docstring: mail-signature

docstring: mail-signature

[T.V Raman]

from sendmail.el:
Note that the docstring says mail-signature determines what happens,
but also says that the .signature file is used if this var is t or
nil.

Shouldn't the sign not happen if that var is nil?
The code actually checks for (memq t nil)

(defun mail-signature (&optional atpoint)
  "Sign letter with signature.
If the variable `mail-signature' is a string, inserts it.
If it is t or nil, inserts the contents of the file `mail-signature-file'.
Otherwise, evals `mail-signature'.
Prefix argument ATPOINT means insert at point rather than the end."
  (interactive "*P")
  ;; Test for an unreadable file here, before we delete trailing
  ;; whitespace, so that we don't modify the buffer needlessly.
  (if (and (memq mail-signature '(t nil))
	   (not (file-readable-p mail-signature-file)))
      (if (called-interactively-p 'interactive)
	  (message "The signature file `%s' could not be read"
		   mail-signature-file))
    (save-excursion
      (unless atpoint
	(goto-char (point-max))
	;; Delete trailing whitespace and blank lines.
	(skip-chars-backward " \t\n")
	(end-of-line)
	(delete-region (point) (point-max)))
      (cond ((stringp mail-signature)
	     (insert mail-signature))
	    ((memq mail-signature '(t nil))
	     (insert "\n\n-- \n")
	     (insert-file-contents (expand-file-name mail-signature-file)))
	    (t
	     ;; FIXME add condition-case error handling?
	     (eval mail-signature))))))


-- 

Thanks,

--Raman(I Search, I Find, I Misplace, I Research)
♉ Id: kg:/m/0285kf1  🦮

--

Thanks,

--Raman(I Search, I Find, I Misplace, I Research)
♉ Id: kg:/m/0285kf1  🦮

Re: docstring: mail-signature

[Emanuel Berg]

T.V Raman wrote:

> -- 
>
> Thanks,
>
> --Raman(I Search, I Find, I Misplace, I Research)
> ♉ Id: kg:/m/0285kf1  🦮

Something is sure up since your signature appeared twice.

-- 
underground experts united
https://dataswamp.org/~incal

Re: docstring: mail-signature

[Eli Zaretskii]

> From: "T.V Raman" <raman@google.com>
> Date: Sat, 2 Sep 2023 18:58:09 -0700
> 
> from sendmail.el:
> Note that the docstring says mail-signature determines what happens,
> but also says that the .signature file is used if this var is t or
> nil.
> 
> Shouldn't the sign not happen if that var is nil?
> The code actually checks for (memq t nil)

Yes, it seems to be a bug (because the doc string of the user option
mail-signature says nil means no signature is inserted, even if
mail-signature-file exists and is readable.  But it's a very old bug,
and the code installed 14 years ago by Glenn (CC'ed) explicitly
handles nil and t the same, not sure why.

Glenn, any comments?

Re: docstring: mail-signature

[T.V Raman]

Emanuel Berg <incal@dataswamp.org> writes:

That was the issue I was trying to trace down (duplicate sig) when I
spotted that doc bug.

> T.V Raman wrote:
>
>> -- 
>>
>> Thanks,
>>
>> --Raman(I Search, I Find, I Misplace, I Research)
>> ♉ Id: kg:/m/0285kf1  🦮
>
> Something is sure up since your signature appeared twice.

-- 

Thanks,

--Raman(I Search, I Find, I Misplace, I Research)
♉ Id: kg:/m/0285kf1  🦮