ICU-551.51.tar.gz

[apple/icu.git] / icuSources / data / brkitr / line_ja.txt

diff --git a/icuSources/data/brkitr/line_ja.txt b/icuSources/data/brkitr/line_ja.txt
index 2d4781b04aa95874df0d5b4090ac28acbfb954e7..205d75aabaa6a27515a7b89d00b5f5e930e7b43a 100644 (file)
--- a/icuSources/data/brkitr/line_ja.txt
+++ b/icuSources/data/brkitr/line_ja.txt
@@ -1,4 +1,4 @@
-# Copyright (c) 2002-2012  International Business Machines Corporation and
+# Copyright (c) 2002-2015  International Business Machines Corporation and
 # others. All Rights Reserved.
 #
 #  file:  line_ja.txt
@@ -97,6 +97,15 @@ $WJ = [:LineBreak =  Word_Joiner:];
 $XX = [:LineBreak =  Unknown:];
 $ZW = [:LineBreak =  ZWSpace:];
 
+# Special character classes for people & body part emoji:
+# Subsets of $CM:
+$ZWJ = \u200D;
+$EmojiVar = \uFE0F;
+# The following are subsets of $ID
+$EmojiForSeqs = [\u2764 \U0001F441 \U0001F466-\U0001F469 \U0001F48B \U0001F5E8];
+$EmojiForMods = [\u261D \u26F9 \u270A-\u270D \U0001F385 \U0001F3C3-\U0001F3C4 \U0001F3C7 \U0001F3CA-\U0001F3CB \U0001F442-\U0001F443 \U0001F446-\U0001F450 \U0001F466-\U0001F469 \U0001F46E-\U0001F478 \U0001F47C \U0001F481-\U0001F483 \U0001F485-\U0001F487 \U0001F4AA \U0001F590 \U0001F595 \U0001F596 \U0001F645-\U0001F647 \U0001F64B-\U0001F64F \U0001F6A3 \U0001F6B4-\U0001F6B6 \U0001F6C0 \U0001F918];
+$EmojiMods = [\U0001F3FB-\U0001F3FF];
+
 #   Dictionary character set, for triggering language-based break engines. Currently
 #   limited to LineBreak=Complex_Context. Note that this set only works in Unicode
 #   5.0 or later as the definition of Complex_Context was corrected to include all
@@ -231,6 +240,10 @@ $LB8Breaks    = [$LB4Breaks $ZW];
 $LB8NonBreaks = [[$LB4NonBreaks] - [$ZW]];
 
 
+# Special forward rule for people & body part emoji:
+# don't break $ZWJ from subsequent $EmojiForSeqs
+$EmojiForSeqs $EmojiVar? $EmojiMods? $ZWJ $EmojiForSeqs;
+
 # LB 9     Combining marks.      X   $CM needs to behave like X, where X is not $SP, $BK $CR $LF $NL 
 #                                $CM not covered by the above needs to behave like $AL   
 #                                See definition of $CAN_CM.
@@ -342,6 +355,10 @@ $BBcm $LB20NonBreaks $CM*;
 #  
 $HLcm ($HYcm | $BAcm) [^$CB]?;
 
+# LB 21b (forward) Don't break between SY and HL
+# (break between HL and SY already disallowed by LB 13 above)
+$SYcm $HLcm;
+
 # LB 22
 ($ALcm | $HLcm) $INcm;
 $CM+     $INcm;     #  by rule 10, any otherwise unattached CM behaves as AL
@@ -398,6 +415,10 @@ $CPcm ($ALcm | $HLcm | $NUcm);
 # LB 30a  Do not break between regional indicators.
 $RIcm $RIcm;
 
+# Special forward rule for people & body part emoji:
+# don't break between relevant emoji and $EmojiMods
+$EmojiForMods $EmojiVar? $EmojiMods;
+
 #
 #  Reverse Rules.
 #
@@ -480,6 +501,10 @@ $LF $CR;
 #           Requires an engine enhancement.
 #   / $SP* $ZW
 
+# Special reverse rule for people & body part emoji:
+# don't break $ZWJ from subsequent $EmojiForSeqs
+$EmojiForSeqs $ZWJ $EmojiMods? $EmojiVar? $EmojiForSeqs;
+
 # LB 9,10  Combining marks.
 #    X   $CM needs to behave like X, where X is not $SP or controls.
 #    $CM not covered by the above needs to behave like $AL
@@ -575,6 +600,8 @@ $CM* [$LB20NonBreaks-$CM] $CM* $BB;                   #  BB x .
 # LB21a
 [^$CB] $CM* ($HY | $BA) $CM* $HL;
 
+# LB21b (reverse)
+$CM* $HL $CM* $SY;
 
 # LB 22
 $CM* $IN $CM* ($ALPlus | $HL);
@@ -620,6 +647,10 @@ $CM* ($ALPlus | $HL | $NU) $CM* $CP;
 # LB 30a
 $CM* $RI $CM* $RI;
 
+# Special reverse rule for people & body part emoji:
+# don't break between relevant emoji and $EmojiMods
+$EmojiMods $EmojiVar? $EmojiForMods;
+
 ## -------------------------------------------------
 
 !!safe_reverse;