Use specific characters instead of punktuation. Makes a difference
authorDennis Schafroth <dennis@indexdata.com>
Tue, 22 May 2012 11:40:35 +0000 (11:40 +0000)
committerDennis Schafroth <dennis@indexdata.com>
Tue, 22 May 2012 11:40:35 +0000 (11:40 +0000)
test/test_icu.4.xml

index cc97b3f..e5e5cdb 100644 (file)
@@ -1,7 +1,7 @@
 <icu_chain locale="en">
-  <tokenize rule="s"/>
+<tokenize rule="s" /> 
   <transliterate>
-[:punctuation:]* } [$] > ; 
+[,. -]* } [$] > ; 
  </transliterate>
 </icu_chain>