Have `define-key` return the keymap

Have `define-key` return the keymap

[Tianxiang Xiong]

[-- Attachment #1: Type: text/plain, Size: 393 bytes --]

Would there be any issues with / objections to having `define-key` return
the keymap it adds a key to? It currently returns `nil`, making it a
statement (bleh!) rather than a function.

If `define-key` returned the keymap, calls to it could be chained together
in convenient ways, e.g.

```
(thread-first (make-sparse-keymap)
  (define-key (kbd "a") #'foo)
  (define-key (kbd "b") #'bar))
```

[-- Attachment #2: Type: text/html, Size: 574 bytes --]

Re: Have `define-key` return the keymap

[Andreas Schwab]

On Mai 13 2017, Tianxiang Xiong <tianxiang.xiong@gmail.com> wrote:

> Would there be any issues with / objections to having `define-key` return
> the keymap it adds a key to? It currently returns `nil`, making it a
> statement (bleh!) rather than a function.

That would cause major inconvenience when define-key is evaluated with
M-: or ielm.  A keymap is a big structure to print.

Andreas.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
"And now for something completely different."

Re: Have `define-key` return the keymap

[Tianxiang Xiong]

[-- Attachment #1: Type: text/plain, Size: 694 bytes --]

That's true...a valid concern.

On Sat, May 13, 2017 at 1:32 AM, Andreas Schwab <schwab@linux-m68k.org>
wrote:

> On Mai 13 2017, Tianxiang Xiong <tianxiang.xiong@gmail.com> wrote:
>
> > Would there be any issues with / objections to having `define-key` return
> > the keymap it adds a key to? It currently returns `nil`, making it a
> > statement (bleh!) rather than a function.
>
> That would cause major inconvenience when define-key is evaluated with
> M-: or ielm.  A keymap is a big structure to print.
>
> Andreas.
>
> --
> Andreas Schwab, schwab@linux-m68k.org
> GPG Key fingerprint = 58CA 54C7 6D53 942B 1756  01D3 44D5 214B 8276 4ED5
> "And now for something completely different."
>

[-- Attachment #2: Type: text/html, Size: 1237 bytes --]

Re: Have `define-key` return the keymap

[Michael Heerdegen]

Tianxiang Xiong <tianxiang.xiong@gmail.com> writes:

> If `define-key` returned the keymap, calls to it could be chained together in
> convenient ways, e.g.
>
> ```
> (thread-first (make-sparse-keymap)
>   (define-key (kbd "a") #'foo)
>   (define-key (kbd "b") #'bar))
> ```

A different solution would be to define something analog to
`thread-first' that substitutes always the first value as first argument
instead of the subsequent return values (in this case, this is always
the same object).

A variant of this approach that works without such a new macro:

#+begin_src emacs-lisp
(defun my-keybinder (map)
  (apply-partially #'define-key map))

(let ((map (make-sparse-keymap)))
  (cl-flet ((define-key (my-keybinder map)))
    (define-key (kbd "a") #'foo)
    (define-key (kbd "b") #'bar)
    map))
#+end_src


Michael.

RE: Have `define-key` return the keymap

[Drew Adams]

> > If `define-key` returned the keymap, calls to it could be chained together
> > in convenient ways, e.g.
> >
> > (thread-first (make-sparse-keymap)
> >   (define-key (kbd "a") #'foo)
> >   (define-key (kbd "b") #'bar))
> 
> A different solution would be to define something analog to
> `thread-first' that substitutes always the first value as first argument
> instead of the subsequent return values (in this case, this is always
> the same object).
> 
> A variant of this approach that works without such a new macro:
> 
> (defun my-keybinder (map)
>   (apply-partially #'define-key map))
> 
> (let ((map (make-sparse-keymap)))
>   (cl-flet ((define-key (my-keybinder map)))
>     (define-key (kbd "a") #'foo)
>     (define-key (kbd "b") #'bar)
>     map))

I'm probably missing something, but (to me, so far) all of this
has the scent of a solution looking for a problem to solve.

Re: Have `define-key` return the keymap

[Kaushal Modi]

[-- Attachment #1: Type: text/plain, Size: 357 bytes --]

On Sat, May 13, 2017, 10:58 AM Drew Adams <drew.adams@oracle.com> wrote:

>
> I'm probably missing something, but (to me, so far) all of this
> has the scent of a solution looking for a problem to solve.
>

+1

bind-keys from use-package already allows for succinct key bindings to the
same or multiple key maps. Have you tried that OP?

> --

Kaushal Modi

[-- Attachment #2: Type: text/html, Size: 864 bytes --]

Re: Have `define-key` return the keymap

[Stefan Monnier]

> I'm probably missing something, but (to me, so far) all of this
> has the scent of a solution looking for a problem to solve.

Indeed.  If we're concerned about the redundancy in the various
define-key calls for a keymap, then I'd rather we provide something like

    (defvar toto-mode-map (fumble-map map '(("a" foo) ("B" bar))))


-- Stefan