1diff --git a/source/data/brkitr/ja.txt b/source/data/brkitr/ja.txt 2index 363a74f6..7ae4b8f6 100644 3--- a/source/data/brkitr/ja.txt 4+++ b/source/data/brkitr/ja.txt 5@@ -11,6 +11,7 @@ ja{ 6 line_phrase:process(dependency){"line_phrase_cj.brk"} 7 line_strict:process(dependency){"line.brk"} 8 line_strict_phrase:process(dependency){"line_phrase_cj.brk"} 9+ word:process(dependency){"word_ja.brk"} 10 } 11 extensions{ 12 } 13diff --git a/source/data/brkitr/rules/word.txt b/source/data/brkitr/rules/word.txt 14index b4603823..58c18cbc 100644 15--- a/source/data/brkitr/rules/word.txt 16+++ b/source/data/brkitr/rules/word.txt 17@@ -84,13 +84,11 @@ $Control = [\p{Grapheme_Cluster_Break = Control}]; 18 $HangulSyllable = [\uac00-\ud7a3]; 19 $ComplexContext = [:LineBreak = Complex_Context:]; 20 $KanaKanji = [$Han $Hiragana $Katakana]; 21-$dictionaryCJK = [$KanaKanji $HangulSyllable]; 22-$dictionary = [$ComplexContext $dictionaryCJK]; 23+$dictionary = [$ComplexContext]; 24 25 # TODO: check if handling of katakana in dictionary makes rules incorrect/void 26 27-# leave CJK scripts out of ALetterPlus 28-$ALetterPlus = [$ALetter-$dictionaryCJK [$ComplexContext-$Extend-$Control]]; 29+$ALetterPlus = [$ALetter [$ComplexContext-$Extend-$Control]]; 30 31 32 ## ------------------------------------------------- 33@@ -185,10 +183,6 @@ $ExtendNumLet $ExFm* $Katakana {400}; # (13b) 34 # 35 ^$Regional_Indicator $ExFm* $Regional_Indicator; 36 37-# special handling for CJK characters: chain for later dictionary segmentation 38-$HangulSyllable $HangulSyllable {200}; 39-$KanaKanji $KanaKanji {400}; # different rule status if both kana and kanji found 40- 41 # Rule 999 42 # Match a single code point if no other rule applies. 43 .; 44