Init commit

Author: storm-ptr

.clang-format

@@ -0,0 +1,67 @@
+---
+Language:        Cpp
+# BasedOnStyle:  Chromium
+AccessModifierOffset: -4
+AlignAfterOpenBracket: Align
+AlignConsecutiveAssignments: false
+AlignEscapedNewlinesLeft: true
+AlignOperands:   true
+AlignTrailingComments: true
+AllowAllParametersOfDeclarationOnNextLine: false
+AllowShortBlocksOnASingleLine: false
+AllowShortCaseLabelsOnASingleLine: false
+AllowShortFunctionsOnASingleLine: Inline
+AllowShortIfStatementsOnASingleLine: false
+AllowShortLoopsOnASingleLine: false
+AlwaysBreakAfterDefinitionReturnType: None
+AlwaysBreakBeforeMultilineStrings: true
+AlwaysBreakTemplateDeclarations: true
+BinPackArguments: false
+BinPackParameters: false
+BreakBeforeBinaryOperators: None
+BreakBeforeBraces: Stroustrup
+BreakBeforeTernaryOperators: true
+BreakConstructorInitializersBeforeComma: true
+ColumnLimit:     80
+CommentPragmas:  '^ IWYU pragma:'
+ConstructorInitializerAllOnOneLineOrOnePerLine: true
+ConstructorInitializerIndentWidth: 4
+ContinuationIndentWidth: 4
+Cpp11BracedListStyle: true
+DerivePointerAlignment: false
+DisableFormat:   false
+ExperimentalAutoDetectBinPacking: false
+ForEachMacros:   [ foreach, Q_FOREACH, BOOST_FOREACH ]
+IndentCaseLabels: true
+IndentWidth:     4
+IndentWrappedFunctionNames: false
+KeepEmptyLinesAtTheStartOfBlocks: false
+MacroBlockBegin: '^IPC_END_MESSAGE_MAP$'
+MacroBlockEnd:   ''
+MaxEmptyLinesToKeep: 1
+NamespaceIndentation: None
+ObjCBlockIndentWidth: 4
+ObjCSpaceAfterProperty: false
+ObjCSpaceBeforeProtocolList: false
+PenaltyBreakBeforeFirstCallParameter: 1
+PenaltyBreakComment: 300
+PenaltyBreakFirstLessLess: 120
+PenaltyBreakString: 1000
+PenaltyExcessCharacter: 1000000
+PenaltyReturnTypeOnItsOwnLine: 200
+PointerAlignment: Left
+SpaceAfterCStyleCast: false
+SpaceBeforeAssignmentOperators: true
+SpaceBeforeParens: ControlStatements
+SpaceInEmptyParentheses: false
+SpacesBeforeTrailingComments: 2
+SpacesInAngles:  false
+SpacesInContainerLiterals: true
+SpacesInCStyleCastParentheses: false
+SpacesInParentheses: false
+SpacesInSquareBrackets: false
+Standard:        Auto
+TabWidth:        8
+UseTab:          Never
+...
+

detail/common.hpp

@@ -0,0 +1,44 @@
+// Andrew Naplavkov
+
+#ifndef STEP_COMMON_HPP
+#define STEP_COMMON_HPP
+
+#include <array>
+#include <utility>
+#include <vector>
+
+namespace step {
+
+struct make_pair {
+    template <typename Lhs, typename Rhs>
+    auto operator()(Lhs&& lhs, Rhs&& rhs)
+    {
+        return std::make_pair(std::forward<Lhs>(lhs), std::forward<Rhs>(rhs));
+    }
+};
+
+struct make_reverse_pair {
+    template <typename Lhs, typename Rhs>
+    auto operator()(Lhs&& lhs, Rhs&& rhs)
+    {
+        return std::make_pair(std::forward<Rhs>(rhs), std::forward<Lhs>(lhs));
+    }
+};
+
+template <typename T, size_t N>
+class ring_table {
+    std::array<std::vector<T>, N> rows_;
+
+public:
+    explicit ring_table(size_t cols)
+    {
+        for (auto& row : rows_)
+            row.resize(cols);
+    }
+
+    auto& operator[](size_t row) { return rows_[row % N]; }
+};
+
+}  // namespace step
+
+#endif  // STEP_COMMON_HPP

detail/hirschberg.hpp

@@ -0,0 +1,73 @@
+// Andrew Naplavkov
+
+#ifndef STEP_HIRSCHBERG_HPP
+#define STEP_HIRSCHBERG_HPP
+
+#include <algorithm>
+#include <iterator>
+#include <utility>
+
+namespace step {
+namespace hirschberg {
+namespace detail {
+
+template <typename RandomIt1,
+          typename RandomIt2,
+          typename DynamicProg,
+          typename BinaryOp>
+auto partition_point(RandomIt1 first1,
+                     RandomIt1 last1,
+                     RandomIt2 first2,
+                     RandomIt2 last2,
+                     DynamicProg dp,
+                     BinaryOp op)
+{
+    auto split1 = first1 + std::distance(first1, last1) / 2;
+    auto top = dp.make_last_row(first1, split1, first2, last2);
+    auto bottom = dp.make_last_row(std::make_reverse_iterator(last1),
+                                   std::make_reverse_iterator(split1),
+                                   std::make_reverse_iterator(last2),
+                                   std::make_reverse_iterator(first2));
+    std::transform(
+        top.begin(), top.end(), bottom.rbegin(), top.begin(), std::plus{});
+    auto split2 =
+        first2 + std::distance(top.begin(),
+                               std::min_element(top.begin(), top.end(), dp));
+    return op(split1, split2);
+}
+
+}  // namespace detail
+
+/// @see https://en.wikipedia.org/wiki/Hirschberg's_algorithm
+template <typename RandomIt1,
+          typename RandomIt2,
+          typename OutputIt,
+          typename DynamicProg>
+OutputIt align(RandomIt1 first1,
+               RandomIt1 last1,
+               RandomIt2 first2,
+               RandomIt2 last2,
+               OutputIt result,
+               DynamicProg dp)
+{
+    auto size1 = std::distance(first1, last1);
+    auto size2 = std::distance(first2, last2);
+    if (size1 < 2 || size2 < 2)
+        return dp.trivial_align(first1, last1, first2, last2, result);
+
+    auto[split1, split2] =
+        size2 < size1
+            ? detail::partition_point(
+                  first1, last1, first2, last2, dp, make_pair{})
+            : detail::partition_point(
+                  first2, last2, first1, last1, dp, make_reverse_pair{});
+
+    result = hirschberg::align(first1, split1, first2, split2, result, dp);
+    result = hirschberg::align(split1, last1, split2, last2, result, dp);
+    return result;
+}
+
+}  // namespace hirschberg
+}  // namespace step
+
+#endif  // STEP_HIRSCHBERG_HPP

edit_distance.hpp

@@ -0,0 +1,117 @@
+// Andrew Naplavkov
+
+#ifndef STEP_EDIT_DISTANCE_HPP
+#define STEP_EDIT_DISTANCE_HPP
+
+#include <algorithm>
+#include <optional>
+#include <step/detail/common.hpp>
+#include <step/detail/hirschberg.hpp>
+#include <utility>
+
+namespace step {
+namespace edit_distance {
+namespace detail {
+
+template <typename ForwardIt, typename T, typename OutputIt, typename BinaryOp>
+auto align(ForwardIt first,
+           ForwardIt last,
+           const T& val,
+           OutputIt result,
+           BinaryOp op)
+{
+    auto it = std::find(first, last, val);
+    if (it == last)
+        it = first;
+    return std::transform(first, last, result, [&](const auto& item) {
+        return op(item, item == *it ? std::optional{val} : std::nullopt);
+    });
+}
+
+struct dynamic_programming {
+    /// @see https://en.wikipedia.org/wiki/Wagner–Fischer_algorithm
+    template <typename RandomIt1, typename RandomIt2>
+    auto make_last_row(RandomIt1 first1,
+                       RandomIt1 last1,
+                       RandomIt2 first2,
+                       RandomIt2 last2) const
+    {
+        auto size1 = std::distance(first1, last1);
+        auto size2 = std::distance(first2, last2);
+        ring_table<size_t, 2> tbl(size2 + 1);
+        for (size_t l = 0; l <= size1; ++l)
+            for (size_t r = 0; r <= size2; ++r) {
+                if (l == 0)
+                    tbl[l][r] = r;
+                else if (r == 0)
+                    tbl[l][r] = l;
+                else if (first1[l - 1] == first2[r - 1])
+                    tbl[l][r] = tbl[l - 1][r - 1];
+                else
+                    tbl[l][r] = 1 + std::min({tbl[l][r - 1],        // insert
+                                              tbl[l - 1][r],        // remove
+                                              tbl[l - 1][r - 1]});  // replace
+            }
+        return std::move(tbl[size1]);
+    }
+
+    bool operator()(size_t lhs, size_t rhs) const { return lhs < rhs; }
+
+    template <typename RandomIt1, typename RandomIt2, typename OutputIt>
+    OutputIt trivial_align(RandomIt1 first1,
+                           RandomIt1 last1,
+                           RandomIt2 first2,
+                           RandomIt2 last2,
+                           OutputIt result) const
+    {
+        if (first1 == last1)
+            return std::transform(first2, last2, result, [&](const auto& item) {
+                return std::make_pair(std::nullopt, item);
+            });
+        else if (first2 == last2)
+            return std::transform(first1, last1, result, [&](const auto& item) {
+                return std::make_pair(item, std::nullopt);
+            });
+        else if (std::next(first1) == last1)
+            return align(first2, last2, *first1, result, make_reverse_pair{});
+        else  // std::next(first2) == last2
+            return align(first1, last1, *first2, result, make_pair{});
+    }
+};
+
+}  // namespace detail
+
+/**
+ * Find the optimal sequence alignment between two strings. Optimality is
+ * measured with the Levenshtein distance, defined to be the sum of the costs of
+ * insertions, replacements, deletions, and null actions needed to change one
+ * string into the other.
+ *
+ * Time complexity: O(N*M), space complexity O(min(N,M)), where:
+ * N = std::distance(first1, last1), M = std::distance(first2, last2).
+ */
+template <typename RandomIt1, typename RandomIt2, typename OutputIt>
+OutputIt align(RandomIt1 first1,
+               RandomIt1 last1,
+               RandomIt2 first2,
+               RandomIt2 last2,
+               OutputIt result)
+{
+    return hirschberg::align(
+        first1, last1, first2, last2, result, detail::dynamic_programming{});
+}
+
+template <typename RandomRng1, typename RandomRng2, typename OutputIt>
+OutputIt align(const RandomRng1& rng1, const RandomRng2& rng2, OutputIt result)
+{
+    return edit_distance::align(std::begin(rng1),
+                                std::end(rng1),
+                                std::begin(rng2),
+                                std::end(rng2),
+                                result);
+}
+
+}  // namespace edit_distance
+}  // namespace step
+
+#endif  // STEP_EDIT_DISTANCE_HPP

example/diff/main.cpp

@@ -0,0 +1,52 @@
+// Andrew Naplavkov
+
+#include <fstream>
+#include <iostream>
+#include <step/edit_distance.hpp>
+#include <string>
+#include <string_view>
+
+auto read_file(const char* file_name)
+{
+    using iterator_t = std::istreambuf_iterator<char>;
+    return std::string{(iterator_t(std::ifstream{file_name})), iterator_t()};
+}
+
+auto split(const std::string& str)
+{
+    std::vector<std::string_view> result;
+    auto first = str.c_str();
+    auto last = first + str.size();
+    while (first != last) {
+        auto it = std::find(first, last, '\n');
+        result.emplace_back(first, std::distance(first, it));
+        first = it == last ? it : std::next(it);
+    }
+    return result;
+}
+
+int main(int argc, char* argv[])
+{
+    using option_t = std::optional<std::string_view>;
+    auto file1 = read_file(argv[1]);
+    auto file2 = read_file(argv[2]);
+    auto rng1 = split(file1);
+    auto rng2 = split(file2);
+    std::vector<std::pair<option_t, option_t>> pairs;
+
+    step::edit_distance::align(rng1, rng2, std::back_inserter(pairs));
+
+    size_t row = 0;
+    for (const auto& pair : pairs) {
+        if (pair.second)
+            ++row;
+        if (pair.first != pair.second) {
+            if (pair.first)
+                std::cout << "--- " << row + (pair.second ? 0 : 1) << " ---\n"
+                          << pair.first.value() << "\n";
+            if (pair.second)
+                std::cout << "+++ " << row << " +++\n"
+                          << pair.second.value() << "\n";
+        }
+    }
+}

example/diff/makefile.windows

@@ -0,0 +1,16 @@
+CXXFLAGS=/EHsc /bigobj /std:c++17 /D_HAS_AUTO_PTR_ETC=1 $(CXXFLAGS)
+EXECUTABLE=diff.exe
+INCLUDE=..\..\;$(INCLUDE)
+OBJECTS=$(SOURCES:.cpp=.obj)
+SOURCES=main.cpp
+
+all: $(SOURCES) $(EXECUTABLE)
+
+$(EXECUTABLE): $(OBJECTS)
+	$(CXX) /Fe:$(EXECUTABLE) $(OBJECTS) $(LIBS)
+
+.cpp.obj:
+	$(CXX) -c $(CXXFLAGS) $<
+
+clean:
+	del *.obj *.exp *.lib $(EXECUTABLE)