    # The first field after the range is a "schematic name"
                    # not used by Perl
                    Properties => [ '<ignored>', 'Joining_Type', 'Joining_Group' ],
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('Blocks.txt', v2.0.0,
                    Property => 'Block',
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Each_Line_Handler => \&filter_blocks_lines
                   ),
    Input_file->new('Index.txt', v2.0.0,
                    Skip => 'Alphabetical index of Unicode characters',
                   ),
    Input_file->new('NamesList.txt', v2.0.0,
                    Skip => 'Annotated list of characters',
                   ),
    Input_file->new('PropList.txt', v2.0.0,
                    Each_Line_Handler => (($v_version lt v3.1.0)
                                            ? \&filter_old_style_proplist
                                            : undef),
                   ),
    Input_file->new('Props.txt', v2.0.0,
                    Withdrawn => v3.0.0,
                    Skip => 'A subset of F<PropList.txt> (which is used instead)',
                   ),
    Input_file->new('ReadMe.txt', v2.0.0,
                    Skip => $Documentation,
                   ),
    Input_file->new('Unihan.txt', v2.0.0,
                    Withdrawn => v5.2.0,
                    Construction_Time_Handler => \&construct_unihan,
                    Pre_Handler => \&setup_unihan,
                    Optional => [ "",
                                  'Unicode_Radical_Stroke'
                                ],
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('SpecialCasing.txt', v2.1.8,
                    Each_Line_Handler => ($v_version eq 2.1.8)
                                         ? \&filter_2_1_8_special_casing_line
                                         : \&filter_special_casing_line,
                    Pre_Handler => \&setup_special_casing,
                    Has_Missings_Defaults => $IGNORED,
                   ),
    Input_file->new(
                    'LineBreak.txt', v3.0.0,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Property => 'Line_Break',
                    # Early versions had problematic syntax
                    Each_Line_Handler => ($v_version ge v3.1.0)
                                          ? undef
                                          : ($v_version lt v3.0.0)
                                            ? \&filter_substitute_lb
                                            : \&filter_early_ea_lb,
                    # Must use long names for property values see comments at
                    # sub filter_substitute_lb
                    Early => [ "LBsubst.txt", '_Perl_LB', 'Alphabetic',
                               'Alphabetic', # default to this because XX ->
                                             # AL

                               # Don't use _Perl_LB as a synonym for
                               # Line_Break in later perls, as it is tailored
                               # and isn't the same as Line_Break
                               'ONLY_EARLY' ],
                   ),
    Input_file->new('EastAsianWidth.txt', v3.0.0,
                    Property => 'East_Asian_Width',
                    Has_Missings_Defaults => $NOT_IGNORED,
                    # Early versions had problematic syntax
                    Each_Line_Handler => (($v_version lt v3.1.0)
                                        ? \&filter_early_ea_lb
                                        : undef),
                   ),
    Input_file->new('CompositionExclusions.txt', v3.0.0,
                    Property => 'Composition_Exclusion',
                   ),
    Input_file->new('UnicodeData.html', v3.0.0,
                    Withdrawn => v4.0.1,
                    Skip => $Documentation,
                   ),
    Input_file->new('BidiMirroring.txt', v3.0.1,
                    Property => 'Bidi_Mirroring_Glyph',
                    Has_Missings_Defaults => ($v_version lt v6.2.0)
                                              ? $NO_DEFAULTS
                                              # Is <none> which doesn't mean
                                              # anything to us, we will use the
                                              # null string
                                              : $IGNORED,
                   ),
    Input_file->new('NamesList.html', v3.0.0,
                    Skip => 'Describes the format and contents of '
                          . 'F<NamesList.txt>',
                   ),
    Input_file->new('UnicodeCharacterDatabase.html', v3.0.0,
                    Withdrawn => v5.1,
                    Skip => $Documentation,
                   ),
    Input_file->new('CaseFolding.txt', v3.0.1,
                    Pre_Handler => \&setup_case_folding,
                    Each_Line_Handler =>
                        [ ($v_version lt v3.1.0)
                                 ? \&filter_old_style_case_folding
                                 : undef,
                           \&filter_case_folding_line
                        ],
                    Has_Missings_Defaults => $IGNORED,
                   ),
    Input_file->new("NormTest.txt", v3.0.1,
                     Handler => \&process_NormalizationsTest,
                     Skip => ($make_norm_test_script) ? 0 : $Validation,
                   ),
    Input_file->new('DCoreProperties.txt', v3.1.0,
                    # 5.2 changed this file
                    Has_Missings_Defaults => (($v_version ge v5.2.0)
                                            ? $NOT_IGNORED
                                            : $NO_DEFAULTS),
                   ),
    Input_file->new('DProperties.html', v3.1.0,
                    Withdrawn => v3.2.0,
                    Skip => $Documentation,
                   ),
    Input_file->new('PropList.html', v3.1.0,
                    Withdrawn => v5.1,
                    Skip => $Documentation,
                   ),
    Input_file->new('Scripts.txt', v3.1.0,
                    Property => 'Script',
                    Each_Line_Handler => (($v_version le v4.0.0)
                                          ? \&filter_all_caps_script_names
                                          : undef),
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('DNormalizationProps.txt', v3.1.0,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Each_Line_Handler => (($v_version lt v4.0.1)
                                      ? \&filter_old_style_normalization_lines
                                      : undef),
                   ),
    Input_file->new('DerivedProperties.html', v3.1.1,
                    Withdrawn => v5.1,
                    Skip => $Documentation,
                   ),
    Input_file->new('DAge.txt', v3.2.0,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Property => 'Age'
                   ),
    Input_file->new('HangulSyllableType.txt', v4.0,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Early => [ \&generate_hst, 'Hangul_Syllable_Type' ],
                    Property => 'Hangul_Syllable_Type'
                   ),
    Input_file->new('NormalizationCorrections.txt', v3.2.0,
                     # This documents the cumulative fixes to erroneous
                     # normalizations in earlier Unicode versions.  Its main
                     # purpose is so that someone running on an earlier
                     # version can use this file to override what got
                     # published in that earlier release.  It would be easy
                     # for mktables to handle this file.  But all the
                     # corrections in it should already be in the other files
                     # for the release it is.  To get it to actually mean
                     # something useful, someone would have to be using an
                     # earlier Unicode release, and copy it into the directory
                     # for that release and recompile.  So far there has been
                     # no demand to do that, so this hasn't been implemented.
                    Skip => 'Documentation of corrections already '
                          . 'incorporated into the Unicode data base',
                   ),
    Input_file->new('StandardizedVariants.html', v3.2.0,
                    Skip => 'Obsoleted as of Unicode 9.0, but previously '
                          . 'provided a visual display of the standard '
                          . 'variant sequences derived from '
                          . 'F<StandardizedVariants.txt>.',
                        # I don't know why the html came earlier than the
                        # .txt, but both are skipped anyway, so it doesn't
                        # matter.
                   ),
    Input_file->new('StandardizedVariants.txt', v4.0.0,
                    Skip => 'Certain glyph variations for character display '
                          . 'are standardized.  This lists the non-Unihan '
                          . 'ones; the Unihan ones are also not used by '
                          . 'Perl, and are in a separate Unicode data base '
                          . 'L<http://www.unicode.org/ivd>',
                   ),
    Input_file->new('UCD.html', v4.0.0,
                    Withdrawn => v5.2,
                    Skip => $Documentation,
                   ),
    Input_file->new("$AUXILIARY/WordBreakProperty.txt", v4.1.0,
                    Early => [ "WBsubst.txt", '_Perl_WB', 'ALetter' ],
                    Property => 'Word_Break',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("$AUXILIARY/GraphemeBreakProperty.txt", v4.1,
                    Early => [ \&generate_GCB, '_Perl_GCB' ],
                    Property => 'Grapheme_Cluster_Break',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new("$AUXILIARY/GCBTest.txt", v4.1.0,
                    Handler => \&process_GCB_test,
                    retain_trailing_comments => 1,
                   ),
    Input_file->new("$AUXILIARY/GraphemeBreakTest.html", v4.1.0,
                    Skip => $Validation_Documentation,
                   ),
    Input_file->new("$AUXILIARY/SBTest.txt", v4.1.0,
                    Handler => \&process_SB_test,
                    retain_trailing_comments => 1,
                   ),
    Input_file->new("$AUXILIARY/SentenceBreakTest.html", v4.1.0,
                    Skip => $Validation_Documentation,
                   ),
    Input_file->new("$AUXILIARY/WBTest.txt", v4.1.0,
                    Handler => \&process_WB_test,
                    retain_trailing_comments => 1,
                   ),
    Input_file->new("$AUXILIARY/WordBreakTest.html", v4.1.0,
                    Skip => $Validation_Documentation,
                   ),
    Input_file->new("$AUXILIARY/SentenceBreakProperty.txt", v4.1.0,
                    Property => 'Sentence_Break',
                    Early => [ "SBsubst.txt", '_Perl_SB', 'OLetter' ],
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('NamedSequences.txt', v4.1.0,
                    Handler => \&process_NamedSequences
                   ),
    Input_file->new('Unihan.html', v4.1.0,
                    Withdrawn => v5.2,
                    Skip => $Documentation,
                   ),
    Input_file->new('NameAliases.txt', v5.0,
                    Property => 'Name_Alias',
                    Each_Line_Handler => ($v_version le v6.0.0)
                                   ? \&filter_early_version_name_alias_line
                                   : \&filter_later_version_name_alias_line,
                   ),
        # NameAliases.txt came along in v5.0.  The above constructor handles
        # this.  But until 6.1, it was lacking some information needed by core
        # perl.  The constructor below handles that.  It is either a kludge or
        # clever, depending on your point of view.  The 'Withdrawn' parameter
        # indicates not to use it at all starting in 6.1 (so the above
        # constructor applies), and the 'v6.1' parameter indicates to use the
        # Early parameter before 6.1.  Therefore 'Early" is always used,
        # yielding the internal-only property '_Perl_Name_Alias', which it
        # gets from a NameAliases.txt from 6.1 or later stored in
        # N_Asubst.txt.  In combination with the above constructor,
        # 'Name_Alias' is publicly accessible starting with v5.0, and the
        # better 6.1 version is accessible to perl core in all releases.
    Input_file->new("NameAliases.txt", v6.1,
                    Withdrawn => v6.1,
                    Early => [ "N_Asubst.txt", '_Perl_Name_Alias', "" ],
                    Property => 'Name_Alias',
                    EOF_Handler => \&fixup_early_perl_name_alias,
                    Each_Line_Handler =>
                                       \&filter_later_version_name_alias_line,
                   ),
    Input_file->new('NamedSqProv.txt', v5.0.0,
                    Skip => 'Named sequences proposed for inclusion in a '
                          . 'later version of the Unicode Standard; if you '
                          . 'need them now, you can append this file to '
                          . 'F<NamedSequences.txt> and recompile perl',
                   ),
    Input_file->new("$AUXILIARY/LBTest.txt", v5.1.0,
                    Handler => \&process_LB_test,
                    retain_trailing_comments => 1,
                   ),
    Input_file->new("$AUXILIARY/LineBreakTest.html", v5.1.0,
                    Skip => $Validation_Documentation,
                   ),
    Input_file->new("BidiTest.txt", v5.2.0,
                    Skip => $Validation,
                   ),
    Input_file->new('UnihanIndicesDictionary.txt', v5.2.0,
                    Optional => "",
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanDataDictionaryLike.txt', v5.2.0,
                    Optional => "",
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanIRGSources.txt', v5.2.0,
                    Optional => [ "",
                                  'kCompatibilityVariant',
                                  'kIICore',
                                  'kIRG_GSource',
                                  'kIRG_HSource',
                                  'kIRG_JSource',
                                  'kIRG_KPSource',
                                  'kIRG_MSource',
                                  'kIRG_KSource',
                                  'kIRG_SSource',
                                  'kIRG_TSource',
                                  'kIRG_USource',
                                  'kIRG_UKSource',
                                  'kIRG_VSource',
                               ],
                    Pre_Handler => \&setup_unihan,
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanNumericValues.txt', v5.2.0,
                    Optional => [ "",
                                  'kAccountingNumeric',
                                  'kOtherNumeric',
                                  'kPrimaryNumeric',
                                ],
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanOtherMappings.txt', v5.2.0,
                    Optional => "",
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanRadicalStrokeCounts.txt', v5.2.0,
                    Optional => [ "",
                                  'Unicode_Radical_Stroke'
                                ],
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanReadings.txt', v5.2.0,
                    Optional => "",
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('UnihanVariants.txt', v5.2.0,
                    Optional => "",
                    Each_Line_Handler => \&filter_unihan_line,
                   ),
    Input_file->new('CJKRadicals.txt', v5.2.0,
                    Skip => 'Maps the kRSUnicode property values to '
                          . 'corresponding code points',
                   ),
    Input_file->new('EmojiSources.txt', v6.0.0,
                    Skip => 'Maps certain Unicode code points to their '
                          . 'legacy Japanese cell-phone values',
                   ),
    # This file is actually not usable as-is until 6.1.0, because the property
    # is provisional, so its name is missing from PropertyAliases.txt until
    # that release, so that further work would have to be done to get it to
    # work properly
    Input_file->new('ScriptExtensions.txt', v6.0.0,
                    Property => 'Script_Extensions',
                    Early => [ sub {} ], # Doesn't do anything but ensures
                                         # that this isn't skipped for early
                                         # versions
                    Pre_Handler => \&setup_script_extensions,
                    Each_Line_Handler => \&filter_script_extensions_line,
                    Has_Missings_Defaults => (($v_version le v6.0.0)
                                            ? $NO_DEFAULTS
                                            : $IGNORED),
                   ),
    # These two Indic files are actually not usable as-is until 6.1.0,
    # because they are provisional, so their property values are missing from
    # PropValueAliases.txt until that release, so that further work would have
    # to be done to get them to work properly.
    Input_file->new('IndicMatraCategory.txt', v6.0.0,
                    Withdrawn => v8.0.0,
                    Property => 'Indic_Matra_Category',
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Skip => $Indic_Skip,
                   ),
    Input_file->new('IndicSyllabicCategory.txt', v6.0.0,
                    Property => 'Indic_Syllabic_Category',
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Skip => (($v_version lt v8.0.0)
                              ? $Indic_Skip
                              : 0),
                   ),
    Input_file->new('USourceData.txt', v6.2.0,
                    Skip => 'Documentation of status and cross reference of '
                          . 'proposals for encoding by Unicode of Unihan '
                          . 'characters',
                   ),
    Input_file->new('USourceGlyphs.pdf', v6.2.0,
                    Skip => 'Pictures of the characters in F<USourceData.txt>',
                   ),
    Input_file->new('BidiBrackets.txt', v6.3.0,
                    Properties => [ 'Bidi_Paired_Bracket',
                                    'Bidi_Paired_Bracket_Type'
                                  ],
                    Has_Missings_Defaults => $NO_DEFAULTS,
                   ),
    Input_file->new("BidiCharacterTest.txt", v6.3.0,
                    Skip => $Validation,
                   ),
    Input_file->new('IndicPositionalCategory.txt', v8.0.0,
                    Property => 'Indic_Positional_Category',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('TangutSources.txt', v9.0.0,
                    Skip => 'Specifies source mappings for Tangut ideographs'
                          . ' and components. This data file also includes'
                          . ' informative radical-stroke values that are used'
                          . ' internally by Unicode',
                   ),
    Input_file->new('VerticalOrientation.txt', v10.0.0,
                    Property => 'Vertical_Orientation',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('NushuSources.txt', v10.0.0,
                    Skip => 'Specifies source material for Nushu characters',
                   ),
    Input_file->new('EquivalentUnifiedIdeograph.txt', v11.0.0,
                    Property => 'Equivalent_Unified_Ideograph',
                    Has_Missings_Defaults => $NOT_IGNORED,
                   ),
    Input_file->new('EmojiData.txt', v11.0.0,
                    # Is in UAX #51 and not the UCD, so must be updated
                    # separately, and the first line edited to indicate the
                    # UCD release we're pretending it to be in.  The UTC says
                    # this is a transitional state, and in fact was moved to
                    # the UCD in 13.0
                    Withdrawn => v13.0.0,
                    Pre_Handler => \&setup_emojidata,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    Each_Line_Handler => \&filter_emojidata_line,
                    UCD => 0,
                   ),
    Input_file->new("$EMOJI/emoji.txt", v13.0.0,
                    Has_Missings_Defaults => $NOT_IGNORED,
                    UCD => 0,
                   ),
    Input_file->new("$EMOJI/ReadMe.txt", v13.0.0,
                    Skip => $Documentation,
                    UCD => 0,
                   ),
    Input_file->new('IdStatus.txt', v13.0.0,
                    Pre_Handler => \&setup_IdStatus,
                    Has_Missings_Defaults => $IGNORED,
                    Property => 'Identifier_Status',

                    # Part of UTS 39, so must be downloaded separately from
                    # unicode.org
                    UCD => 0,
                   ),
    Input_file->new('IdType.txt', v13.0.0,
                    Pre_Handler => \&setup_IdType,
                    Has_Missings_Defaults => $IGNORED,
                    Each_Line_Handler => \&filter_IdType_line,
                    Property => 'Identifier_Type',

                    # Part of UTS 39, so must be downloaded separately from
                    # unicode.org
                    UCD => 0,
                   ),
    Input_file->new('confusables.txt', v15.0.0,
                    Skip => $Unused_Skip,
                    UCD => 0,
                   ),
    Input_file->new('confusablesSummary.txt', v15.0.0,
                    Skip => $Unused_Skip,
                    UCD => 0,
                   ),
    Input_file->new('intentional.txt', v15.0.0,
                    Skip => $Unused_Skip,
                    UCD => 0,
                   ),
    Input_file->new('Unikemet.txt', v16.0.0,
                    # For Egyptian Hieroglyphs; is in an alien format to the
                    # other files Unicode furnishes.
                    Skip => $Unused_Skip,
                    UCD => 0,
                   ),
    Input_file->new('DoNotEmit.txt', v16.0.0,
                    # Advice about characters that are unwise to create; not
                    # any properties, though we could create some.
                    Skip => $Unused_Skip,
                    UCD => 0,
                   ),
);

# End of all the preliminaries.
# Do it...

if (@missing_early_files) {
    print simple_fold(join_lines(<<END

The compilation cannot be completed because one or more required input files,
listed below, are missing.  This is because you are compiling Unicode version
$unicode_version, which predates the existence of these file(s).  To fully
function, perl needs the data that these files would have contained if they
had been in this release.  To work around this, create copies of later
versions of the missing files in the directory containing '$0'.  (Perl will
make the necessary adjustments to the data to compensate for it not being the
same version as is being compiled.)  The files are available from unicode.org,
via either ftp or http.  If using http, they will be under
www.unicode.org/versions/.  Below are listed the source file name of each
missing file, the Unicode version to copy it from, and the name to store it
as.  (Note that the listed source file name may not be exactly the one that
Unicode calls it.  If you don't find it, you can look it up in 'README.perl'
to get the correct name.)
END
    ));
    print simple_fold(join_lines("\n$_")) for @missing_early_files;
    exit 2;
}

if ($compare_versions) {
    Carp::my_carp(<<END
Warning.  \$compare_versions is set.  Output is not suitable for production
END
    );
}

# Put into %potential_files a list of all the files in the directory structure
# that could be inputs to this program
File::Find::find({
    wanted=>sub {
        return unless / \. ( txt | htm l? ) $ /xi;  # Some platforms change the
                                                    # name's case
        my $full = lc(File::Spec->rel2abs($_));
        $potential_files{$full} = 1;
        return;
    }
}, File::Spec->curdir());

my @mktables_list_output_files;
my $old_start_time = 0;
my $old_options = "";

if (! -e $file_list) {
    print "'$file_list' doesn't exist, so forcing rebuild.\n" if $verbosity >= $VERBOSE;
    $write_unchanged_files = 1;
} elsif ($write_unchanged_files) {
    print "Not checking file list '$file_list'.\n" if $verbosity >= $VERBOSE;
}
else {
    print "Reading file list '$file_list'\n" if $verbosity >= $VERBOSE;
    my $file_handle;
    if (! open $file_handle, "<", $file_list) {
        Carp::my_carp("Failed to open '$file_list'; turning on -globlist option instead: $!");
        $glob_list = 1;
    }
    else {
        my @input;

        # Read and parse mktables.lst, placing the results from the first part
        # into @input, and the second part into @mktables_list_output_files
        for my $list ( \@input, \@mktables_list_output_files ) {
            while (<$file_handle>) {
                s/^ \s+ | \s+ $//xg;
                if (/^ \s* \# \s* Autogenerated\ starting\ on\ (\d+)/x) {
                    $old_start_time = $1;
                    next;
                }
                if (/^ \s* \# \s* From\ options\ (.+) /x) {
                    $old_options = $1;
                    next;
                }
                next if /^ \s* (?: \# .* )? $/x;
                last if /^ =+ $/x;
                my ( $file ) = split /\t/;
                push @$list, $file;
            }
            @$list = uniques(@$list);
            next;
        }

        # Look through all the input files
        foreach my $input (@input) {
            next if $input eq 'version'; # Already have checked this.

            # Ignore if doesn't exist.  The checking about whether we care or
            # not is done via the Input_file object.
            next if ! file_exists($input);

            # The paths are stored with relative names, and with '/' as the
            # delimiter; convert to absolute on this machine
            my $full = lc(File::Spec->rel2abs(internal_file_to_platform($input)));
            $potential_files{lc $full} = 1;
        }
    }

    close $file_handle;
}

if ($glob_list) {

    # Here wants to process all .txt files in the directory structure.
    # Convert them to full path names.  They are stored in the platform's
    # relative style
    my @known_files;
    foreach my $object (@input_file_objects) {
        my $file = $object->file;
        next unless defined $file;
        push @known_files, File::Spec->rel2abs($file);
    }

    my @unknown_input_files;
    foreach my $file (keys %potential_files) {  # The keys are stored in lc
        next if grep { $file eq lc($_) } @known_files;

        # Here, the file is unknown to us.  Get relative path name
        $file = File::Spec->abs2rel($file);
        push @unknown_input_files, $file;

        # What will happen is we create a data structure for it, and add it to
        # the list of input files to process.  First get the subdirectories
        # into an array
        my (undef, $directories, undef) = File::Spec->splitpath($file);
        $directories =~ s;/$;;;     # Can have extraneous trailing '/'
        my @directories = File::Spec->splitdir($directories);

        # If the file isn't extracted (meaning none of the directories is the
        # extracted one), just add it to the end of the list of inputs.
        if (! grep { $EXTRACTED_DIR eq $_ } @directories) {
            push @input_file_objects, Input_file->new($file, v0);
        }
        else {

            # Here, the file is extracted.  It needs to go ahead of most other
            # processing.  Search for the first input file that isn't a
            # special required property (that is, find one whose first_release
            # is non-0), and isn't extracted.  Also, the Age property file is
            # processed before the extracted ones, just in case
            # $compare_versions is set.
            for (my $i = 0; $i < @input_file_objects; $i++) {
                if ($input_file_objects[$i]->first_released ne v0
                    && lc($input_file_objects[$i]->file) ne 'dage.txt'
                    && $input_file_objects[$i]->file !~ /$EXTRACTED_DIR/i)
                {
                    splice @input_file_objects, $i, 0,
                                                Input_file->new($file, v0);
                    last;
                }
            }

        }
    }
    if (@unknown_input_files) {
        print STDERR simple_fold(join_lines(<<END

The following files are unknown as to how to handle.  Assuming they are
typical property files.  You'll know by later error messages if it worked or
not:
END
        ) . " " . join(", ", @unknown_input_files) . "\n\n");
    }
} # End of looking through directory structure for more .txt files.

# Create the list of input files from the objects we have defined, plus
# version
my @input_files = qw(version Makefile);
foreach my $object (@input_file_objects) {
    my $file = $object->file;
    next if ! defined $file;    # Not all objects have files
    next if defined $object->skip;;
    push @input_files,  $file;
}

if ( $verbosity >= $VERBOSE ) {
    print "Expecting ".scalar( @input_files )." input files. ",
         "Checking ".scalar( @mktables_list_output_files )." output files.\n";
}

# We set $most_recent to be the most recently changed input file, including
# this program itself (done much earlier in this file)
foreach my $in (@input_files) {
    next unless -e $in;        # Keep going even if missing a file
    my $mod_time = (stat $in)[9];
    $most_recent = $mod_time if $mod_time > $most_recent;

    # See that the input files have distinct names, to warn someone if they
    # are adding a new one
    if ($make_list) {
        my ($volume, $directories, $file ) = File::Spec->splitpath($in);
        $directories =~ s;/$;;;     # Can have extraneous trailing '/'
        my @directories = File::Spec->splitdir($directories);
        construct_filename($file, 'mutable', \@directories);
    }
}

# We use 'Makefile' just to see if it has changed since the last time we
# rebuilt.  Now discard it.
@input_files = grep { $_ ne 'Makefile' } @input_files;

my $rebuild = $write_unchanged_files    # Rebuild: if unconditional rebuild
              || ! scalar @mktables_list_output_files  # or if no outputs known
              || $old_start_time < $most_recent        # or out-of-date
              || $old_options ne $command_line_arguments; # or with different
                                                          # options

# Now we check to see if any output files are older than youngest, if
# they are, we need to continue on, otherwise we can presumably bail.
if (! $rebuild) {
    foreach my $out (@mktables_list_output_files) {
        if ( ! file_exists($out)) {
            print "'$out' is missing.\n" if $verbosity >= $VERBOSE;
            $rebuild = 1;
            last;
         }
        #local $to_trace = 1 if main::DEBUG;
        trace $most_recent, (stat $out)[9] if main::DEBUG && $to_trace;
        if ( (stat $out)[9] <= $most_recent ) {
            #trace "$out:  most recent mod time: ", (stat $out)[9], ", youngest: $most_recent\n" if main::DEBUG && $to_trace;
            print "'$out' is too old.\n" if $verbosity >= $VERBOSE;
            $rebuild = 1;
            last;
        }
    }
}
if (! $rebuild) {
    print "$0: Files seem to be ok, not bothering to rebuild.  Add '-w' option to force build\n";
    exit(0);
}
print "$0: Must rebuild tables.\n" if $verbosity >= $VERBOSE;

# Ready to do the major processing.  First create the perl pseudo-property.
$perl = Property->new('perl', Type => $NON_STRING, Perl_Extension => 1);

# Process each input file
foreach my $file (@input_file_objects) {
    $file->run;
}

# Finish the table generation.

print "Finishing processing Unicode properties\n" if $verbosity >= $PROGRESS;
finish_Unicode();

# For the very specialized case of comparing two Unicode versions...
if (DEBUG && $compare_versions) {
    handle_compare_versions();
}

print "Compiling Perl properties\n" if $verbosity >= $PROGRESS;
compile_perl();

print "Creating Perl synonyms\n" if $verbosity >= $PROGRESS;
add_perl_synonyms();

print "Writing tables\n" if $verbosity >= $PROGRESS;
write_all_tables();

# Write mktables.lst
if ( $file_list and $make_list ) {

    print "Updating '$file_list'\n" if $verbosity >= $PROGRESS;
    foreach my $file (@input_files, @files_actually_output) {
        my (undef, $directories, $basefile) = File::Spec->splitpath($file);
        my @directories = grep length, File::Spec->splitdir($directories);
        $file = join '/', @directories, $basefile;
    }

    my $ofh;
    if (! open $ofh,">",$file_list) {
        Carp::my_carp("Can't write to '$file_list'.  Skipping: $!");
        return
    }
    else {
        my $localtime = localtime $start_time;
        print $ofh <<"END";
#
# $file_list -- File list for $0.
#
#   Autogenerated starting on $start_time ($localtime)
#   From options $command_line_arguments
#
# - First section is input files
#   ($0 itself is not listed but is automatically considered an input)
# - Section separator is /^=+\$/
# - Second section is a list of output files.
# - Lines matching /^\\s*#/ are treated as comments
#   which along with blank lines are ignored.
#

# Input files:

END
        print $ofh "$_\n" for sort(@input_files);
        print $ofh "\n=================================\n# Output files:\n\n";
        print $ofh "$_\n" for sort @files_actually_output;
        print $ofh "\n# ",scalar(@input_files)," input files\n",
                "# ",scalar(@files_actually_output)+1," output files\n\n",
                "# End list\n";
        close $ofh
            or Carp::my_carp("Failed to close $ofh: $!");

        print "Filelist has ",scalar(@input_files)," input files and ",
            scalar(@files_actually_output)+1," output files\n"
            if $verbosity >= $VERBOSE;
    }
}

# Output these warnings unless -q explicitly specified.
if ($verbosity >= $NORMAL_VERBOSITY && ! $debug_skip) {
    if (@unhandled_properties) {
        print "\nProperties and tables that unexpectedly have no code points\n";
        foreach my $property (sort @unhandled_properties) {
            print $property, "\n";
        }
    }

    if (%potential_files) {
        print "\nInput files that are not considered:\n";
        foreach my $file (sort keys %potential_files) {
            print File::Spec->abs2rel($file), "\n";
        }
    }
    print "\nAll done\n" if $verbosity >= $VERBOSE;
}

if ($version_of_mk_invlist_bounds lt $v_version) {
    Carp::my_carp("WARNING: \\b{} algorithms (regen/mk_invlist.pl) need"
                . " to be checked and possibly updated to Unicode"
                . " $string_version.  Failing tests will be marked TODO");
}

exit(0);

# TRAILING CODE IS USED BY make_property_test_script()
__DATA__

use strict;
use warnings;

use feature 'signatures';

no warnings 'experimental::uniprop_wildcards';

# Test qr/\X/ and the \p{} regular expression constructs.  This file is
# constructed by mktables from the tables it generates, so if mktables is
# buggy, this won't necessarily catch those bugs.  Tests are generated for all
# feasible properties; a few aren't currently feasible; see
# is_code_point_usable() in mktables for details.

# Standard test packages are not used because this manipulates SIG_WARN.  It
# exits 0 if every non-skipped test succeeded; -1 if any failed.

my $Tests = 0;
my $Fails = 0;

# loc_tools.pl requires this function to be defined
sub ok($pass, @msg) {
    print "not " unless $pass;
    print "ok ";
    print ++$Tests;
    print " - ", join "", @msg if @msg;
    print "\n";
}

sub Expect($expected, $ord, $regex, $warning_type='') {
    my $line   = (caller)[2];

    # Convert the code point to hex form
    my $string = sprintf "\"\\x{%04X}\"", $ord;

    my @tests = "";

    # The first time through, use all warnings.  If the input should generate
    # a warning, add another time through with them turned off
    push @tests, "no warnings '$warning_type';" if $warning_type;

    foreach my $no_warnings (@tests) {

        # Store any warning messages instead of outputting them
        local $SIG{__WARN__} = $SIG{__WARN__};
        my $warning_message;
        $SIG{__WARN__} = sub { $warning_message = $_[0] };

        $Tests++;

        # A string eval is needed because of the 'no warnings'.
        # Assumes no parentheses in the regular expression
        my $result = eval "$no_warnings
                            my \$RegObj = qr($regex);
                            $string =~ \$RegObj ? 1 : 0";
        if (not defined $result) {
            print "not ok $Tests - couldn't compile /$regex/; line $line: $@\n";
            $Fails++;
        }
        elsif ($result ^ $expected) {
            print "not ok $Tests - expected $expected but got $result for $string =~ qr/$regex/; line $line\n";
            $Fails++;
        }
        elsif ($warning_message) {
            if (! $warning_type || ($warning_type && $no_warnings)) {
                print "not ok $Tests - for qr/$regex/ did not expect warning message '$warning_message'; line $line\n";
                $Fails++;
            }
            else {
                print "ok $Tests - expected and got a warning message for qr/$regex/; line $line\n";
            }
        }
        elsif ($warning_type && ! $no_warnings) {
            print "not ok $Tests - for qr/$regex/ expected a $warning_type warning message, but got none; line $line\n";
            $Fails++;
        }
        else {
            print "ok $Tests - got $result for $string =~ qr/$regex/; line $line\n";
        }
    }
    return;
}

sub Error($regex) {
    $Tests++;
    if (eval { 'x' =~ qr/$regex/; 1 }) {
        $Fails++;
        my $line = (caller)[2];
        print "not ok $Tests - re compiled ok, but expected error for qr/$regex/; line $line: $@\n";
    }
    else {
        my $line = (caller)[2];
        print "ok $Tests - got and expected error for qr/$regex/; line $line\n";
    }
    return;
}

# Break test files (e.g. GCBTest.txt) character that break allowed here
my $breakable_utf8 = my $breakable = chr(utf8::unicode_to_native(0xF7));
utf8::upgrade($breakable_utf8);

# Break test files (e.g. GCBTest.txt) character that indicates can't break
# here
my $nobreak_utf8 = my $nobreak = chr(utf8::unicode_to_native(0xD7));
utf8::upgrade($nobreak_utf8);

my $are_ctype_locales_available;
my $utf8_locale;
chdir 't' if -d 't';
eval { require "./loc_tools.pl" };
if (defined &locales_enabled) {
    $are_ctype_locales_available = locales_enabled('LC_CTYPE');
    if ($are_ctype_locales_available) {
        $utf8_locale = &find_utf8_ctype_locale;
    }
}

# Eval'd so can run on versions earlier than the property is available in
my $WB_Extend_or_Format_re = eval 'qr/[\p{WB=Extend}\p{WB=Format}\p{WB=ZWJ}]/';
if (! defined $WB_Extend_or_Format_re) {
    $WB_Extend_or_Format_re = eval 'qr/[\p{WB=Extend}\p{WB=Format}]/';
}

sub _test_break($template, $break_type) {
    # Test various break property matches.  The 2nd parameter gives the
    # property name.  The input is a line from auxiliary/*Test.txt for the
    # given property.  Each such line is a sequence of Unicode (not native)
    # code points given by their hex numbers, separated by the two characters
    # defined just before this subroutine that indicate that either there can
    # or cannot be a break between the adjacent code points.  All these are
    # tested.
    #
    # For the gcb property extra tests are made.  if there isn't a break, that
    # means the sequence forms an extended grapheme cluster, which means that
    # \X should match the whole thing.  If there is a break, \X should stop
    # there.  This is all converted by this routine into a match: $string =~
    # /(\X)/, Each \X should match the next cluster; and that is what is
    # checked.

    my $line   = (caller 1)[2];   # Line number
    my $comment = "";

    if ($template =~ / ( .*? ) \s* \# (.*) /x) {
        $template = $1;
        $comment = $2;

        # Replace leading spaces with a single one.
        $comment =~ s/ ^ \s* / # /x;
    }

    # The line contains characters above the ASCII range, but in Latin1.  It
    # may or may not be in utf8, and if it is, it may or may not know it.  So,
    # convert these characters to 8 bits.  If knows is in utf8, simply
    # downgrade.
    if (utf8::is_utf8($template)) {
        utf8::downgrade($template);
    } else {

        # Otherwise, if it is in utf8, but doesn't know it, the next lines
        # convert the two problematic characters to their 8-bit equivalents.
        # If it isn't in utf8, they don't harm anything.
        use bytes;
        $template =~ s/$nobreak_utf8/$nobreak/g;
        $template =~ s/$breakable_utf8/$breakable/g;
    }

    # Perl customizes wb.  So change the official tests accordingly
    if ($break_type eq 'wb' && $WB_Extend_or_Format_re) {

        # Split into elements that alternate between code point and
        # break/no-break
        my @line = split / +/, $template;

        # Look at each code point and its following one
        for (my $i = 1; $i <  @line - 1 - 1; $i+=2) {

            # The customization only involves changing some breaks to
            # non-breaks.
            next if $line[$i+1] =~ /$nobreak/;

            my $lhs = chr utf8::unicode_to_native(hex $line[$i]);
            my $rhs = chr utf8::unicode_to_native(hex $line[$i+2]);

            # And it only affects adjacent space characters.
            next if $lhs !~ /\s/u;

            # But, we want to make sure to test spaces followed by a Extend
            # or Format.
            next if $rhs !~ /\s|$WB_Extend_or_Format_re/;

            # To test the customization, add some white-space before this to
            # create a span.  The $lhs white space may or may not be bound to
            # that span, and also with the $rhs.  If the $rhs is a binding
            # character, the $lhs is bound to it and not to the span, unless
            # $lhs is vertical space.  In all other cases, the $lhs is bound
            # to the span.  If the $rhs is white space, it is bound to the
            # $lhs
            my $bound;
            my $span;
            if ($rhs =~ /$WB_Extend_or_Format_re/) {
                if ($lhs =~ /\v/) {
                    $bound = $breakable;
                    $span = $nobreak;
                }
                else {
                    $bound = $nobreak;
                    $span = $breakable;
                }
            }
            else {
                $span = $nobreak;
                $bound = $nobreak;
            }

            splice @line, $i, 0, ( '0020', $nobreak, '0020', $span);
            $i += 4;
            $line[$i+1] = $bound;
        }
        $template = join " ", @line;
    }

    # The input is just the break/no-break symbols and sequences of Unicode
    # code points as hex digits separated by spaces for legibility. e.g.:
    # ÷ 0020 × 0308 ÷ 0020 ÷
    # Convert to native \x format
    $template =~ s/ \s* ( [[:xdigit:]]+ ) \s* /sprintf("\\x{%02X}", utf8::unicode_to_native(hex $1))/gex;
    $template =~ s/ \s* //gx;   # Probably the line above removed all spaces;
                                # but be sure

    # Make a copy of the input with the symbols replaced by \b{} and \B{} as
    # appropriate
    my $break_pattern = $template =~ s/ $breakable /\\b{$break_type}/grx;
    $break_pattern =~ s/ $nobreak /\\B{$break_type}/gx;

    my $display_string = $template =~ s/[$breakable$nobreak]//gr;
    my $string = eval "\"$display_string\"";

    # The remaining massaging of the input is for the \X tests.  Get rid of
    # the leading and trailing breakables
    $template =~ s/^ \s* $breakable \s* //x;
    $template =~ s/ \s* $breakable \s* $ //x;

    # Delete no-breaks
    $template =~ s/ \s* $nobreak \s* //xg;

    # Split the input into segments that are breakable between them.
    my @should_display = split /\s*$breakable\s*/, $template;
    my @should_match = map { eval "\"$_\"" } @should_display;

    # If a string can be represented in both non-ut8 and utf8, test both cases
    my $display_upgrade = "";
    UPGRADE:
    for my $to_upgrade (0 .. 1) {

        if ($to_upgrade) {

            # If already in utf8, would just be a repeat
            next UPGRADE if utf8::is_utf8($string);

            utf8::upgrade($string);
            $display_upgrade = " (utf8-upgraded)";
        }

        my @modifiers = qw(a aa d u i);
        if ($are_ctype_locales_available) {
            push @modifiers, "l$utf8_locale" if defined $utf8_locale;

            # The /l modifier has C after it to indicate the locale to try
            push @modifiers, "lC";
        }

        # Test for each of the regex modifiers.
        for my $modifier (@modifiers) {
            my $display_locale = "";

            # For /l, set the locale to what it says to.
            if ($modifier =~ / ^ l (.*) /x) {
                my $locale = $1;
                $display_locale = "(locale = $locale)";
                POSIX::setlocale(POSIX::LC_CTYPE(), $locale);
                $modifier = 'l';
            }

            no warnings qw(locale regexp surrogate);
            my $pattern = "(?$modifier:$break_pattern)";

            # Actually do the test
            my $matched_text;
            my $matched = $string =~ qr/$pattern/;
            if ($matched) {
                $matched_text = "matched";
            }
            else {
                $matched_text = "failed to match";
                print "not ";

                if (TODO_FAILING_BREAKS) {
                    $comment = " # $comment" unless $comment =~ / ^ \s* \# /x;
                    $comment =~ s/#/# TODO/;
                }
            }
            print "ok ", ++$Tests, " - \"$display_string\" $matched_text /$pattern/$display_upgrade; line $line $display_locale$comment\n";

            # Only print the comment on the first use of this line
            $comment = "";

            # Repeat with the first \B{} in the pattern.  This makes sure the
            # code in regexec.c:find_byclass() for \B gets executed
            if ($pattern =~ / ( .*? : ) .* ( \\B\{ .* ) /x) {
                my $B_pattern = "$1$2";
                $matched = $string =~ qr/$B_pattern/;
                print "not " unless $matched;
                $matched_text = ($matched) ? "matched" : "failed to match";
                print "ok ", ++$Tests, " - \"$display_string\" $matched_text /$B_pattern/$display_upgrade; line $line $display_locale";
                print " # TODO" if TODO_FAILING_BREAKS && ! $matched;
                print "\n";
            }
        }

        next if $break_type ne 'gcb';

        # Finally, do the \X match.
        my @matches = $string =~ /(\X)/g;

        # Look through each matched cluster to verify that it matches what we
        # expect.
        my $min = (@matches < @should_match) ? @matches : @should_match;
        for my $i (0 .. $min - 1) {
            $Tests++;
            if ($matches[$i] eq $should_match[$i]) {
                print "ok $Tests - ";
                if ($i == 0) {
                    print "In \"$display_string\" =~ /(\\X)/g, \\X #1";
                } else {
                    print "And \\X #", $i + 1,
                }
                print " correctly matched $should_display[$i]; line $line\n";
            } else {
                $matches[$i] = join("", map { sprintf "\\x{%04X}", ord $_ }
                                                    split "", $matches[$i]);
                print "not ok $Tests -";
                print " # TODO" if TODO_FAILING_BREAKS;
                print " In \"$display_string\" =~ /(\\X)/g, \\X #",
                    $i + 1,
                    " should have matched $should_display[$i]",
                    " but instead matched $matches[$i]",
                    ".  Abandoning rest of line $line\n";
                next UPGRADE;
            }
        }

        # And the number of matches should equal the number of expected matches.
        $Tests++;
        if (@matches == @should_match) {
            print "ok $Tests - Nothing was left over; line $line\n";
        } else {
            print "not ok $Tests - There were ", scalar @should_match, " \\X matches expected, but got ", scalar @matches, " instead; line $line";
            print " # TODO" if TODO_FAILING_BREAKS;
            print "\n";
        }
    }

    return;
}

sub Test_GCB($t) {
    _test_break($t, 'gcb');
}

sub Test_LB($t) {
    _test_break($t, 'lb');
}

sub Test_SB($t) {
    _test_break($t, 'sb');
}

sub Test_WB($t) {
    _test_break($t, 'wb');
}

sub Finished() {
    print "1..$Tests\n";
    exit($Fails ? -1 : 0);
}