| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | MERGE lacks ruleutils.c decompiling support!? |
| Date: | 2023-05-05 13:36:41 |
| Message-ID: | 3579737.1683293801@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
I made this function:
CREATE OR REPLACE FUNCTION test_fun()
RETURNS void
LANGUAGE SQL
BEGIN ATOMIC
MERGE INTO target
USING source s on s.id = target.id
WHEN MATCHED THEN
UPDATE SET data = s.data
WHEN NOT MATCHED THEN
INSERT VALUES (s.id, s.data);
end;
It appears to work fine, but:
regression=# \sf+ test_fun()
ERROR: unrecognized query command type: 5
and it also breaks pg_dump. Somebody screwed up pretty badly
here. Is there any hope of fixing it for Monday's releases?
(I'd guess that decompiling the WHEN clause would take a nontrivial
amount of new code, so maybe fixing it on such short notice is
impractical. But ugh.)
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Padmavathi G | 2023-05-05 13:56:56 | Re: Tables getting stuck at 's' state during logical replication |
| Previous Message | Andrew Dunstan | 2023-05-05 13:23:37 | Re: Add LZ4 compression in pg_dump |