bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Drew Adams]

In isearch.el, we have this:
(define-key map "\M-\t" 'isearch-complete)
 
We should also have this, so `ESC TAB' works too:
(define-key map [escape tab] 'isearch-complete)

In GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600)
 of 2010-06-07 on 3249CTO
Windowing system distributor `Microsoft Corp.', version 5.1.2600
configured using `configure --with-gcc (4.4) --no-opt --cflags -Ic:/xpm/include'
 

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Juri Linkov]

> In isearch.el, we have this:
> (define-key map "\M-\t" 'isearch-complete)
>
> We should also have this, so `ESC TAB' works too:
> (define-key map [escape tab] 'isearch-complete)

I wonder why `ESC C-i' works in isearch, but `ESC TAB' doesn't work.
(There is no special key binding for `ESC C-i' in isearch.)

-- 
Juri Linkov
http://www.jurta.org/emacs/

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Drew Adams]

> > In isearch.el, we have this:
> > (define-key map "\M-\t" 'isearch-complete)
> >
> > We should also have this, so `ESC TAB' works too:
> > (define-key map [escape tab] 'isearch-complete)
> 
> I wonder why `ESC C-i' works in isearch, but `ESC TAB' doesn't work.
> (There is no special key binding for `ESC C-i' in isearch.)

See this comment in the code:

    ;; To handle local bindings with meta char prefix keys, define
    ;; another full keymap.  This must be done for any other prefix
    ;; keys as well, one full keymap per char of the prefix key.  It
    ;; would be simpler to disable the global keymap, and/or have a
    ;; default local key binding for any key not otherwise bound.
    (let ((meta-map (make-sparse-keymap)))
      (define-key map (char-to-string meta-prefix-char) meta-map)
      (define-key map [escape] meta-map))
    (define-key map (vector meta-prefix-char t) 'isearch-other-meta-char)

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Stefan Monnier]

> I wonder why `ESC C-i' works in isearch, but `ESC TAB' doesn't work.
> (There is no special key binding for `ESC C-i' in isearch.)

It probably has to do with the fact that isearch uses default-bindings
(i.e. bindings to the event gt which stands for "any event").
Such bindings prevent function-key-map (and similar hardcoded key
translations such as A -> a, S-left -> left, double-mouse-1 -> mouse-1,
...) from doing its job (since function-key-map is only used if the
untranslated sequence has no bindings which never happens thanks to the
default binding which always applies).

I've tried to change read-key-sequence such that a default binding is
only applied *after* function-key-map (or at least that a default
binding does not prevent function-key-map translations), but so far this
has failed.


        Stefan

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Alex]

"Drew Adams" <drew.adams@oracle.com> writes:

> In isearch.el, we have this:
> (define-key map "\M-\t" 'isearch-complete)
>  
> We should also have this, so `ESC TAB' works too:
> (define-key map [escape tab] 'isearch-complete)
>
> In GNU Emacs 24.0.50.1 (i386-mingw-nt5.1.2600)
>  of 2010-06-07 on 3249CTO
> Windowing system distributor `Microsoft Corp.', version 5.1.2600
> configured using `configure --with-gcc (4.4) --no-opt --cflags -Ic:/xpm/include'
>  

ESC TAB works for me since at least 24.4. Does it work for you now?

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Drew Adams]

> > In isearch.el, we have this:
> > (define-key map "\M-\t" 'isearch-complete)
> >
> > We should also have this, so `ESC TAB' works too:
> ESC TAB works for me since at least 24.4. Does it work for you now?

Yes.  The bug was fixed in 24.4.  Earlier versions had an explicit binding of <escape> TAB to `isearch-complete', but ESC TAB did not work.

bug#6387: 24.0.50; ESC TAB should be bound to `isearch-complete'

[Nicolas Petton]

[-- Attachment #1: Type: text/plain, Size: 154 bytes --]

Drew Adams <drew.adams@oracle.com> writes:

close 6387
thanks

> Yes.  The bug was fixed in 24.4.

Great, I'm closing this bug report then.

Cheers,
Nico

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 512 bytes --]