v0.9.0

[kanru]

What's New in libchewing 0.9.0 (August 11, 2024)

• Features

  • Support two new input modes:
    • Fuzzy mode - support prefix search, partial zhuyin search, toneless input.
    • Simple mode - like the traditional input method on Windows 95 which does
      not perform any intelligent phrasing.
  • Allow space key for pagination regardless of space_is_select_key mode.
  • Merge DOWN and SPACE key behavior in pagination.
  • A new option to disable fullwidth toggle key.
  • Workman layout support.
  • Load embedded mini dictionary if system dictionaries were not found.
  • Automatically load extra dictionaries found in search path
    • ${CHEWING_PATH}/dictionary.d/*.dat
  • Automatic snapshot selections when the curser is moved.

• Developer Features

  • BREAKING: Remove language_mode and character_form methods (Rust).
  • BREAKING: Remove unused pinyin.tab file
  • BREAKING: Remove most unused C code but keep compat C headers.
  • Bump minimum supported Rust version to 1.77
  • Enable ELF symbol versioning. All existing symbols are marked as version
    CHEWING_0.5 and new symbols from this release are marked as CHEWING_0.9.
  • A new set of configuration API:
    • chewing_config_has_option
    • chewing_config_get_int
    • chewing_config_set_int
    • chewing_config_get_str
    • chewing_config_set_str
  • New API to get runtime library version:
    • chewing_version
    • chewing_version_major
    • chewing_version_minor
    • chewing_version_patch
    • chewing_version_extra
  • New API to acknowledge output buffers:
    • chewing_ack
  • Define version macros
    • CHEWING_VERSION_MAJOR
    • CHEWING_VERSION_MINOR
    • CHEWING_VERSION_PATCH

• Bug Fixes

  • Separate special handling by pinyin variants (無法打「ㄐ」 #298)
  • Panic when selecting phrases backwards at the end of buffer. (introduced in
    v0.8.0)

• Dictionary

  • Default dictionary updates and fixes (工作上遇到的一些常用錯字。「獨占」「身分」「香菸」等等 #331)

• Performance

  • Optimize K shortest path algorithm to handle long pre-edit buffer. This
    allowed us to remove the special DP algorithm and keep only the graph search
    based algorithm in conversion.

• Testing

  • Add benchmarks for the conversion module
  • Add tests for options and default values
  • Add test for chewing_config_set_str
  • Test chewing_config_set_str with chewing.keyboard_type
  • Allow switching keyboard layout in genkeystroke
  • Add test for simple engine symbol input
  • Support new editor options in the fuzzer.
  • Generate test cases from fuzzer input.

• Miscellaneous Tasks

  • Drop outdated ChangeLog
  • Add option to turn off tests run
  • Use system corrosion if available
  • Regenerate chewing.h using latest cbindgen
  • Include CTest and use the built-in BUILD_TESTING option
  • Manage SQLite dependency with CMake
  • Print status about SQLite3 module
  • Remove unused test files
  • Bump Corrosion to v0.5
  • Fix building without using CMake presets
  • Force linking chewing_version obj file
  • Append git describe to version string for unreleased version.

• New Contributors

  • @chenrui333 made their first contribution in chore(build): add option to turn off tests run #560
  • @wusyong made their first contribution in Update documentation in bopomofo module #582

Full Changelog: v0.8.5...v0.9.0

---

This discussion was created from the release v0.9.0.

[yan12125]

FWIW, 我先放了 0.9.0 到 Arch Linux testing repo https://archlinux.org/packages/extra-testing/x86_64/libchewing/. 0.9 RC版我用了一陣子感覺似乎比較常選到不常用的字，但是一直找不到 reproduce 的方法，所以這次預計會多放在 testing 幾天看看情況。

[kanru]

看看有沒有開到 Fuzzy matching

[yan12125]

我最近沒有改設定，我用的 fcitx5-chewing 應該也還沒支援 enable fuzzy matching。另外，我遇到的案例都是同音但比較少用的詞(聲母、韻母、聲調都相同)。