bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Drew Adams]

Enhancement request.  Keys shown by `describe-bindings' and similar
(including `substitute-command-keys' for \{...}) currently show `??' as
their binding if the binding is not a command symbol.

Please turn the `??' occurrences into links that when clicked show the
source code for the key's binding, when available.  If it is
byte-compiled, and the uncompiled code is unknown, show the byte code,
but with control chars filtered out (so somewhat human-readable).

This lets users see what is behind the `??', giving them a somewhat
better idea what the key does.

In GNU Emacs 25.1.50.1 (i686-pc-mingw32)
 of 2015-12-10
Repository revision: 6148555ee5a3d0139ae517803718b3e0357933c7
Windowing system distributor 'Microsoft Corp.', version 6.1.7601
Configured using:
 'configure --prefix=/c/Devel/emacs/snapshot/trunk --enable-checking=yes
 --enable-check-lisp-object-type --without-compress-install 'CFLAGS=-Og
 -ggdb3' LDFLAGS=-Lc:/Devel/emacs/lib 'CPPFLAGS=-DGC_MCHECK=1
 -Ic:/Devel/emacs/include''

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Lars Ingebrigtsen]

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

> Enhancement request.  Keys shown by `describe-bindings' and similar
> (including `substitute-command-keys' for \{...}) currently show `??' as
> their binding if the binding is not a command symbol.
>
> Please turn the `??' occurrences into links that when clicked show the
> source code for the key's binding, when available.  If it is
> byte-compiled, and the uncompiled code is unknown, show the byte code,
> but with control chars filtered out (so somewhat human-readable).
>
> This lets users see what is behind the `??', giving them a somewhat
> better idea what the key does.

The calling sequence for this command is slightly convoluted, but it
turns out that help.el calls basically

  (describe-buffer-bindings buffer nil nil)

which is a C function (but why?) which inserts all the

C-n		next-line
C-o		open-line
C-H-w		??

stuff, which is then post-processed to add the xref links.  So at the
point where the xref is added, then there's no feasible way to determine
what the ?? referred to, really, without redoing the work the C function
had already done...

I think.

Anybody?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Drew Adams]

> The calling sequence for this command is slightly convoluted, but it
> turns out that help.el calls basically
>   (describe-buffer-bindings buffer nil nil)
> which is a C function (but why?) which inserts all the
> 
> C-n		next-line
> C-o		open-line
> C-H-w		??
> 
> stuff, which is then post-processed to add the xref links.  So at the
> point where the xref is added, then there's no feasible way to
> determine what the ?? referred to, really, without redoing the work the C
> function had already done...
> 
> I think.
> 
> Anybody?

I don't have a solution suggestion.  But maybe
a start would be to move this from C to Lisp?
(Just a suggestion.  Dunno whether there's a
good reason why it's still in C.)

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Lars Ingebrigtsen]

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

> Please turn the `??' occurrences into links that when clicked show the
> source code for the key's binding, when available.  If it is
> byte-compiled, and the uncompiled code is unknown, show the byte code,
> but with control chars filtered out (so somewhat human-readable).

This is now fixed in Emacs 29.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Juri Linkov]

>> Please turn the `??' occurrences into links that when clicked show the
>> source code for the key's binding, when available.  If it is
>> byte-compiled, and the uncompiled code is unknown, show the byte code,
>> but with control chars filtered out (so somewhat human-readable).
>
> This is now fixed in Emacs 29.

One minor problem is that [closure] and [lambda] buttons
have no 'help-echo' and no 'mouse-face' with 'highlight'
like other buttons in 'C-h b'.

bug#24235: 25.1.50; `describe-bindings' request: show `??' definitions

[Lars Ingebrigtsen]

Juri Linkov <juri@linkov.net> writes:

> One minor problem is that [closure] and [lambda] buttons
> have no 'help-echo' and no 'mouse-face' with 'highlight'
> like other buttons in 'C-h b'.

I've now fixed the mouse-face...  but...  I think we kinda go overboard
on the help-echo stuff: It's a button, so RET/mouse is what works, and I
don't think anybody needs to be told?

But I guess it could say what it's doing of you click on it.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no