Re: Common table expression - parsing questions

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Thomas Kellerer <spam_eater(at)gmx(dot)net>
Cc: pgsql-sql(at)postgresql(dot)org
Subject: Re: Common table expression - parsing questions
Date: 2009-10-04 15:00:36
Message-ID: 8813.1254668436@sss.pgh.pa.us
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-sql

Thomas Kellerer <spam_eater(at)gmx(dot)net> writes:
> the6campbells wrote on 29.09.2009 04:54:
>> 2. Do you intend to remove the requirement to include the recursive
>> keyword - as other vendors allow

> The standard *requires* the keyword.

The reason the standard requires the keyword is that the WITH-name
scoping rules are different in the two cases. This means it is possible
to construct 100% valid queries that mean different things depending on
whether RECURSIVE is present or not. Admittedly, they'd be uncommon
corner cases, but the short answer is that an implementation that does
not require the keyword will fail to process some standard-conforming
queries in a standard-conforming way.

IOW, no, we are not going to change this.

regards, tom lane

In response to

Browse pgsql-sql by date

  From Date Subject
Next Message Osvaldo Kussama 2009-10-04 18:37:40 Re: reading last inserted record withoud any autoincrement field
Previous Message mohammad qoreishy 2009-10-04 14:49:54 reading last inserted record withoud any autoincrement field