[Biojava-l] Proposed SeqFileFormer interface change

Thomas Down td2@sanger.ac.uk
Fri, 21 Sep 2001 14:43:14 +0100


That sounds sensible.  In cases like the EMBL/Genbank feature
table format, the location line is definitely a function of
the whole Feature -- not just the biojava Location.

   Thomas.

On Fri, Sep 21, 2001 at 09:44:30AM -0400, Cox, Greg wrote:
> I have implemented location formatting for remote features [i.e. join(5..10,
> AL121903:(10..105))], but in doing so I had to introduce a new method to the
> SeqFileFormer interface.  Since there's an interface change, I'd like some
> consensus before comitting to CVS.  Right now, the only way to format a
> location is via SeqFileFormer.formatLocation(StringBuffer, Location,
> Strand).  I added SeqFileFormer.formatLocation(Feature), which will detect a
> RemoteFeature and format its location properly, and for non-remote features
> it extracts the location field and formats that.
> 
> 	This is needed because the location field of the RemoteFeature class
> doesn't completely describe the location of the feature, being the
> projection of the feature onto the sequence it's attached to.  A field
> specific to RemoteFeature (RegionList) is needed to describe the complete
> location of the feature.  In the next few days, I hope to add the capability
> to distinguish between order(a,b) and join(a,b) and this function will be
> needed to format an order() style location.