[Bioperl-l] [ANNOUNCEMENT] BioPerl 1.6 release agenda

Chris Fields cjfields at illinois.edu
Thu Nov 27 18:58:30 UTC 2008


On Nov 27, 2008, at 11:05 AM, Dave Messina wrote:

> Chris et al,
>
> A few quick notes on the SeqIO test reorg I just committed:
>
> - each SeqIO driver now has its own .t file.
>
> - six modules have all tests skipped (except for use_ok) because  
> there isn't a sample input file against which to test. Those are:
>
> agave
> alf
> chadoxml
> chaos
> flybase_chadoxml
> strider

Possible candidates to be moved out of core into a dev (unless we come  
up with some tests!).

> - where possible, I renamed the sample file in data/ to be  
> test.<format> (e.g. test.fasta, test.embl). There were several  
> already named this way, and future refactoring will benefit from the  
> consistency.

I agree.

> - a few of the modules throw off some warnings (which don't cause  
> any test failures).
>
>  An example is Bio::Location::Fuzzy, called by Bio::SeqIO::genbank,  
> complaining about not finding valid fuzzy encodings.
>
> The modules throwing warnings are:
>
> SeqIO::chadoxml
> SeqIO::genbank (really Location::Fuzzy)
> SeqIO::phd
>
> I don't think it's serious.

Still warrants some investigation, but I agree.  We can turn off  
warning if needed, but I would rather try to find what is triggering  
them (it's almost always something).

> - I opted to organize all of the .t in a subdirectory, SeqIO. So it's
>
> t/SeqIO.t
> t/SeqIO/abi.t
> t/SeqIO/ace.t
> ...etc...
>
> Chris, I know you've been doing SearchIO_blast.t.
>
> I propose this alternative, hierarchical structure so that the t/  
> directory would be more manageable, but I will happily rename the  
> SeqIO test files to conform to your standard if you prefer.

I like!  I'll change AlignIO and SearchIO over to this and work on  
cleaning up test file names.  We'll need to update the wiki accordingly.

> Dave

Thanks Dave!  Feel free to do more restructuring along these lines; we  
have the other IO's (Assembly, Resstriction, Tree, Feature, and a few  
others I think), and I could see collecting various modules into a  
collective structure, such as Ontology, Tools, Graphics, etc.  If you  
do just post a note here or on the wiki (I'll do likewise) so we don't  
work on the same thing.  Anyone else wanting to join in please let us  
know likewise.

chris



More information about the Bioperl-l mailing list