🚀 go-pugleaf

All,

Following a report by a user at Sanger I have found a serious bug has been 
introduced into the way acedb parses in data, the bug has been in the code for 
several years so its surprising it has not come to light before.

The problem can be summarised as follows, if you parse in the following file:


//------start--------
CDS : "R09E10.1"

Sequence : "Eds_sequence"
CDS_child "R09E10.1" 11111 22222

Sequence : "R09E10"
-D CDS_child "R09E10.1" 11111 22222
CDS_child "R09E10.1" 33333 44444
//------------end-----

You end up with "R09E10" looking like this:

Sequence R09E10
   SMap    S_Child     CDS_child   R09E10.1

i.e. you have lost the new coordinates from your change.

There is a work around in that if you do the deletion and addition in two stages 
then all will be fine:

Sequence : "R09E10"
-D CDS_child "R09E10.1" 11111 22222

Sequence : "R09E10"
CDS_child "R09E10.1" 33333 44444


I am urgently looking to find when this problem was introduced and to fix it.

Ed
-- 
  ------------------------------------------------------------------------
| Ed Griffiths, Acedb development, Informatics Group,                    |
|  The Morgan Building, Sanger Institute, Wellcome Trust Genome Campus   |
|  Hinxton, Cambridge CB10 1HH                                           |
|                                                                        |
| email: edgrif@sanger.ac.uk  Tel: +44-1223-496844  Fax: +44-1223-494919 |
  ------------------------------------------------------------------------