Philip Yarra wrote:
> Oliver Jowett wrote:
>
>> Essentially I think it boiled down to "getTableName() should return
>> the table alias name", and since we don't have that available we
>> return "". There's a postgresql-specific interface to get at the
>> underlying table name (in the cases where that info is available).
>
>
> Hmmm... maybe I'm missing something, but the attached test case doesn't
> use any aliases, and it returns "" (at least it does for me). Is this
> expected behaviour? If there's a postgresql-specific way to get it,
> couldn't this method implement the same way of getting it?
We don't know whether aliases have been used or not so we can't do this.
-O