Re: [HACKERS] Conditional trigger missing from pg_dump

2010-03-25 Thread Josh Berkus
>> This was discovered by Dave Olszewski while testing conditional triggers. > > Please show a specific example of what you're worried about. I'll get you a reproduceable test case later today. In our testing, the conditional part of the trigger was not being dumped. --

Re: [HACKERS] Conditional trigger missing from pg_dump

2010-03-25 Thread Tom Lane
Josh Berkus writes: > It looks like pg_dump was never modified to include conditional > triggers. If you look in pg_dump.c, there's no provision to dump the > FOR UPDATE OF part of the trigger declaration. That proves nothing; it could be relying on server-side code to produce the declaration.

[HACKERS] Conditional trigger missing from pg_dump

2010-03-25 Thread Josh Berkus
All, It looks like pg_dump was never modified to include conditional triggers. If you look in pg_dump.c, there's no provision to dump the FOR UPDATE OF part of the trigger declaration. This was discovered by Dave Olszewski while testing conditional triggers. --