From 2b714f2bcef3753d5ff18e0804bde7ca252af9cc Mon Sep 17 00:00:00 2001
From: Andy Heninger <andy.heninger@gmail.com>
Date: Thu, 31 Mar 2005 01:45:27 +0000
Subject: [PATCH] ICU-4157 Word Break, fix problem with CR <combining>
 sequences

X-SVN-Rev: 17427
---
 icu4c/source/data/brkitr/word.txt      | 4 ++--
 icu4c/source/test/intltest/rbbitst.cpp | 3 +++
 2 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/icu4c/source/data/brkitr/word.txt b/icu4c/source/data/brkitr/word.txt
index aaf98bfc84..07e7862cb2 100644
--- a/icu4c/source/data/brkitr/word.txt
+++ b/icu4c/source/data/brkitr/word.txt
@@ -52,7 +52,7 @@ $ExtendNumLetEx = $ExtendNumLet $Extend*;
 #          see character breaks.
 
 $CR $LF;
-[^$Control] $Extend+;
+[^$Control $CR $LF] $Extend+;
 $NumericEx $Extend* {100};
 $ALetterEx $Extend* {200};
 
@@ -114,7 +114,7 @@ $LF $CR;
 
 # see character breaks
 
-$Extend* [^$Control];
+$Extend* [^$Control $CR $LF];
 
 # rule 5
 
diff --git a/icu4c/source/test/intltest/rbbitst.cpp b/icu4c/source/test/intltest/rbbitst.cpp
index 071a7c18e4..693622a784 100644
--- a/icu4c/source/test/intltest/rbbitst.cpp
+++ b/icu4c/source/test/intltest/rbbitst.cpp
@@ -2139,6 +2139,8 @@ RBBIWordMonkey::RBBIWordMonkey() : fGCFMatcher(0),
     fOtherSet->removeAll(*fMidNumSet);
     fOtherSet->removeAll(*fNumericSet);
     fOtherSet->removeAll(*fExtendNumLetSet);
+    fOtherSet->removeAll(*fFormatSet);
+    fOtherSet->removeAll(*fExtendSet);
 
     fSets->addElement(fALetterSet,   status);
     fSets->addElement(fKatakanaSet,  status);
@@ -2146,6 +2148,7 @@ RBBIWordMonkey::RBBIWordMonkey() : fGCFMatcher(0),
     fSets->addElement(fMidNumSet,    status);
     fSets->addElement(fNumericSet,   status);
     fSets->addElement(fFormatSet,    status);
+    fSets->addElement(fExtendSet,    status);
     fSets->addElement(fOtherSet,     status);
     fSets->addElement(fExtendNumLetSet, status);