Re [HACKERS]: Still not happy with psql's multiline history behavior - Mailing list pgsql-patches

From Sergey E. Koposov
Subject Re [HACKERS]: Still not happy with psql's multiline history behavior
Date
Msg-id Pine.LNX.4.64.0606040807340.27473@lnfm1.sai.msu.ru
Whole thread Raw
List pgsql-patches
Hello,

I'm sending the patch fixing the \r bug:

On Wed, 31 May 2006, Tom Lane wrote:

> The example that seriously sucks is:
>
> regression=# select foo ...
> regression-# \r
> Query buffer reset (cleared).
> now control-P brings back:
> regression=# select foo ...
> \r

In my fix the currently-written query is completely erased if \r is
invoked. I think that's the most logic behavior.
So, the example of work:

wsdb=# select 1;
  ?column?
----------
         1
(1 row)

wsdb=# select
wsdb-# \r
Query buffer reset (cleared).
wsdb=#
Control-P now brings back:
wsdb=# \r

Another Control-P brings back:
wsdb=# select 1;


> Also, \e is seriously broken: after you edit the text and exit the
> editor, the text is not reloaded into the visible display (although it
> does seem to still be "behind the scenes" somewhere).

For that part, I currently do not see the possible fix, and I don't even
think that it should be fixed, because there with current multi-line
implementation there is no need for that.

For example, when we do
wsdb=# select 1+
wsdb-#
And press Control-P , the string "select 1+" is not in the history, and
there is no need for that.

And in the case of \e editing, if in the external editor we have typed
the incomplete query like "select 1+", the situation will be exactly the
same and for the same reason...

So, I don't think, that this require fixing. In principal, probably it is
possible to fix that in a different way:
Suppose we executed the \e command and in the external editor we typed
"select 1 +". And after the exit from the editor, psql produce
wsdb=# select 1 +
(and put cursor in the end of the line).

That can be reasonable (I'm not sure that it's possible to do that with
readline, but should be).
But I would consider that as a new feature (probably not very useful), not
as a bug fix.

Regards,
     Sergey

*******************************************************************
Sergey E. Koposov
Max Planck Institute for Astronomy/Sternberg Astronomical Institute
Tel: +49-6221-528-349      Web: http://lnfm1.sai.msu.ru/~math
E-mail: math@sai.msu.ru

Attachment

pgsql-patches by date:

Previous
From: "Sergey E. Koposov"
Date:
Subject: Re: Re [HACKERS]: Still not happy with psql's multiline history
Next
From: Bruce Momjian
Date:
Subject: Re: Re [HACKERS]: Still not happy with psql's multiline history behavior