MIME-Version: 1.0
References: 
 <CAAAe_zAuHwqUfqJOD4PDUkWsxTfTytNaandq11Kddw2bfCcpvQ@mail.gmail.com>
 <20260602.142009.1903449512796629664.ishii@postgresql.org>
 <20260602.142356.343550180613194547.ishii@postgresql.org>
 <20260602.191912.1197273977474107705.ishii@postgresql.org>
In-Reply-To: <20260602.191912.1197273977474107705.ishii@postgresql.org>
Reply-To: assam258@gmail.com
From: Henson Choi <assam258@gmail.com>
Date: Tue, 2 Jun 2026 23:11:05 +0900
Message-ID: 
 <CAAAe_zAv0FZQfdjAX05QpE5qtdQQUrqEfZgOZRb1P9Qincircg@mail.gmail.com>
Subject: Re: Row pattern recognition
To: Tatsuo Ishii <ishii@postgresql.org>
Cc: jian.universality@gmail.com, zsolt.parragi@percona.com,
	sjjang112233@gmail.com, vik@postgresfriends.org, er@xs4all.nl,
	jacob.champion@enterprisedb.com, david.g.johnston@gmail.com,
	peter@eisentraut.org, li.evan.chao@gmail.com, pgsql-hackers@postgresql.org
Content-Type: multipart/alternative; boundary="0000000000006d340c065345e293"
Archived-At: 
 <https://www.postgresql.org/message-id/CAAAe_zAv0FZQfdjAX05QpE5qtdQQUrqEfZgOZRb1P9Qincircg%40mail.gmail.com>
Precedence: bulk

--0000000000006d340c065345e293
Content-Type: text/plain; charset="UTF-8"

Hi Tatsuo,

Catching up on today's notes -- all straightforward, so I'm folding them
into one reply.

> I'd prefer (3). Yes, I agree that 240 pattern variables is enough.

Then (3) it is -- and with Jian on the same choice, that closes it. I'll
go ahead with it. Since it's a narrowing it's only free pre-release, so
it goes in the v48 core rather than a later follow-up.

> static void
> nfa_reevaluate_dependent_vars(WindowAggState *winstate, RPRNFAContext
*ctx,
>                               int64 currentPos)
>
> This function does not have a forward declaration. Moreover, it is put
> in the "API exposed to nodeWindowAgg.c" section. [...] I suggest to add
> a forward declaration for it and move it before "API exposed to
> nodeWindowAgg.c" section.

Agreed -- that was just an oversight. I'll add the forward declaration
and move it out of the API-exposed section.

> nocfbot-0025-...:55:+   (errcode(ERRCODE_WINDOWING_ERROR),
> Maybe this is the only one worth to look into?

Right -- I'll update 0025 in the next posting.

Thanks,
Henson

--0000000000006d340c065345e293
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hi Tatsuo,<br><br>Catching up on today&#39;s notes -- all =
straightforward, so I&#39;m folding them<br>into one reply.<br><br>&gt; I&#=
39;d prefer (3). Yes, I agree that 240 pattern variables is enough.<br><br>=
Then (3) it is -- and with Jian on the same choice, that closes it. I&#39;l=
l<br>go ahead with it. Since it&#39;s a narrowing it&#39;s only free pre-re=
lease, so<br>it goes in the v48 core rather than a later follow-up.<br><br>=
&gt; static void<br>&gt; nfa_reevaluate_dependent_vars(WindowAggState *wins=
tate, RPRNFAContext *ctx,<br>&gt; =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0=
 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 int64 curre=
ntPos)<br>&gt;<br>&gt; This function does not have a forward declaration. M=
oreover, it is put<br>&gt; in the &quot;API exposed to nodeWindowAgg.c&quot=
; section. [...] I suggest to add<br>&gt; a forward declaration for it and =
move it before &quot;API exposed to<br>&gt; nodeWindowAgg.c&quot; section.<=
br><br>Agreed -- that was just an oversight. I&#39;ll add the forward decla=
ration<br>and move it out of the API-exposed section.<br><br>&gt; nocfbot-0=
025-...:55:+ =C2=A0 (errcode(ERRCODE_WINDOWING_ERROR),<br>&gt; Maybe this i=
s the only one worth to look into?<br><br>Right -- I&#39;ll update 0025 in =
the next posting.<br><br>Thanks,<br>Henson<br></div>

--0000000000006d340c065345e293--