ohp@pyrenet.fr writes:
> I've got two problems:
> Looking at the errors, ISTM foreign statement is the over way round :
> levt_tevt_cod is in ligne_evt NOT in type_evt
No, that's just how we've worded FK violation errors for some time.
The real question is how did FK violations get into your dump?
> And, looking at the generated foreign keys, I don't see ON UPDATE/ON
> DELETE clauses
The ones we can see here are default (NO ACTION) cases, so I'm not
convinced you're describing a real problem.
regards, tom lane