Christopher Kings-Lynne <chriskl(at)familyhealth(dot)com(dot)au> writes:
>> Applied by agreement that this is a bug fix.
> You have actually technically done more than just fix the bug. Because
> you are passing in 'true' as the second parameter to
> pg_get_constraintdef, you've enabled the 'pretty print' option... I'm
> not sure it's that big a thing though...
Yeah, I know; that was deliberate. Any objections?
regards, tom lane