v0.10.1

Fixes a bug that caused Dactyl to fail on Python 3.5.

v0.10.0

Dactyl v0.10.0 is a major refactor of Dactyl that includes minor breaking changes and some new functionality:

• Breaking: New header ID formula
• Breaking: xrefs errors
• Frontmatter improvements
• New hierarchy features
• Other new automatic fields
• Path filenames
• OpenAPI parsing improvements
• Multiline callouts
• Other bug fixes and improvements

See releasenotes.md for details on each of these.

Dactyl v0.9.0

This version adds two experimental new features:

• Parsing of Jekyll-style frontmatter. Unfortunately, more refactoring is needed before you can effectively use Jekyll-style frontmatter in Dactyl pages, but for the time being, Dactyl at least strips out the frontmatter before parsing the Markdown document.
• OpenAPI spec parsing. You can provide an OpenAPI v3.0 specification and Dactyl will generate documentation directly from the spec. For more information, see the updated README.

v0.8.4

This release fixes a bug with the link checker similar to v0.8.3, but
applying to images instead of normal links. Now the link checker will
properly skip over images with absolute src paths such as
<img src="/path/to/something">

v0.8.3

This release fixes a bug with the link checker where it would try (and fail) to check links with absolute paths if those links had anchors in them. It now correctly skips links with absolute paths regardless of whether they have anchors.

v0.8.2

This release fixes a file-decoding bug on Windows. It also incorporates the changes from (unreleased) v0.8.1, which quiets CSS warnings from Prince.

v0.8.0

v0.8.0 updates the requirements for Dactyl, fixes some bugs, and adds some small features.

Updated Dependencies

• Python v3.5 or higher is now required.
• Changes to ruamel.yaml from from PyYAML since PyYAML is not much maintained.
• Updates to the latest versions of other packages.

Breaking Changes:

• Dactyl now raises an error when it reads a YAML config file with any duplicate keys. You can use --bypass_errors to skip past these errors and use the first instance of each key only. (Previously, Dactyl used only the last instance of any duplicate keys.)
• YAML Config files are now loaded in "safe" mode.

New and updated features:

• Adds the option to use "strict" undefined fields in templates or the preprocessor. This raises an error if you refer to a field that's undefined. (By default, Dactyl ignores keys that are not defined in both preprocessor and template syntax.)
  Use --template_strict_undefined on the commandline or template_allow_undefined = false in the config file to enable strict undefined for templates. Use --pp_strict_undefined on the commandline or preprocessor_allow_undefined = false in the config file to enable strict undefined for the preprocessor.
• Adds two new custom tests to the templating syntax: defined_and_equalto and undefined_or_ne. These tests combine existing Jinja tests for to make it easier to use selectattr() on the pages array when using strict undefined is enabled.

Bug Fixes:

• Fixes namespace overlapping when loading more than one custom filter at a time. This previously could cause some custom filters' effects to run twice or not at all.
• Duplicate keys in YAML files are no longer ignored.

Dactyl v0.7.3

This minor release fixes the link substitution filter's behavior in Markdown mode, and more gracefully handles config files that list missing files without explicit page titles.

Dactyl v0.7.2

This patch release provides more information in the error message when template/preprocessor parsing encounters an error.

Dactyl v0.7.1

This minor upgrade improves the flexibility of the ElasticSearch uploader introduced in Dactyl v0.7.0 (the --es_upload command).

This version supports new config options, es_index_fields and es_index_separator which are analogous to pdf_filename_fields and pdf_filename_separator but for generating the names of indexes where Dactyl uploads things to ElasticSearch. Previously, Dactyl used the same formula for both PDF filenames and ElasticSearch index names.

Caution: If you are using the ElasticSearch uploader and non-default values for pdf_filename_fields or pdf_filename_separator, Dactyl v0.7.1 uploads files to different ES indexes than Dactyl v0.7.0 unless you also define the es_index_fields and es_index_separator config fields to match.

There are no other changes to Dactyl functionality so there's no need to upgrade from 0.7.0 if you're not using Dactyl's ElasticSearch uploader.