Converting string to symbol

Converting string to symbol

[excalamus--- via Users list for the GNU Emacs text editor]

Hi, 

I want to define a global abbrev with case-fixed property t.  I want this abbrev to appear in the define-abbrev-table definition created by write-abbrev-file.

To my knowledge, there is no native option to do this.  So, I have defined a function which wraps define-abbrev and reads in the table, abbrev name, and expansion from the user.

(defun my-add-case-fixed-abbrev (name expansion &optional table)
  "Add abbrev with case-fixed t property."
  (interactive
   (let ((table (intern-soft (completing-read
          "Abbrev table (global-abbrev-table): "
          abbrev-table-name-list nil t nil nil "global-abbrev-table" )))
         (name (read-string "Abbrev name: "))
         (expansion (read-string "Expansion: ")))
     (list name expansion table)))
  (let ((table (or table global-abbrev-table)))
    (define-abbrev table name expansion nil :case-fixed t)))

Since completing-read returns a string, I convert the string representing the abbrev table to the corresponding symbol using intern-soft.  

However, this causes the error "Wrong-type-argument vectorp global-abbrev-table" when calling explicitly or interactively.

 (my-add-case-fixed-abbrev "my-abbrev" "my-expansion" "global-abbrev-table") ;; causes error

The error doesn't appear when the default argument is used

(my-add-case-fixed-abbrev "my-abbrev" "my-expansion")  ;; uses default arg, no error

I can see that (eq 'global-abbrev-table (intern-soft "global-abbrev-table")) returns t.  How does the symbol created by intern-soft differ from the symbol 'global-abbrev-table' used as the default within let?

I can see that (vectorp (intern-soft "global-abbrev-table")) is nil, so fair enough, table is not a vector.  

If I convert table to a vector using

   (let ((table (vector (intern-soft (completing-read
          "Abbrev table (global-abbrev-table): "
          abbrev-table-name-list nil t nil nil "global-abbrev-table" ))))

I get the error "Wrong type argument: number-or-marker-p, nil".

I have run each of these with debug-on-error set, but I'm not sure how to interpret the error message.   

Any hint on what's going on here?

Re: Converting string to symbol

[Thien-Thi Nguyen]

[-- Attachment #1: Type: text/plain, Size: 706 bytes --]


() excalamus--- via <help-gnu-emacs@gnu.org>
() Sun, 3 Jan 2021 21:54:32 +0100 (CET)

   Any hint on what's going on here?

The name of the thing is not the thing.  You don't want the
symbol ‘global-abbrev-table’, you want the object that that
symbol references.  See ‘interactive’ form, specifically ‘X’.

-- 
Thien-Thi Nguyen -----------------------------------------------
 (defun responsep (query)               ; (2021) Software Libero
   (pcase (context query)               ;       = Dissenso Etico
     (`(technical ,ml) (correctp ml))
     ...))                              748E A0E8 1CB8 A748 9BFA
--------------------------------------- 6CE4 6703 2224 4C80 7502


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 215 bytes --]

Re: Converting string to symbol

[excalamus--- via Users list for the GNU Emacs text editor]

I see, thank you very much.

Ideally, I would like global-abbrev-table to be the default.  With (interactive X...) I don't see how to do that.  

When I run M-x my-test <RET>, I get read--expression: End of file during parsing.  The function halts immediately and doesn't evaluate the body.  I'm not sure how I could catch that.  

(defun my-test (expr)
  "test interactive X behavior."
  (interactive  "XLisp expression: ")
  (message "before compare")  (if expr
      (message "%s" expr)
    (message "is nil"))
  (message "after compare"))

I assume read--expression is being used within (interactive X...).  It's not documented and I didn't have much luck trying to use it (not detailing those efforts here since it's really not meant for "public" use).

Re: Converting string to symbol

[Emanuel Berg via Users list for the GNU Emacs text editor]

excalamus wrote:

> When I run M-x my-test <RET>, I get read--expression: End of
> file during parsing. The function halts immediately and
> doesn't evaluate the body. I'm not sure how I could
> catch that.
>
> (defun my-test (expr)
>  "test interactive X behavior."
>  (interactive "XLisp expression: ")
>  (message "before compare") (if expr
>  (message "%s" expr)
>  (message "is nil"))
>  (message "after compare"))
>
> I assume read--expression is being used within (interactive
> X...). It's not documented and I didn't have much luck
> trying to use it (not detailing those efforts here since
> it's really not meant for "public" use).

I don't understand a lot but

(defun print-lisp (expr)
  (interactive "xlisp: ")
  (message "lisp: %s" expr) )

M-x print-lisp RET (+ 1 1) RET

"lisp: (+ 1 1)"

-- 
underground experts united
http://user.it.uu.se/~embe8573
https://dataswamp.org/~incal

Re: Converting string to symbol

[excalamus--- via Users list for the GNU Emacs text editor]

I got it.  Simply grab the symbol's value once it's been interned.  Thanks again for your help :)
(defun my-add-case-fixed-abbrev (name expansion &optional table)
  "Add fixed case abbrev with NAME and EXPANSION into TABLE.

TABLE is optional; defaults to `global-abbrev-table'."
  (interactive
   (let ((table (symbol-value (intern-soft (completing-read
          "Abbrev table (global-abbrev-table): "
          abbrev-table-name-list nil t nil nil "global-abbrev-table"))))
         (name (read-string "Abbrev name: "))
         (expansion (read-string "Expansion: ")))
     (list name expansion table)))
  (let ((table (or table global-abbrev-table)))
    (define-abbrev table name expansion nil :case-fixed t)))

Re: Converting string to symbol

[Andreas Röhler]

On 03.01.21 21:54, excalamus--- via Users list for the GNU Emacs text 
editor wrote:
> Hi,
>
> I want to define a global abbrev with case-fixed property t.

AFAIK Emacs never was able to store and use case-fixed abbreviations.

For example function add-abbrev  closes with the following:

(define-abbrev table (downcase name) exp)

Re: Converting string to symbol

[Stefan Monnier]

> AFAIK Emacs never was able to store and use case-fixed abbreviations.

Where "never" means "not before Emacs-24"?


        Stefan