[Biopython-dev] changes

[Andrew Dalke]

I've committed some changes in CVS:

 - added 'fasta' reader
 - modified the 'genbank' reader to take the new-style Std tags
    (also fixed the "" bug in feature qualifier values)
    (the parsing is about 1/2 the performance of SWISS-PROT;
     haven't figured out why)
 - added a 'block' parser, but no builder yet
 - added the beginnings of a 'blast' parser
 - added a DBXRef module for database cross references
 - a couple additions to Martel:
     - SkipLinesUntil / SkipLinesTo ... equivalent to
           while not pattern.match(line):
               line = infile.next()
     - can now iterate HeaderFooter records
 - the SeqRecord now stores features and database cross references
 - the SeqRecord stores a Seq instead of a string
     - genbank and swissprot records set the correct alphabet type
 - the 'parse' and 'identify' commands now take XML 'source' objects,
     which can be a URL or a file handle.

Huh.  Guess I was busy this evening.

Still need to run full regressions against the new formats.  BTW,
the FASTA reader tries to parse the NCBI fields and generate appropriate
dbxref fields for the SeqRecord.

                    Andrew
                    dalke at dalkescientific.com