Releases: elixir-unicode/unicode_unihan
Release list
Unicode Unihan version 0.5.0
Breaking Changes
-
kTotalStrokesis now a single integer rather than a%{Hans: n, Hant: n}map, matching the single-valued property in Unicode 16.0 and later. -
Decoders for the properties Unicode has withdrawn (kFrequency, kGB7, kHKSCS, kIRGDaeJaweon, kIRGDaiKanwaZiten, kIRGKangXi, kJa, kKPS0, kKPS1, kKSC0, kKSC1, kRSKangXi) are removed along with their data.
-
The
:syntaxof each property definition is the regex source string rather than a compiledRegex, since a compiled pattern is tied to the OTP release that built it.
Enhancements
-
Updated to Unicode 18.0 data (103,000 code points), adding kFanqie, kZhuang, kTayNumeric, kJapaneseNewVariant and kJapaneseOldVariant.
-
kRSUnicodeentries carry a:scriptkey (:Hant,:Hans,:Hanjor:Hanv) identifying the radical variant, andUnicode.Unihan.Radicalgains the:Hanvvariant introduced in Unicode 18. -
kMorohashidecodes supplemental-volume (Hnnn) indices and variation selectors, andkStrangegains the:symmetriccategory. -
Structured decoding for kMojiJoho, kSMSZD2003Index, kSMSZD2003Readings, kZhuang and the new numeric properties.
-
Documentation is generated with the markdown formatter as well as HTML.
Unicode Unihan version 0.4.0
Enhancements
-
Updated to Unicode 15.1
-
Property information is autogenerated from TR38 UNICODE HAN DATABASE (UNIHAN.
-
Diagnostic output uses
Logger.info/2rather thanIO.puts/1
Unicode Unihan version 0.3.0
Bug Fixes
-
Ensure all data paths are resolved at runtime, not compile time.
-
Move data/ to priv/ since only priv/ is preserved in releases.
Unicode Unihan version 0.2.0
Enhancements
- Improves compilation time by orders of magnitude by switching to use
:persistent_termfor storing the Unihan database.