Re: How to schema-qualify "IS DISTINCT FROM" on trigger definition? (I created a db that cannot be dump/restored) - Mailing list pgsql-general

From Vick Khera
Subject Re: How to schema-qualify "IS DISTINCT FROM" on trigger definition? (I created a db that cannot be dump/restored)
Date
Msg-id CALd+dccdvmt+kHor7erfSkZo+8-CKz1XM6OuVcGM8ZORPedrSQ@mail.gmail.com
Whole thread Raw
In response to Re: How to schema-qualify "IS DISTINCT FROM" on trigger definition? (I created a db that cannot be dump/restored)  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: How to schema-qualify "IS DISTINCT FROM" on trigger definition? (I created a db that cannot be dump/restored)  (Erik Wienhold <ewie@ewie.name>)
List pgsql-general
On Wed, Feb 21, 2024 at 4:27 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
For the moment, I think the only feasible solution is for your trigger
function to set the search path it needs by adding a "SET search_path
= whatever" clause to the function's CREATE command.

The error is not in the function, it is the WHEN clause of the trigger. There's no way to set a search path on the trigger as far as I see.

The only option I see is to remove the WHEN clause on the trigger and wrap my function with an IF with those same conditions. I hope this will not result in any noticeable difference in speed.

It is rather unfortunate that one can end up with a schema that a backup with pg_dump cannot be restored. Feel free to keep my example for regression testing when postgres does grow the ability to schema-qualify such operators.

pgsql-general by date:

Previous
From: Thomas Munro
Date:
Subject: Re: How do you optimize the disk IO when you cannot assume a file will start at a boundary then?
Next
From: Vick Khera
Date:
Subject: Re: Performance issue debugging