[Biopython-dev] "Online" tests, was [Bug 1972]

[Peter (BioPython-dev)]

>>>>In some cases (e.g. GenBank, Fasta) once the sample file is downloaded 
>>>>there are multiple parsers to be checked (e.g. record and feature parsers).
> 
> I have tests using EUtils that check the record and feature parsers for
> GenBank and Fasta files.

You don't check the iteration over files with multiple records... that 
shouldn't be a problem as other than changing the number of blank lines 
between records I can't think of anything the NCBI might change, but you 
never know.

>>I'll volunteer to add cases for GenBank, Fasta and GEO files.
> 
> I've not yet looked at GEO files, so there's that yet to do. I expect to
> have some free time soon so I may look at the GEO and other parsers at
> that time.

Checking the basic "did it load something with the expected ID" 
shouldn't be too tricky for GEO Soft files.

 > I'd certainly be interested in any feedback you have on these
> tests, or to see additional test cases. Finding the parsers to test and
> implementing the tests seems pretty straightforward.

You are only testing XML parsing with blastp - do you think its worth 
including tests of any of the other variants?  Especially the slightly 
more exotic ones offered by the NBCI.

Also, you are only submitting a single query, so only a single blast 
result is returned.  Can multiple queries be submitted as this would 
give you an excuse to test the Blast iterator support.

Loading a PDB file strikes me as another fairly simple one to include. 
If you can get a link to the "PDB file of the month" then even better.

Are there any "online" phylogenetic tree resources we should use to 
valid ate the new Nexus parser?

> I've attached the test code to this email, and added a .txt file
> extension just in case the listserv won't allow attaching a code file.

I can see the test code file.  I haven't run it yet.

Thanks Bill.

Peter