Sérgio
-----Original Message----- From: corpora-bounces at uib.no [mailto:corpora-bounces at uib.no] On Behalf Of Edward Ivanovic Sent: 26 November 2009 04:55 To: corpora at uib.no Subject: [Corpora-List] Java-based chunk parser?
Dear Colleagues,
I'm looking for a Java-based tool that will let me define a simple grammar based on regular expressions to parse a given string. For example:
"2q34w-6q8w 5q8w-11q87w" (etc)
Could be parsed by the following rules:
A: \d+q B: \d+w C: <A>\-<B> D: (<C> )+
mixing regex with my own labels (A,B,C,D). The actual syntax for the rules isn't important.
Parsing D will then give me the groupings for C (of which there will be two), and access to the other labels.
Something like the RegexpChunkParser in NLTK does this very well, but I can't use Python for this (needs to be Java), so was hoping someone would know of something before I write my own.
Many thanks, Edward
_______________________________________________ Corpora mailing list Corpora at uib.no http://mailman.uib.no/listinfo/corpora