Changelog for libime-devel-1.1.10-1.1.x86_64.rpm :

* Wed Jan 29 2025 Marguerite Su - update to 1.1.10
* New language model to fix some frequency issue
* Correction support for shuangpin
* Improving segmentation algorithm for pinyin- refresh libime-1.1.10-boost166.patch- drop libime-1.1.9-boost166.patch
* Sun Jan 05 2025 Marguerite Su - update to 1.1.9:
* update model and dict
* Manually adjust certain character weight in dict to avoid some meaningless score- add patch libime-1.1.9-boost166.patch
* was splitted from since boost 1.67
* Sun Sep 08 2024 Dirk Müller - update to 1.1.8:
* When parsing with correction flag, u,i,v should not be considered as invalid
* Fix possible prediction duplicate
* Refactor pinyin map generatation with meaningful enum
* Allow un -> vn for jqxy
* Wed May 22 2024 Marguerite Su - update to 1.1.7:
* Support more pinyin correction
* Improve candidates to cursor logic
* Remove shuang pin\'s ng -> mapping to pinyin ng
* Sun Mar 17 2024 Dirk Müller - update to 1.1.6:
* Fix TscanCode detections
* clang tidy and coverity clean ups
* Handle customized single char final mapping
* Sun Feb 04 2024 Marguerite Su - update to 1.1.5
* Replace PUA character with non-PUA unicode in the table data.- changes in 1.1.4
* Improve Pinyin prediction order
* Fix two shuangpin profile bugs
* Fix build on musl
* Fix for exact match mode
* libime_tabledict adds a new mode flag for extra dict.- changes in 1.1.3
* reduce user model weight and unigram weight.
* new pinyin prediction API to allow prediction from dictionaries.
* Add more pinyin correction for wrong order e.g. (hzi -> zhi)
* Fix exact match mode in table.
* Tue Oct 03 2023 Marguerite Su - update to 1.1.2
* Allow table to include words without code
* allow table ordering ignore the code length
* Allow value to be quoted
* Apply ZSTD compression the pinyin/table dict and user history.
* fix some performance issue for certain pinyin input
* Allow table rule to specify character from backwards.
* Sun Aug 27 2023 Marguerite Su - update to 1.1.1
* hotfix for user pinyin parsing
* Sun Aug 20 2023 Marguerite Su - update to 1.1.0
* new language model
* fix full pinyin shuangpin preedit to use the pinyin of first candidate
* fuzzy pinyin now allows jqxy with van
* libime_pinyindict now skips invalid lines when building dictionary file
* improve partial match flag for chaizi
* Thu Apr 27 2023 Dirk Müller - update to 1.0.17:
* Accept lue or nue when parsing text formatted dictionary
* Don\'t make partial sp like z/c/s match zh/ch/sh
* Set a search size hard limit on the nbest search
* use int64_t for scoped clock
* Thu Dec 08 2022 Dirk Müller - update to 1.0.16:
* Use our own hasher for std::string because boost stream may not use a fixed implementation for string hash
* Sun Nov 20 2022 Fusion Future - Update to 1.0.15
* Update translation
* Tue Sep 20 2022 Marguerite Su - update version 1.0.14
* Update language model data and dictionary
* Add support for libime_history to compile text to binary data- drop libime-1.0.13-return-0.0-for-other-platforms-in-usage.patch, upstreamed
* Mon Jul 25 2022 Takashi Iwai - Remove yet another stale file, table.tar.gz
* Sun Jul 24 2022 Fusion Future - Remove one unused file
* libime-1.0.13.tar.xz
* Tue Jul 05 2022 Fusion Future - Update to 1.0.13
* fix a crash in pinyin under certain rare cases- Use _dict tar which contains all the needed dictionaries- Fix build by adding return statement (gh#kpu/kenlm#391)
* libime-1.0.13-return-0.0-for-other-platforms-in-usage.patch- Remove unneeded patch
* libime-no-download.patch
* Tue Jan 25 2022 Marguerite Su - update version 1.0.11
* Partial shuangpin support. It\'s a new option