[Biojava-l] Problems with SAX parsing

Simon Brocklehurst simon.brocklehurst at cambridgeantibody.com
Fri Jan 31 11:40:29 EST 2003



Alexander Churbanov wrote:
>    Hello Matthew or Thomas,
> 
>    I just use SAX BLAST parser to parse a massive of
> BLAST output. Unfortunately this package does a really
> poor job in a batch mode. It runs out of memory
> because of the recursive self-calls which adversely
> affects the heap.
>    I know that you have borrowed this from another
> team, but are there any ways to fix it, outherwise I
> run into strange situation while parsing 18,000 Blast
> runs.

We've heard of this before - this *may* be due to an old bug that was 
*supposed* to have been fixed in the biojava code base long ago i.e. 
something not getting popped of a stack when it should be.  Should be 
easy to fix if this bug is somehow still in the biojava code.

If could e-mail the code of the ContentHandler you are using it would be 
helpful. It would be useful to confirm that the problem you are 
experiencing is nothing to do with object creation *outside* of the SAX 
driver itself.

An obvious work around for your problem is to split the blast output 
into smaller chunks.

Simon
-- 
Dr Simon M. Brocklehurst, Ph.D.
Director of Informatics & Robotics

Cambridge Antibody Technology
Milstein Building
Granta Park
Cambridge
CB1 6GH
UK

Telephone: + 44 (0) 1763 263233
Facsimile + 44 (0) 1763 263413
Email: mailto:simon.brocklehurst at cambridgeantibody.com
http://www.cambridgeantibody.com

Cambridge Antibody Technology Limited *
Registered Office: The Science Park, Melbourn, Cambridgeshire,
SG8 6JJ, UK. Registered in England and Wales number 2451177
(* Cambridge Antibody Technology Limited is a member of the
Cambridge Antibody Technology Group of Companies)

Confidentiality Note: This information and any attachments is
confidential and only for use by the individual or entity to
whom it has been sent. Any unauthorised dissemination,
distribution or copying of this message is strictly prohibited.
If you are not the intended recipient please inform the sender
immediately by reply e-mail and delete this message from your system.
Thank you for your co-operation.



More information about the Biojava-l mailing list