[Biojava-l] The org.biojava.bio.program.* hierarchy

[Thomas Down]

On Tue, Jun 19, 2001 at 01:49:34PM +0100, Keith James wrote:
> 
> Related to this, I'm working on Blast/Fasta SAX2 -> Biojava objects
> (org.biojava.bio.search.* objects so far, rather than
> org.biojava.bio.seq.homol) and was puzzling over where this should
> go. Any preferences?

Cool.  That would be a good thing to get in before the 1.2
freeze.  Do you have any idea of timescale, yet?

I'd suggest that this code probably /is/ a good candidate
for putting in a program.* subpackage, since it's targetted
at specific applications (albeit quite nicely abstracted by
the SAX parsers).  How about we keep the search package
reserved for the interfaces (plus any pure-java searchers
which might get added in the future)?

> As an aside, this means that I'm making necessary changes to
> org.biojava.bio.search.* as the hit objects were missing key methods
> such as getting the position/strand of the hit on the query and
> subject sequences. For some reason they were missing from the original
> interfaces.

Yes, those interfaces make quite good starting points,
but there are plenty of wrinkles -- iron away :).

   Thomas.