Specifying arguments to interactive functions

Specifying arguments to interactive functions

[dont.spam.earl]

Hi all,

I'd like to run multiple comint sessions of the same type in different buffers (e.g. multiple shells, multiple sql connections, etc). I'd like to hit e.g. M-S for shell, then the single digit for the session ID to either create a new session or switch to the existing session with that ID. The following code works, but I'm wondering if I can refactor it to eliminate switch-to-shell and switch-to-sql while still keeping the global-set-key calls simple.

In essence, my problem is that I'd like to bind the keys to a function that interactively elicit arguments (the session #) while also accepting other preset arguments (the buffer-type and buffer-creation function).

The relevant code is included below. Any suggestions on how to simplify this?

Thanks!

Earl


(global-set-key "\M-Q" 'switch-to-sql)
(global-set-key "\M-S" 'switch-to-shell)

(defun switch-to-shell (buffer-id)
  (interactive
   (list (- (read-char-choice
             (format "Enter id for %s buffer: " 'shell)
             '(48 49 50 51 52 53 54 55 56 57))
            48)))
  (message nil)
  (let ((buffer-type "shell")
        (buffer-fn 'shell))
    (switch-to-interactive-buffer '() buffer-id buffer-type buffer-fn)))

(defun switch-to-sql (buffer-id)
  (interactive
   (list (- (read-char-choice
             (format "Enter id for %s buffer: " 'sql)
             '(48 49 50 51 52 53 54 55 56 57))
            48)))
  (message nil)
  (let ((buffer-type "sql")
        (buffer-fn 'sql-connect-fn))
    (switch-to-interactive-buffer '() buffer-id buffer-type buffer-fn)))

(defun switch-to-interactive-buffer (buffer-args buffer-id buffer-type buffer-fn)
  "Switch to the interactive buffer of the specified ID.
ID is elicited from user in minibuffer as a single digit.
Looks for buffer type name (e.g. \"shell\") in *interactive-buffer-type*.
Looks for buffer creation function (e.g. 'shell) in *interactive-buffer-fn*.
"
  ;;(message "in buffer-name: %s" (buffer-name))
  (let* ((buffer-name (format "*%s-%d*" buffer-type buffer-id))
         (buffer (or (get-buffer buffer-name)
                     (let ((new-buffer-args (append buffer-args (list buffer-name))))
                       (apply buffer-fn new-buffer-args)))))
    ;;(message "switching to buffer %s" buffer)
    (switch-to-buffer buffer)
    ;;(message "switched to buffer: %s" (buffer-name))
    ))

Re: Specifying arguments to interactive functions

[Thien-Thi Nguyen]

[-- Attachment #1: Type: text/plain, Size: 830 bytes --]

() dont.spam.earl@gmail.com
() Wed, 14 May 2014 22:20:51 -0700 (PDT)

   Any suggestions on how to simplify this?

                '(48 49 50 51 52 53 54 55 56 57)

These are code points for ASCII "0" through "9", and are already as
simple as possible, in one sense, but not in the "wait six months and
re-grok quickly the code" sense.  Better to be symbolic (not necessarily
w/ Lisp symbols, but rather w/ human-friendly expressions).

For example, you can:
- compress:       (number-sequence 48 57)
- use characters: (number-sequence ?0 ?9)
- go up a level:  (interactive "P") + SMOP

-- 
Thien-Thi Nguyen
   GPG key: 4C807502
   (if you're human and you know it)
      read my lisp: (responsep (questions 'technical)
                               (not (via 'mailing-list)))
                     => nil

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: Specifying arguments to interactive functions

[dont.spam.earl]

On Thursday, May 15, 2014 1:04:01 AM UTC-7, Thien-Thi Nguyen wrote:
> () dont.spam.earl@gmail.com
> 
> () Wed, 14 May 2014 22:20:51 -0700 (PDT)
> 
> 
> 
>    Any suggestions on how to simplify this?
> 
> 
> 
>                 '(48 49 50 51 52 53 54 55 56 57)
> 
> 
> 
> These are code points for ASCII "0" through "9", and are already as
> 
> simple as possible, in one sense, but not in the "wait six months and
> 
> re-grok quickly the code" sense.  Better to be symbolic (not necessarily
> 
> w/ Lisp symbols, but rather w/ human-friendly expressions).
> 
> 
> 
> For example, you can:
> 
> - compress:       (number-sequence 48 57)
> 
> - use characters: (number-sequence ?0 ?9)
> 
> - go up a level:  (interactive "P") + SMOP
> 
> 
> 
> -- 
> 
> Thien-Thi Nguyen
> 
>    GPG key: 4C807502
> 
>    (if you're human and you know it)
> 
>       read my lisp: (responsep (questions 'technical)
> 
>                                (not (via 'mailing-list)))
> 
>                      => nil

Thanks. Any tips on passing an argument to 'switch-to-interactive-buffer while having it still call 'interactive to get the digit from the user?

Re: Specifying arguments to interactive functions

[Thien-Thi Nguyen]

[-- Attachment #1: Type: text/plain, Size: 675 bytes --]

() dont.spam.earl@gmail.com
() Thu, 15 May 2014 10:43:56 -0700 (PDT)

   Any tips on passing an argument to 'switch-to-interactive-buffer
   while having it still call 'interactive to get the digit from the
   user?

First, ‘C-h f interactive’ and also (info "(elisp) Using Interactive").
Next, experiment.  Next, read some source.  Next, post questions on the
discrepencies and lacunae you might (think you) experience.  That's it.

-- 
Thien-Thi Nguyen
   GPG key: 4C807502
   (if you're human and you know it)
      read my lisp: (responsep (questions 'technical)
                               (not (via 'mailing-list)))
                     => nil

[-- Attachment #2: Type: application/pgp-signature, Size: 197 bytes --]

Re: Specifying arguments to interactive functions

[dont.spam.earl]

On Thursday, May 15, 2014 11:06:24 AM UTC-7, Thien-Thi Nguyen wrote:

> 
> First, 'C-h f interactive' and also (info "(elisp) Using Interactive").
> 
> Next, experiment.  Next, read some source.  Next, post questions on the
> 
> discrepencies and lacunae you might (think you) experience.  That's it.


Well, even after looking at the docs, I'm still looking for an answer to the same question: Is there a way to call interactive to elicit only _some_ of the arguments of a function.

Alternately, if there's a standard workaround for this, I'd be interested in that too.