[Bioperl-l] Question about using XML::Parser::PerlSAX for sequence parsing

[Donald G. Jackson]

Hi,

I'm working on a SeqIO adapter for the NCBI TinySeq XML format and have 
a question about XML::Parser::PerlSAX.

Currently,  my parser goes through the whole XML file and stores the 
results in a stack.  My next_seq() method pulls the next element off the 
stack, converts it to a  Bio::Seq object, and returns it.  This works, 
but I'd rather be able to step through the input XML one sequence at a 
time.  I thought bioperl selected PerlSAX because it could do that, but 
I'm having trouble figuring out how.  I looked at Bio::SeqIO::game, but 
that also loads all of the parsed data into memory.

I'd like to implement something analagous to how Bio::SeqIO::fasta steps 
through an input file, though without having to get to such low-level 
methods.  If anyone on the list has any suggestions, I'd be grateful!

Thanks,

Don Jackson
BMS Bioinformatics  

PS - I'm putting together a test suite and hope to have the preliminary 
version of Bio::SeqIO::tinyseq checked into the cvs repository in the 
next day or so.