[Biojava-dev] NCBI Blast XML Parser update

[Muhammad Ali]

Hi,

Sorry for the delay, I got caught up with my other project. Anyway,
I'm here now and I've attached the patch file with this email. Please
take a look and let me know if changes are required.

With regards to testing, there is a BlastParser.java file in the
/demos/blastxml folder that I've been using to test on my end. It
takes a Blast output file as its input. If you want, I can provide
some Blast output file with multiple iteration entries for use as a
sample input file for testing.

I think I'll do without SVN write access. I don't expect to be working
on other parts of biojava right now, unless someone else in my lab
complains about something else being broken ;). But if I do make any
local patches, I'll definitely try and get them back to you guys.

Cheers,
Ali.

On Tue, Aug 19, 2008 at 9:36 PM, Andreas Prlic <andreas at sdsc.edu> wrote:
> Hi Ali,
>
> that's good news, can you send your patch to this list, so we can have
> a look? At this stage it would be also great to provide a new Junit
> test that makes sure that the parsing works ok and to demonstrate the
> new feature. If you provide more patches in the future we can also set
> up write access to the SVN repository.
>
> Andreas
>
> On Tue, Aug 19, 2008 at 6:10 AM, Muhammad Ali <talk2ali at gmail.com> wrote:
>> Hello,
>>
>> The current version of the BlastXMLParser (used for parsing NCBI BLAST
>> output files) is not handling multiple Iteration entries in the file
>> correctly. It lumps them all together, resulting in a loss of
>> search-specific parameters. The end result is a single
>> SeqSimilaritySearchResult object. The expected output should be one
>> SeqSimilaritySearchResult for each Iteration entry in the file.
>>
>> I've fixed this issue on my locally checked out copy by modifying a
>> few files in the  org.biojava.bio.program.sax.blastxml package. I'm
>> interested in submitting the updated code back to the main repository.
>> Can someone tell me how I can go about doing so?
>>
>> Also the javadocs for the BlastXMLParser seem to be outdated. They
>> mention malformed XML being generated by NCBI, but that doesn't seem
>> to be the case anymore.
>>
>> Thanks,
>> Ali.
>> _______________________________________________
>> biojava-dev mailing list
>> biojava-dev at lists.open-bio.org
>> http://lists.open-bio.org/mailman/listinfo/biojava-dev
>>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: changes.patch
Type: application/octet-stream
Size: 34641 bytes
Desc: not available
URL: <http://lists.open-bio.org/pipermail/biojava-dev/attachments/20081121/9adcfb8b/attachment-0002.obj>