[Biojava-dev] How to write a Blast-like parser?

[Christian Gruber]

Hello!

I'd like to write a parser for some sequence similarity based search 
programs of biological databases, for example BLIMPS etc. The idea now 
was to extend BioJava, that is to write a SAX2 parser for this purpose.

So, I've looked into some org.biojava.bio.program.sax classes, and as it 
seems, there are 2 classes for each program: One Facade class and one 
actual parser (at least, for blast there is BlastLikeSAXParser and 
BlastSAXParser, and for fasta there are FastaSearchSAXParser and 
FastaSearchParser), which confuses me a bit, especially the naming of 
the classes.

 From this point on, I need some guiding hand...

My question is now: When I want to write an additional parser, what 
classes do I have to write, which interfaces do they have to implement 
or which base classes can they extend, which functionality goes where 
etc. I don't expect an elaborate answer to my question here, just a few 
hints where to look first in order to help me understand properly what's 
happening here. And after that, I'd need some advice how to commit my 
extensions ;-)

Thank you very much!

Christian