On Wed, Nov 24, 2010 at 7:03 PM, Bruce Momjian <bruce@momjian.us> wrote:
> Daniel Farina wrote:
>> On Wed, Nov 24, 2010 at 4:30 PM, Bruce Momjian <bruce@momjian.us> wrote:
>> > Robert Haas wrote:
>> >> With respect to the syntax itself, I have mixed feelings. ?On the one
>> >> hand, I'm a big fan of CREATE IF NOT EXISTS and DROP IF EXISTS
>> >> precisely because I believe they handle many common cases that people
>> >> want in real life without much hullabaloo. ?But, there's clearly some
>> >> limit to what can reasonably be done this way. ?At some point, what
>> >> you really want is some kind of meta-language where you can write
>> >> things like:
>> >>
>> >> IF EXISTS TABLE t1 THEN
>> >> ? ?ALTER TABLE t1 DROP CONSTRAINT IF EXISTS t1_constr;
>> >> END IF;
>> >
>> > FYI, I have felt this way for a while. ?IF EXISTS seemed like something
>> > that should never have been added as an inline SQL command option; it
>> > just crept in, and kept growing.
>>
>> Okay, that being the case: would it make sense to have pg_dump emit DO
>> blocks? I have a feeling this might draw fire, but I don't see any
>> reason why the mechanism would not work to more or less equivalent
>> effect. Certainly making dumps harder to use for those who insist on
>> disabling PL/PGSQL is probably a negative side effect, if one can
>> identify this hypothetical class of person.
>
> Not being able to recover a dump is serious problem for a user.
Even if it only involves enabling PLPGSQL to do the restore? Also take
into consideration that plpgsql is enabled by default. A user would
have to change the template database (which, in general, can cause
restores to fail in at least a few other ways) or drop the procedural
language explicitly to make that mechanism not work with a fresh and
normal-looking createdb.
--
fdr