locales33a: #i111152# treat Indic aksaras with virama as one grapheme cluster for character iterator; patch from <cmc>

1 files changed, 86 insertions, 26 deletions

diff --git a/i18npool/source/breakiterator/data/char_in.txt b/i18npool/source/breakiterator/data/char_in.txt
index 72c4a44720cd..5e1ed67596c0 100644
--- a/i18npool/source/breakiterator/data/char_in.txt
+++ b/i18npool/source/breakiterator/data/char_in.txt
@@ -1,48 +1,108 @@
 #
-#   Copyright (C) 2002-2003, International Business Machines Corporation and others.
+#   Copyright (C) 2002-2009, International Business Machines Corporation and others.
 #       All Rights Reserved.
 #
 #   file:  char.txt 
 #
 #   ICU Character Break Rules, also known as Grapheme Cluster Boundaries
 #      See Unicode Standard Annex #29.
-#      These rules are based on TR29 Version 4.0.0
+#      These rules are based on TR29 Revision 13, for Unicode Version 5.1
 #
 
 #
 #  Character Class Definitions.
-#    The names are those from TR29.
 #
-$CR = \r;
-$LF = \n;
-$Control    = [[:Zl:] [:Zp:] [:Cc:] [:Cf:]];
-
-# add Japanese Half Width voicing marks to $Extend
-$VoiceMarks = [\uff9e\uff9f];
-$cmcextend = [ \u0903 \u093e-\u0940 \u0949-\u094C \u09bf-\u09c0 \u09c7-\u09c8 \u09cb-\u09cc \u0bc1-\u0bc2 \u0bc6-\u0bc8 \u0bca-\u0bcc \u0c01-\u0c03 \u0c41-\u0c44];
-$Extend     = [[:Grapheme_Extend = TRUE:] $VoiceMarks $cmcextend]; 
+$CR          = [\p{Grapheme_Cluster_Break = CR}];
+$LF          = [\p{Grapheme_Cluster_Break = LF}];
+$Control     = [\p{Grapheme_Cluster_Break = Control}];
+$Prepend     = [\p{Grapheme_Cluster_Break = Prepend}];
+$Extend      = [\p{Grapheme_Cluster_Break = Extend}];
+$SpacingMark = [\p{Grapheme_Cluster_Break = SpacingMark}];
+$BengaliLetter = [\u0985-\u09B9 \u09CE \u09DC-\u09E1 \u09F0-\u09F1];
+$BengaliSignVirama = \u09CD;
+$GujaratiLetter = [\u0A85-\u0A8C \u0A8F-\u0A90 \u0A93-\u0AB9 \u0AE0-\u0AE1];
+$GujaratiSignVirama = \u0ACD;
+$DevanagariLetter = [\u0904-\u0939 \u0958-\u0961 \u0972-\u097F];
+$DevanagariSignVirama = \u094D;
+$KannadaLetter = [\u0C85-\u0CB9 \u0CDE-\u0CE1];
+$KannadaSignVirama = \u0CCD;
+$MalayalamLetter = [\u0D05-\u0D39 \u0D60-\u0D61 \u0D7A-\u0D7F];
+$MalayalamSignVirama = \u0D4D;
+$OriyaLetter = [\u0B05-\u0B39 \u0B5C-\u0B61 \u0B71];
+$OriyaSignVirama = \u0B4D;
+$GurmukhiLetter = [\u0A05-\u0A39 \u0A59-\u0A5E];
+$GurmukhiSignVirama = \u0A4D;
+$TamilLetter = [\u0B85-\u0BB9];
+$TamilSignVirama = \u0BCD;
+$TeluguLetter = [\u0C05-\u0C39 \u0C58-\u0C61];
+$TeluguSignVirama = \u0C4D;
 
 #
 # Korean Syllable Definitions
 #
-$L   = [:Hangul_Syllable_Type = L:];
-$V   = [:Hangul_Syllable_Type = V:];
-$T   = [:Hangul_Syllable_Type = T:];
+$L       = [\p{Grapheme_Cluster_Break = L}];
+$V       = [\p{Grapheme_Cluster_Break = V}];
+$T       = [\p{Grapheme_Cluster_Break = T}];
 
-$LV  = [:Hangul_Syllable_Type = LV:];
-$LVT = [:Hangul_Syllable_Type = LVT:];
+$LV      = [\p{Grapheme_Cluster_Break = LV}];
+$LVT     = [\p{Grapheme_Cluster_Break = LVT}];
 
-$HangulSyllable = $L+ | ($L* ($LV? $V+ | $LV | $LVT) $T*) | $T+;
 
-#
-#  Forward Break Rules
-#
+## -------------------------------------------------
+!!chain;
+
+!!forward;
+
 $CR $LF;
-([^$Control] | $HangulSyllable) $Extend*;
-.;
 
+$BengaliLetter ($BengaliSignVirama $BengaliLetter?)+;
+$GujaratiLetter ($GujaratiSignVirama $GujaratiLetter?)+;
+$DevanagariLetter ($DevanagariSignVirama $DevanagariLetter?)+;
+$KannadaLetter ($KannadaSignVirama $KannadaLetter?)+;
+$MalayalamLetter ($MalayalamSignVirama $MalayalamLetter?)+;
+$OriyaLetter ($OriyaSignVirama $OriyaLetter?)+;
+$GurmukhiLetter ($GurmukhiSignVirama $GurmukhiLetter?)+;
+$TamilLetter ($TamilSignVirama $TamilLetter?)+;
+$TeluguLetter ($TeluguSignVirama $TeluguLetter?)+;
+
+$L ($L | $V | $LV | $LVT);
+($LV | $V) ($V | $T);
+($LVT | $T) $T;
+
+[^$Control $CR $LF] $Extend;
+
+[^$Control $CR $LF] $SpacingMark;
+$Prepend [^$Control $CR $LF];
+
+
+## -------------------------------------------------
+
+!!reverse;
+$LF $CR;
+($BengaliLetter? $BengaliSignVirama)+ $BengaliLetter;
+($GujaratiLetter? $GujaratiSignVirama)+ $GujaratiLetter;
+($DevanagariLetter? $DevanagariSignVirama)+ $DevanagariLetter;
+($KannadaLetter? $KannadaSignVirama)+ $KannadaLetter;
+($MalayalamLetter? $MalayalamSignVirama)+ $MalayalamLetter;
+($OriyaLetter? $OriyaSignVirama)+ $OriyaLetter;
+($GurmukhiLetter? $GurmukhiSignVirama)+ $GurmukhiLetter;
+($TamilLetter? $TamilSignVirama)+ $TamilLetter;
+($TeluguLetter? $TeluguSignVirama)+ $TeluguLetter;
+($L | $V | $LV | $LVT) $L;
+($V | $T) ($LV | $V);
+$T ($LVT | $T);
+
+$Extend      [^$Control $CR $LF];
+$SpacingMark [^$Control $CR $LF];
+[^$Control $CR $LF] $Prepend;
+
+
+## -------------------------------------------------
+
+!!safe_reverse;
+
+
+## -------------------------------------------------
+
+!!safe_forward;
 
-#
-#  Reverse Rule, back up to the beginning of some preceding grapheme cluster.
-#
-! ($Extend | $V | $T )*   ($LF $CR | ($LV | $LVT)?$L* | .);