| From: | Sami Imseih <samimseih(at)gmail(dot)com> |
|---|---|
| To: | David Rowley <dgrowleyml(at)gmail(dot)com> |
| Cc: | Bykov Ivan <i(dot)bykov(at)modernsys(dot)ru>, Michael Paquier <michael(at)paquier(dot)xyz>, "pgsql-hackers(at)postgresql(dot)org" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Query ID Calculation Fix for DISTINCT / ORDER BY and LIMIT / OFFSET |
| Date: | 2025-03-10 17:20:57 |
| Message-ID: | CAA5RZ0sG42-Hyes4U9BToNHfxGx8tkW_HuRkQ0pOmmx6Ud7eiw@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
> transformation. For these reasons, variant A where we put the
> LimitOption between the two int8 expression nodes feels like the
> "okay" approach here. But we must document this expectation in the
> structure, and check for more grammar variants of LIMIT and OFFSET
> clauses in pgss.
Please see the attached. Variant A with comments and some additional
test cases.
It should be noted that we currently have "WITH TIES/ROWS ONLY" tests in pg_s_s,
so I added another case to show "FETCH FIRST 2 ROW ONLY" and "LIMIT 2" are
the same queryId and also added a query that uses both a LIMIT and OFFSET.
I could not think of other cases we need to cover.
--
Sami
| Attachment | Content-Type | Size |
|---|---|---|
| v1-0001-Fix-QueryId-collision-for-LIMIT-and-OFFSET.patch | application/octet-stream | 2.9 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tomas Vondra | 2025-03-10 17:35:56 | Re: Changing the state of data checksums in a running cluster |
| Previous Message | Ranier Vilela | 2025-03-10 16:55:41 | Re: table_tuple_lock's snapshot argument |