> Here is the appropriate documentation link, where they have an example: > http://www.postgresql.org/docs/8.3/static/plpgsql-control-structures.html#PLPGSQL-ERROR-TRAPPING > > The primary difference is that they use a loop, which is more robust. In > theory, if you delete the record between when the INSERT happens and > when the UPDATE happens, you will get no effect, which isn't what you're > looking for. The loop will correct for this by ensuring that something > happens before it terminates.
Yes, it could have been removed. I seem to remember that it is not possible to lock the table with a transaction from within a function, this has to be done on the level from which I call the function? > However, you can do the same thing on the command line using > subtransactions, a.k.a. SAVEPOINTs: > http://www.postgresql.org/docs/8.3/static/sql-savepoint.html Oh, fun to learn new stuff :-) Thank you for your reply -- Sent via pgsql-general mailing list (pgsql-general@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
