xref: /aosp_15_r20/external/cronet/third_party/icu/cast/brkitr.patch (revision 6777b5387eb2ff775bb5750e3f5d96f37fb7352b)

diff --git a/source/data/brkitr/ja.txt b/source/data/brkitr/ja.txt
index 363a74f6..7ae4b8f6 100644
--- a/source/data/brkitr/ja.txt
+++ b/source/data/brkitr/ja.txt
@@ -11,6 +11,7 @@ ja{
         line_phrase:process(dependency){"line_phrase_cj.brk"}
         line_strict:process(dependency){"line.brk"}
         line_strict_phrase:process(dependency){"line_phrase_cj.brk"}
+	word:process(dependency){"word_ja.brk"}
     }
     extensions{
     }
diff --git a/source/data/brkitr/rules/word.txt b/source/data/brkitr/rules/word.txt
index b4603823..58c18cbc 100644
--- a/source/data/brkitr/rules/word.txt
+++ b/source/data/brkitr/rules/word.txt
@@ -84,13 +84,11 @@ $Control        = [\p{Grapheme_Cluster_Break = Control}];
 $HangulSyllable = [\uac00-\ud7a3];
 $ComplexContext = [:LineBreak = Complex_Context:];
 $KanaKanji      = [$Han $Hiragana $Katakana];
-$dictionaryCJK  = [$KanaKanji $HangulSyllable];
-$dictionary     = [$ComplexContext $dictionaryCJK];
+$dictionary     = [$ComplexContext];

 # TODO: check if handling of katakana in dictionary makes rules incorrect/void

-# leave CJK scripts out of ALetterPlus
-$ALetterPlus  = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]];
+$ALetterPlus  = [$ALetter [$ComplexContext-$Extend-$Control]];


 ## -------------------------------------------------
@@ -185,10 +183,6 @@ $ExtendNumLet  $ExFm* $Katakana     {400};    #  (13b)
 #
 ^$Regional_Indicator $ExFm* $Regional_Indicator;

-# special handling for CJK characters: chain for later dictionary segmentation
-$HangulSyllable $HangulSyllable {200};
-$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found
-
 # Rule 999
 #     Match a single code point if no other rule applies.
 .;