Tom Lane wrote:
> Oskar Berggren <beo@sgs.o.se> writes:
>
>>Would it be difficult to make PostgreSQL case insensitive for real when
>>comparing identifiers?
>
>
> You could try altering the name comparison functions --- don't forget
> the btree 3-way comparator, too. And something would have to be done
> with the hash function for names. And I'm not sure what else; but I
> suspect there are a lot of strcmp() calls that would have to be changed.
Thank you for your help, but this seems to be a bit too much work that
almost certainly will lead to stability problems of some sort.
I am now working on a another theory instead. If I can get the
application developers to correct all their query strings so that
they always use the exact same capitalization, and then make PostgreSQL
be always case sensitive, quoted or not. This I believe would only
require that I comment out a few rows in scan.l. Provided that
everything that tries to talk to the database always use the correct
capitalization.
I have tried this and it seems to work for the query i mentioned in
my original posting, which is as expected since that one actually
use correct capitalization. However, shortly afterwards the
application issue the query:
SELECT "PassWord" FROM "Operators" WHERE Name='...' AND Password='...';
Since the column is called PassWord, you can see why I'm feeling
I bit down right now...
regards,
Oskar
--
/----------------------------------------------------------------------\
| Oskar Berggren beo@sgs.studenthem.gu.se |
| Network and Software Engineer SGS Datanätgrupp |
| Gothenburg, Sweden |
\----------------------------------------------------------------------/