Re: No error when column doesn't exist - Mailing list pgsql-general

From Tom Lane
Subject Re: No error when column doesn't exist
Date
Msg-id 1902.1221067706@sss.pgh.pa.us
Whole thread Raw
In response to No error when column doesn't exist  (Dean Rasheed <dean_rasheed@hotmail.com>)
Responses Re: No error when column doesn't exist  (Artacus <artacus@comcast.net>)
Re: No error when column doesn't exist  (Dean Rasheed <dean_rasheed@hotmail.com>)
List pgsql-general
Dean Rasheed <dean_rasheed@hotmail.com> writes:
> CREATE TABLE foo(a int, b int);
> INSERT INTO foo VALUES(1,2);
> SELECT foo.text FROM foo;

> I expected that to generate an error: column foo.text does not exist.
> Instead it treats "foo.text" as "foo.*::text AS text":
> Is this a feature or a bug?

Hmm.  It's a feature, but maybe a dangerous one.  The expression is
being treated as text(foo), which is intentional in order to allow
use of functions as if they were virtual columns.  However, then it
decides that what you've got there is a cast request.  There wasn't
any ability to cast composite types to text before 8.3, so this fails
in the expected way in 8.2 and before; but in 8.3 the cast
interpretation succeeds, and away we go.

foo.char and foo.varchar have similarly unexpected behavior; I think
that's probably the end of it, though, since those are the only types
that CoerceViaIO will take as targets.

Maybe we could/should restrict things so that the syntax continues to
fail, but I can't think of any restrictions that don't seem like warts.
What's worse, they might break stuff that used to work.

            regards, tom lane

pgsql-general by date:

Previous
From: "Markova, Nina"
Date:
Subject: Re: FW: How to upload data to postgres
Next
From: Zdenek Kotala
Date:
Subject: Re: 64-bit Compile Failure on Solaris 10 with OpenSSL