v0.6

• support single Chinese character search

v0.5.1

• add CREDIT.txt and Licence infomation

v0.5

• use regular expression to compatible with defualt English white space splitter
• removed configuration file, much simpler code, easy to install, easy to use
• support multiple encodings: unicode/utf-8/gb18030/gbk/gb2312/mbcs/big5. provide 3 splitters:
  • CJK splitter : support unicode/utf-8 encoding. this encoding is compatible with version 0.1
  • CJK GB splitter : support unicode/gb18030/gbk/gb2312/mbcs encodings.
  • CJK BIG5 splitter : support unicode/big5/mbcs encodings
• unicode encoding is detected automatically. this make CJKSplitter compatible with Archtypes 1.2+ (string stored as unicode)
• better encoding handling to avoid exception (replace)
• smaller index storage for CJK: index stored as unicode(2 byts) but not utf-8(3 bytes)
• support english globing
• precise CJK char recongnize (\u4E00-\u9FFF)
• maybe better performance, not tested
• better documentations (thanks bjorn!)

v0.2

this is bjorn's([email protected]) contributes

v0.1

initial release, support utf-8 encoding only.