| From: | Henson Choi <assam258(at)gmail(dot)com> |
|---|---|
| To: | Tatsuo Ishii <ishii(at)postgresql(dot)org> |
| Cc: | vik(at)postgresfriends(dot)org, er(at)xs4all(dot)nl, jacob(dot)champion(at)enterprisedb(dot)com, david(dot)g(dot)johnston(at)gmail(dot)com, peter(at)eisentraut(dot)org, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Row pattern recognition |
| Date: | 2026-03-01 11:02:38 |
| Message-ID: | CAAAe_zBdtRuAfyEX6eqtprY8ejGXqo6ndrpy0LPEHSzGWDvpxw@mail.gmail.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hi Tatsuo,
Attached are two more incremental patches (nocfbot-0011, nocfbot-0012)
on top of v43, continuing the nocfbot-0001..0010 series.
nocfbot-0011: Add RPR DEFINE expression cost to WindowAgg cost estimation
Revised version of your cost estimation patch [1]. Fixes the
`(char *)` cast and moves DEFINE cost outside the windowFuncs loop.
nocfbot-0012: Reject qualified column references in RPR DEFINE clause
Revised version of your qualified column reference patch [2].
Exposes `patternVarNames` via `ParseState.p_rpr_pattern_vars` to
distinguish pattern variable qualifiers ("not supported") from
FROM-clause range variable qualifiers ("not allowed").
Variables appearing only in DEFINE (not in PATTERN) are also
collected into `p_rpr_pattern_vars` so they get the "not supported"
message, but the RPR_VARID_MAX count check still applies only to
PATTERN variables.
[1]
https://www.postgresql.org/message-id/20260227.225456.33226875991025537.ishii@postgresql.org
[2]
https://www.postgresql.org/message-id/20260227.145539.1921177948671828231.ishii@postgresql.org
Attachment:
nocfbot-0011-define-cost.txt
nocfbot-0012-qualified-refs.txt
Best regards,
Henson
| Attachment | Content-Type | Size |
|---|---|---|
| nocfbot-0011-define-cost.txt | text/plain | 1.9 KB |
| nocfbot-0012-qualified-refs.txt | text/plain | 13.7 KB |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jim Vanns | 2026-03-01 11:22:21 | Re: [PATCH] Add support for SAOP in the optimizer for partial index paths |
| Previous Message | Madhav Madhusoodanan | 2026-03-01 10:45:22 | Re: [WiP] B-tree page merge during vacuum to reduce index bloat |