Re: [BUGS] ORDER BY $1 behaves inconsistently - Mailing list pgsql-bugs

From Jordan Lewis
Subject Re: [BUGS] ORDER BY $1 behaves inconsistently
Date
Msg-id CAALgziJKJgU=UP7FSfyLvHGKEJ2O=zjYCoLVUcJnTu+7rkD_SA@mail.gmail.com
Whole thread Raw
In response to Re: [BUGS] ORDER BY $1 behaves inconsistently  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-bugs
Thanks for the explanation - that makes sense.

On Fri, Oct 27, 2017 at 2:33 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
Jordan Lewis <jordanthelewis@gmail.com> writes:
> Actually, it's even worse than I thought. It seems that placeholders in
> ORDER BY clauses get entirely ignored,
> as running `EXECUTE x(1)` on the previous example with unsorted table data
> does not sort the table by the 1st
> column as expected.

"ORDER BY $1" is asking to order by some constant value (constant within
any one execution of the command, anyway), which is useless, because all
rows in the query will have the same sort key.  The planner will throw
that away as being a no-op.

"ORDER BY 1" is asking to order by the first output column.  This is
not the same thing.  You cannot get that effect with a parameter;
if you could, it would probably represent a SQL-injection hazard.

If the value of $1 were "foo" and you complained that it didn't order
by column foo, it would be exactly the same issue...

                        regards, tom lane

pgsql-bugs by date:

Previous
From: Cristiano Coelho
Date:
Subject: [BUGS] pg_trgm word_similarity inconsistencies or bug
Next
From: Tom Lane
Date:
Subject: Re: [BUGS] BUG #14874: Dublicate values in primary key