From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Joe Conway <mail(at)joeconway(dot)com> |
Cc: | snpe <snpe(at)snpe(dot)co(dot)yu>, pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: SQL syntax (column alias) |
Date: | 2002-11-10 17:44:45 |
Message-ID: | 20681.1036950285@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general pgsql-hackers pgsql-jdbc |
Joe Conway <mail(at)joeconway(dot)com> quotes:
> In the SQL92 standard, the optional keyword AS is just noise and can be
> omitted without affecting the meaning. The PostgreSQL parser requires this
> keyword when renaming output columns because the type extensibility features
> lead to parsing ambiguities in this context. AS is optional in FROM items,
> however.
Actually, I think it's not so much datatype extensibility as operator
extensibility, and specifically the fact that we allow postfix
operators. If AS were optional, then
SELECT 1 + x FROM foo;
could be parsed either as "(1 + x)" (infix +, x presumably a column name)
or as "(1 +) x" (postfix +, x an AS-name).
So allowing AS to be optional would at minimum require taking out
postfix operators. There might be other features we'd have to lose,
too; I haven't tried messing with the grammar to see what would happen.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Aurangzeb M. Agha | 2002-11-10 19:16:57 | Upgrading from 7.1.3 to 7.2.3 |
Previous Message | Doug McNaught | 2002-11-10 17:15:09 | Re: Stale Process |
From | Date | Subject | |
---|---|---|---|
Next Message | Michael Meskes | 2002-11-10 17:51:17 | Re: c/sql |
Previous Message | Joe Conway | 2002-11-10 16:38:37 | Re: SQL syntax (column alias) |
From | Date | Subject | |
---|---|---|---|
Next Message | M. A. Sridhar | 2002-11-11 02:55:30 | Help with NPE |
Previous Message | Joe Conway | 2002-11-10 16:38:37 | Re: SQL syntax (column alias) |