java - Tokenize a string using apache lucene -

how tokenize string based on patter?

example. in following string

arg1:aaa,bbb , arg2:ccc or arg3:ddd,eee,fff 

first want tokenize based on , and or

so

token set 1 arg1:aaa,bbb  token set 2 arg2:ccc  token set 3 arg3:ddd,eee,fff 

later want pass these individual token sets method , tokenize based on ":"

token set 1 token 1 aaa token 2 bbb  token set 2 token 1 ccc  token set 3 token 1 ddd token 2 eee token 3 fff 

how tokenize using custom patter using lucene?

to perform custom tokenization implementation, implement own tokenizer. primary method needs implemented tokenstream.incrementtoken().

your tokenizer can incorporated analyzer.