[Biopython-dev] Martel-0.5

[Andrew Dalke]

Johann:
>BTW, it seems the small bugfix patch against Generate.py
>which you posted to this list on 6 December did not make it into 0.5(?)

Good catch, but I actually what happened was that I withdrew
that patch because of its huge performance impact.  The details
are given in:

http://www.biopython.org/pipermail/biopython-dev/2000-December/000215.html

I pulled the patch and changed the PIR format definition (see
http://www.biopython.org/~dalke/Martel/Martel/formats/PIR_3_0.py )
to:

# The hack is to work around a Martel bug
sequence_data = Rep1(Re("   (?P<start_pos>.....)") + \
                     Rep(Group("hack", sequence_punc_letter)) + \
                     Opt(sequence_punc) + AnyEol())

My current goal is to make sure the Martel behaviour for valid
matches is a proper subset of regexp behaviour and wait for
someone to write a full engine  :)

                    Andrew
                    dalke at acm.org