IUBio GIL .. BIOSCI/Bionet News .. Biosequences .. Software .. FTP

AQL vs. tags beginning with a digit

Richard Durbin rd at sanger.ac.uk
Thu Aug 21 18:20:05 EST 2003

Hmm.  Jean and I did agree it would be better not to have tags beginning 
with numbers quite some time ago, which was why AQL doesn't allow them. 
  I guess we didn't outlaw them for practical backward compatibility 
reasons.  But consider them deprecated - better to use standard 
progamming language type identifiers, except that ones prefixed with an 
underscore are reserved to the system (is that documented anywhere?).


Keith Bradnam wrote:
> On 13 Aug 2003, Dave Matthews wrote:
>>It's legal for tag names to begin with a number, right?  It works fine
>>in all other respects but in AQL it gives a syntax error.
>>E.g. at WormBase, http://www.wormbase.org/db/searches/aql_query,
>>"select U from U in class UTR where exists_tag U->5_UTR" gives
>>"AQL error 700 around: '_tag U->5_UTR"
> Hi Dave,
> It's legal as far as I know, and we have often used numbers as part of
> tag names in WormBase (as your example shows with the '5_UTR' tag).
> Never been aware of any querying problems though, but then I hardly ever
> use AQL.  If it is a number-related problem, you could maybe check by
> writing an AQL query against the '2_point' tag in the ?Rearrangement
> class.  If you still get a bug, then this should be enough info for the
> acedb guys to track it down (note though that Ed is on holiday at the
> moment).
> Regards,
> Keith
> ~  Keith Bradnam - WormBase group: http://www.wormbase.org/
> ~
> ~  The Wellcome Trust Sanger Institute,
> ~  Hinxton, Cambridge, CB10 1SA, UK.  Tel: 01223 494922

More information about the Acedb mailing list

Send comments to us at archive@iubioarchive.bio.net