Skip to content

Latest commit

 

History

History
82 lines (73 loc) · 4.04 KB

CHANGELOG.md

File metadata and controls

82 lines (73 loc) · 4.04 KB

Changelog

Only the most important stuff is listed here.
See commit history between tags for details.

v0.8.0 (TBD)

  • Added scripts and script extensions support (UAX #24: Unicode Script Property)
  • Added cursor support for text segmentation (doc/CURSOR.md)
  • Optimized UTF conversion of ASCII strings
  • Changed versions in namespace una::version to classes (#18)
  • File uni_algo/version.h is not included by other files anymore
  • Improved una::error class added una::error::code
  • Renamed search functions to find (#23)
  • Renamed una::search class to una::found (#23)
  • Renamed config defines:
    • UNI_ALGO_DISABLE_SHRINK_TO_FIT to UNI_ALGO_NO_SHRINK_TO_FIT
    • UNI_ALGO_DISABLE_BREAK_GRAPHEME to UNI_ALGO_DISABLE_SEGMENT_GRAPHEME
    • UNI_ALGO_DISABLE_BREAK_WORD to UNI_ALGO_DISABLE_SEGMENT_WORD

v0.7.1 (2023-02-27)

  • Fixed bug when normalization and UTF-16 views are used with input ranges (#22). Thanks @tocic.

v0.7.0 (2023-02-07)

  • Renamed main namespace from uni:: to una::
  • Splitted .h files with ranges and functions:
    • Added files: uni_algo/ranges_norm.h and uni_algo/ranges_conv.h
    • Renamed file uni_algo/break_grapheme.h to uni_algo/ranges_grapheme.h
    • Renamed file uni_algo/break_word.h to uni_algo/ranges_word.h
  • The library is full constexpr now (requeries C++20 or higher and header-only enabled)
  • Added safe layer (bounds checks enabled in debug disabled in release)
  • In C++23 std::string::resize_and_overwrite is now used for better performance
  • Added ASCII extension
  • Removed function uni::codepoint::to_decompose_hangul_u32

v0.6.0 (2022-11-07)

  • Added uni::is_valid_utf8/16/32 functions for strings
  • Added custom allocator support for template functions
  • Added custom allocator support for output ranges
  • Added header-only support (CMake option UNI_ALGO_HEADER_ONLY=ON)
  • Added shared library support (CMake option BUILD_SHARED_LIBS=ON)
  • Renamed file uni_algo/convert.h to uni_algo/conv.h
  • Lowered requirements for word/grapheme breaks to bidirectional range
  • Extensions now use unx namespace instead of uni

v0.5.0 (2022-10-01)

  • Updated to Unicode 15.0.0
  • Added single include version
  • Renamed files from cpp_uni_* to uni_algo/*
  • Renamed most functions, utf8/utf16 now at the end of a function
  • Added better locale support
  • Added word properties
  • Added reverse grapheme/word algorithms
  • Added alternative syntax for code point general category
  • Removed function uni::codepoint::is_same_category
  • Redesigned transliterators more than two times faster now

v0.4.0 (2022-08-30)

  • Added code point basic properties
  • Added code point case properties
  • Added code point case tranformations
  • Added code point normalization properties
  • Added code point normalization tranformations
  • Added ranges to_utf8_reserve and to_utf16_reserve

v0.3.0 (2022-08-07)

  • Added support for char8_t strings
  • Switched from iterators to ranges
  • Switched to ranges in transliterators
  • Added utf8, utf16, basic views
  • Added to_utf8, to_utf16 ranges
  • Added normalization views
  • Added break: grapheme, word, word_only views
  • Added internal translit view
  • Removed utf8, utf16, output, translit, func iterators
  • Removed normalization iterators
  • Removed break: grapheme, word iterators
  • Renamed function is_pos() to has_pos() in class error

v0.2.0 (2022-07-14)

  • Updated to Unicode 14.0.0

v0.1.0 (2022-07-07)

  • Initial commit