[Bioperl-l] Thoughts on Bio::Tools::Glimmer

Mark Johnson johnsonm at gmail.com
Thu Apr 12 13:26:33 EDT 2007


    I'd call that a buggy regexp.  Sounds like a good (but minimal)
fix.  Torsten, I don't have cvs write access, I think you do, can you
fix that up?  Andrew, can you file that as a bug:

http://bugzilla.bioperl.org/

    Everything else sounds like enhancements.  I'm not necessarily
opposed, but a little discussion is probably in order before putting
any tickets in for any of that.  Also, I'm not sure when I'll be able
to spare some time to work on the module.  It was easy to justify
spending time from my day job getting the module up to where is now,
as I needed a BioPerl-ish glimmer2/glimmer3 parser.  It's working
quite well for my purposes.  Again, I'm not opposed to further
enhancements, but If I'm going to work on any of them, they'll have to
fit into everything else I'm doing and it could be a while.  However,
there's no reason somebody else can't do what I did.  Discuss the
changes here, work out a plan, implement it, send along the diff(s)
attached to a bug in bugzilla.  Next thing you know, your changes are
in cvs.  8)

On 4/11/07, Torsten Seemann <torsten.seemann at infotech.monash.edu.au> wrote:
> Andrew,
>
> >                 # Glimmer 3.X prediction
> >                 (/\w+(\d+)\s+       # orf (numeric portion)
> > ...isn't picking up more than the last digit in the orf-number.  Not
> > sure if that's intentional.  A sample of the feature output using -
> >  >gff_string shows up as ...
>
> I think that regexp should be \w+?(\d+)
>
> ie. the \w+ should be non-greedy, otherwise it will swallow up all but
> one of the following \d+ (as \d is a subset of \w)
>
> I've CC:ed this to Mark Johnson who made the recent changes to this module.
>
> Thanks for your feedback,
>
> --Torsten Seemann


More information about the Bioperl-l mailing list