To: "Radu-Adrian Popescu" <radu.popescu@aldratech.com>
Cc: pgsql-sql@postgresql.org
Subject: Re: SQL function parse error ? 
In-reply-to: <00a501c2b7dc$a5c626b0$0600a8c0@rpopescu> 
References: <008f01c2b7d0$3733fdf0$0600a8c0@rpopescu>
	<3E1D6362.80701@klaster.net>
	<00a501c2b7dc$a5c626b0$0600a8c0@rpopescu>
Comments: In-reply-to "Radu-Adrian Popescu" <radu.popescu@aldratech.com>
	message dated "Thu, 09 Jan 2003 14:43:02 +0200"
Date: Thu, 09 Jan 2003 09:48:14 -0500
Message-ID: <29440.1042123694@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>

"Radu-Adrian Popescu" <radu.popescu@aldratech.com> writes:
> This is SQL, and people who are using PostgreSql write SQL, not
> whitespace-sensitive SQL, bash or whatever.

Nonsense.  SQL syntax is space-sensitive.  Or have you successfully
written
	SELECTXFROMY;
lately?

There has occasionally been talk of disallowing '$' as a valid character
in operator names, which would eliminate the syntactic ambiguity in this
example.  But undoubtedly it would also break a few applications that
use '$' in user-defined operator names, so the proposal hasn't passed
to date.

> Is there any operator named >$ ?

Whether there is one in the standard distribution is quite irrelevant.
It's a valid operator name according to the current rules, and so open
to definition by anyone who wants to.

The most recent discussion I can find about this is the pgsql-hackers
thread "Dollar in identifiers" from Aug 2001, eg
http://archives.postgresql.org/pgsql-hackers/2001-08/msg00629.php
There didn't seem to be a consensus to change things, so the old
behavior stands, for the moment.

			regards, tom lane