Re: Proposal : REINDEX xxx VERBOSE

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Sawada Masahiko <sawada(dot)mshk(at)gmail(dot)com>
Cc: Fujii Masao <masao(dot)fujii(at)gmail(dot)com>, Fabrízio Mello <fabriziomello(at)gmail(dot)com>, Jim Nasby <Jim(dot)Nasby(at)bluetreble(dot)com>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Michael Paquier <michael(dot)paquier(at)gmail(dot)com>, PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Proposal : REINDEX xxx VERBOSE
Date: 2015-05-05 20:42:40
Message-ID: CA+TgmoZm7d6mvKA7y=4tw-sfGDBz9_KjQGdXUNRKAbnCwP67Qw@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, May 5, 2015 at 11:10 AM, Sawada Masahiko <sawada(dot)mshk(at)gmail(dot)com> wrote:
> On Fri, May 1, 2015 at 9:04 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> On Thu, Apr 30, 2015 at 11:05 PM, Sawada Masahiko <sawada(dot)mshk(at)gmail(dot)com> wrote:
>>> VACUUM has both syntax: with parentheses and without parentheses.
>>> I think we should have both syntax for REINDEX like VACUUM does
>>> because it would be pain to put parentheses whenever we want to do
>>> REINDEX.
>>> Are the parentheses optional in REINDEX command?
>>
>> No. The unparenthesized VACUUM syntax was added back before we
>> realized that that kind of syntax is a terrible idea. It requires
>> every option to be a keyword, and those keywords have to be in a fixed
>> order. I believe the intention is to keep the old VACUUM syntax
>> around for backward-compatibility, but not to extend it. Same for
>> EXPLAIN and COPY.
>
> REINDEX will have only one option VERBOSE for now.
> Even we're in a situation like that it's not clear to be added newly
> additional option to REINDEX now, we should need to put parenthesis?

In my opinion, yes. The whole point of a flexible options syntax is
that we can add new options without changing the grammar. That
involves some compromise on the syntax, which doesn't bother me a bit.
Our previous experiments with this for EXPLAIN and COPY and VACUUM
have worked out quite well, and I see no reason for pessimism here.

> Also I'm not sure that both implementation and documentation regarding
> VERBOSE option should be optional.

I don't know what this means.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Alvaro Herrera 2015-05-05 21:22:22 Re: BRIN range operator class
Previous Message Robert Haas 2015-05-05 20:39:48 Re: Disabling trust/ident authentication configure option