Re: Move un-parenthesized syntax docs to "compatibility" for few SQL commands - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: Move un-parenthesized syntax docs to "compatibility" for few SQL commands
Date
Msg-id ZGMigV+oBj4LjsFB@paquier.xyz
Whole thread Raw
In response to Re: Move un-parenthesized syntax docs to "compatibility" for few SQL commands  (Nathan Bossart <nathandbossart@gmail.com>)
Responses Re: Move un-parenthesized syntax docs to "compatibility" for few SQL commands
List pgsql-hackers
On Mon, May 15, 2023 at 12:36:51PM -0700, Nathan Bossart wrote:
> There's still some time before we'll be able to commit any of these, but
> here is an attempt at addressing all the feedback thus far.

-   The parenthesized syntax was added in
-   <productname>PostgreSQL</productname> 9.0;  the unparenthesized
-   syntax is deprecated.
[...]
+           | CLUSTER '(' utility_option_list ')'
+               {
+                   ClusterStmt *n = makeNode(ClusterStmt);
+
+                   n->relation = NULL;
+                   n->indexname = NULL;
+                   n->params = $3;
+                   $$ = (Node *) n;
+               }

Hmm.  This is older than the oldest version we have to support for
pg_upgrade and co.  Would it be worth switching clusterdb to use the
parenthesized grammar, adding on the way a test for this new grammar
flavor without a table in the TAP tests (too costly for the main
regression test suite)?
--
Michael

Attachment

pgsql-hackers by date:

Previous
From: "Drouvot, Bertrand"
Date:
Subject: Re: walsender performance regression due to logical decoding on standby changes
Next
From: Amit Kapila
Date:
Subject: Re: running logical replication as the subscription owner