From: Tom Lane <tgl@sss.pgh.pa.us>
To: Robert Haas <robertmhaas@gmail.com>
cc: Andreas Karlsson <andreas@proxel.se>, Andres Freund <andres@anarazel.de>,
        Thomas Munro <thomas.munro@enterprisedb.com>,
        Andrew Gierth <andrew@tao11.riddles.org.uk>,
        David Fetter <david@fetter.org>,
        Pg Hackers <pgsql-hackers@postgresql.org>
Subject: Re: Early WIP/PoC for inlining CTEs
In-reply-to: 
 <CA+TgmoYZYj8eLdV9AcROtmPW4AJA7ANh3xQh7BcafaPfCnmv+g@mail.gmail.com>
References: <dccb37dd-2f52-dbd7-a4c4-72a5853f4838@proxel.se>
 <e02af96a-3e25-5de4-09a4-a7ebf3452184@proxel.se>
 <8810.1542402910@sss.pgh.pa.us>
 <d4daf011-01cf-6bf9-3f15-85ae1572e909@proxel.se>
 <87r2dxrw1z.fsf@news-spur.riddles.org.uk>
 <d0619160-10af-dcc9-4e7d-57845b92570c@proxel.se>
 <CAEepm=2x3kJ5AdkKmEf_ODAZm7+e=7DW1u-+ZwfncWvd8xDaiA@mail.gmail.com>
 <CA+Tgmob=Dpje+oGCYgJCoUok=iPQarGjPqRLYS1qKzChkFfKcA@mail.gmail.com>
 <20190111181934.g5qig23o7rxubnjt@alap3.anarazel.de>
 <CAEepm=2n8hChj548H0R5g5JAx3cokpZwTqRzYhN2oVGqBONwOA@mail.gmail.com>
 <20190111190403.77ye7m5mxypxsf4l@alap3.anarazel.de>
 <CA+TgmobVMPX1wJxPaRQ=FZBzp-AQ=JWK=bzxztqwj=MsOsrnrA@mail.gmail.com>
 <00818be9-2f63-f584-efcd-d97ee0662905@proxel.se>
 <CA+TgmoapASa3RQi=DuSikPoPF4363czSD8KxbOxsfW5XmO3WpQ@mail.gmail.com>
 <5715d93e-6ebc-be17-c2c3-56a873100a38@proxel.se>
 <29576.1548084476@sss.pgh.pa.us>
 <CA+TgmoYZYj8eLdV9AcROtmPW4AJA7ANh3xQh7BcafaPfCnmv+g@mail.gmail.com>
Comments: In-reply-to Robert Haas <robertmhaas@gmail.com>
	message dated "Mon, 28 Jan 2019 14:00:33 -0500"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <12407.1548707732.1@sss.pgh.pa.us>
Date: Mon, 28 Jan 2019 15:35:32 -0500
Message-ID: <12408.1548707732@sss.pgh.pa.us>
Precedence: bulk

Robert Haas <robertmhaas@gmail.com> writes:
> However, generally we have not had great luck with just sticking
> keywords in there (cf. VACUUM, ANALYZE, EXPLAIN, COPY) which is why I
> suggested using a flexible syntax with parenthesized options.

Fair, except that as you then proceed to point out, that does not work
either before or after the AS.

Conceivably we could make it work without the parens:

    WITH ctename AS [ option = value [ , .... ] ] ( query .... )

which for the immediate feature I'd be tempted to spell as

    WITH ctename AS [ materialize = on/off ] ( query ... )

I think the only reason the syntax is MATERIALIZED with a D is that
that's already a keyword; it reads a bit awkwardly IMO.  But if we
were accepting a ColId there, there'd be room to adjust the spelling.

That said, this is still clunkier than the existing proposal, and
I'm not really convinced that we need to leave room for more options.

			regards, tom lane