\set \e and newline

On Thu, 2022-02-24 at 18:16 +0100, Wim Bertels wrote:

it seems that \set does not interpret an 'enter' interactively the same
as an 'enter' in a short script made with \e

* case 1:
postgres=# \set x 1
postgres=# select :x;
 ?column?
----------
        1

Ok.

* case 2:
postgres=# \e

-- enter the following the editor (the same code as above)

\set x 1
select :x;

-- save and quit

postgres=#

-- no output

This was interpreted as if you had entered it in a single line:

\set x 1 select :x

So "x" is now "1select1;".

-- curiosly: again \e

postgres=#\e

-- shows select 1; in the editor in v14

That was the last executed SQL statement.

-- shows nothing in the editor in v13 (or recursive the content being
cut off)

I don't quite understand what you mean, but the behavior of \e changed
(got more sane) in v14.

variation of case 2:
postgres=# \e

-- enter the following the editor (the same code as above)

\set x 1
select :x;

-- save and quit

postgres=# select :x;
 select1select1
----------------
              1

Now "x" is "1select1select1;;", so you are running

select 1select1select1;;;

Here "select1select1" is interpreted as alias, so you get that column heading.

You can use "\echo :x" to see the value of a variable.

Yours,
Laurenz Albe
--
Cybertec | https://www.cybertec-postgresql.com

Laurenz Albe schreef op vr 25-02-2022 om 10:33 [+0100]:

On Thu, 2022-02-24 at 18:16 +0100, Wim Bertels wrote:
This was interpreted as if you had entered it in a single line:

\set x 1 select :x

So "x" is now "1select1;".

yes

I don't quite understand what you mean, but the behavior of \e
changed
(got more sane) in v14.

yes,
(this is just a summary of different cases)

variation of case 2:
postgres=# \e

-- enter the following the editor (the same code as above)

\set x 1
select :x;

-- save and quit

postgres=# select :x;
 select1select1
----------------
              1

Now "x" is "1select1select1;;", so you are running

 select 1select1select1;;;

Here "select1select1" is interpreted as alias, so you get that column
heading.

You can use "\echo :x" to see the value of a variable.

Thanks for the feedback Laurenz

I guess the main remark is, it not so intuitive that \e behaves
differently then \i

From https://www.postgresql.org/docs/current/app-psql.html
"
\e..
Type semicolon or \g to send it, or \r to cancel it by clearing the
query buffer.

Treating the buffer as a single line primarily affects meta-commands:
whatever is in the buffer after a meta-command will be taken as
argument(s) to the meta-command, even if it spans multiple lines. (Thus
you cannot make meta-command-using scripts this way. Use \i for that.)
"

### case 1: (\e)
\set x 1
select :x ;

### case 2: (\e)
\set x 1
select :x \g

resulting in the same value for x (\echo :x), but different
intermediate output with case 1 and case 2,

"
\g..
If the current query buffer is empty, the most recently sent query is
re-executed instead. Except for that behavior, \g without any arguments
is essentially equivalent to a semicolon.
"

mvg,
Wim

On Mon, 2022-02-28 at 12:34 +0100, Wim Bertels wrote:

I guess the main remark is, it not so intuitive that \e behaves
differently then \i

Ok, but if the differences are well documented, it should be ok.

You could submit a patch that changes the behavior, but people
might be unhappy if established, well documented behavior changes...

Yours,
Laurenz Albe