[antlr-interest] Re: ANTLR 2.7.2a2 released :)

[Thomas Dudziak]

Sure (its not really obvious ;-).
I use an input stream which converts the input encoding from UTF or
comparable to unicode, which retains the original input stream position such
that the parser can query it and attach it to a node. With this I'm able to
reparse only certain nodes by seeking to this position and using only the
associated rule in the parser. In order not to destroy the position
information as provided by the LexerShardedInputStream, I manipulate it with
such a setter method.

Ciao, Tom



 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/