canonical transcript.

return undef (which is not a good array reference).

Code originally allowed a user to not specify a DBConncetion during construction. This is not the right behaviour and has now been tidied up

calculations, implemented by Steve Searle.

# patch_59_60_c.sql
#
# Title:
#   A patch to fix a couple of inconsistencies in the schema.
#
# Description:
#   QC turned up issues with the signedness of a number of fields in
#   the Ensembl Core schema.  This patch fixes these.  The fields
#   are: karyotype.seq_region_start, karyotype.seq_region_end and
#   seq_region.length (should all be UNSIGNED).

to find.

#
# Title:
#   A patch to fix a couple of inconsistencies in the schema.
#
# Description:
#   A couple of fixes to the schema to do with inconsistencies found
#   during QC.

get_all_DBEntries().  From Michael S.

unneccesarily in add_Attributes().  Transcript attributes do not affect
exon sequences.  Also, the code tried to fiddle with the internals of
the exon object, which is a no-no (especially since those particular
internals did not exist!).

While at it, I changed a few of "$hash->{'key'} = undef" into
"delete($hash->{'key'})".

conditional operator.

In fetch_all_by_Transcript():  Be absolutely sure not to add the
exon_transcript table twice.  Should solve a bug reported by Jan.

In translation_id(): Update for this method.

All from Michael S.

throw() has the correct message (but leave it commented out since it
would stop circular genomes working).

In overlap_size(): Correct documentation.


All from Michael S.

Michael S.

Also, reformat method.

Michael S.

Michael S.

Also, reformat method.

message.  From Michael S.

Also, reformat method.

mentioning of return values as none are returned (and the rest of the
API is not expecting anything).  From Michael S.


Also, reformat method.