[Biojava-l] Re: Biojava-l digest, Vol 1 #334 - 2 msgs

Thomas Down td2@sanger.ac.uk
Tue, 12 Jun 2001 12:56:46 +0100


Sarath's right.  This issue probably /IS/ more Genbank's
fault than ours, but realistically the BioJava parser is
being far too strict.  Especially since it doesn't validate
for the presence of other header entries -- just GI.

I've added an extra conditional in the code, and it now
tolerates VERSION lines which just contain a single token.

Sarath: this is checked into CVS now -- do you have access
to the CVS versions?


   Thomas.




On Tue, Jun 12, 2001 at 01:37:23PM +0200, Sarath wrote:
> Hi keith 
>   I would have accepted your point very well if the files found at
> 
>    ftp://ftp.ncbi.nlm.nih.gov/genomes/Bacteria/Staphylococcus_aureus/
> 
>  read the sequences as these are supposed to contain the actual gene bank
> formatas per the README you have refered to but neverthless they failed
> and the README at this url below
>    ftp://ftp.ncbi.nlm.nih.gov/genomes/Bacteria/
> says even the so called better versions have the same sequences
> 
>   "At present only Archaeoglobus fulgidus genomes is considered
> "reviewed", other genomes are presented as found in the source GenBank
> records "