Re: BUG #17579: 15beta2: strange error when trying to use MERGE statement as a CTE

Tom Lane <tgl@sss.pgh.pa.us>

From: Tom Lane <tgl@sss.pgh.pa.us>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
Cc: Richard Guo <guofenglinux@gmail.com>, borz_off@cs.msu.su, PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Date: 2022-08-11T14:23:15Z
Lists: pgsql-bugs
Alvaro Herrera <alvherre@alvh.no-ip.org> writes:
> On 2022-Aug-11, Tom Lane wrote:
>> The parse error location seems quite oddly chosen.  Can you make
>> it point at the MERGE instead?  I think exprLocation(cte->ctequery)
>> might work, but not sure.

> I tried, but couldn't find the location anywhere.  None of the nodes for
> ModifyTable statements have location :-(

Ah, right, in particular MergeStmt doesn't.

> We could add one now -- while
> it's a bit late for 15, we're likely going to have a catversion bump due
> to the JSON revert, so it might be fine.

Nah, not worth it just for this.

			regards, tom lane



Commits

  1. Reject MERGE in CTEs and COPY