Fixed a bug when parsing embl files: accession ids end with ';' in recent embl…

Fixed a bug when parsing embl files: accession ids end with ';' in recent embl formats. The ';' was kept by the old embl parser, resulting in many possible bugs.

src/obitools/seqdb/embl/parser.py

@@ -25,10 +25,12 @@ def __emblparser(text):
         seq    = _cleanSeq.sub('',seq).upper()
         
         acs    = _acMatcher.search(text).group()
+        acs    = acs.replace(';', ' ')
         acs    = acs.split()
         ac     = acs[0]
         acs    = acs[1:]
 
+        
         de     = _deMatcher.search(header).group()
         de     = _cleanDe.sub(' ',de).strip().strip('.')
     except AttributeError,e: