RE: buff-menu.el in CVS today has utf-8 characters?

[Drew Adams]

My bad. I clicked the download link, which opened the file in my browser (IE); 
I then selected the text and pasted it into Emacs.

When, instead, I right-click to Save Target As, I get the file with em dashes.

I wonder if other users might do the same as I.

I byte-compiled in Emacs 20, for compatibility, and got this warning: 

While compiling list-buffers-noselect:
  ** reference to free variable \200\224

If I byte-compile in Emacs 22 I don't get the warning. I want to adapt the file 
(locally) for Emacs 20 too. Any idea what I'll need to do wrt "free variable 
\200\224"? I guess that the byte-compiler is interpreting the UTF em-dash 
character as a variable named \200\244 following character ?\342 when it sees a 
character code like this: ?\342\200\224. 

What's the best way to prevent execution of the Emacs 22 code for this 
character? I know you will say that testing the Emacs version is not the way to 
go. What's a better test to use here?

Thanks,

  Drew

    -----Original Message-----
    From: Stefan Monnier [mailto:address@hidden
    Sent: Tuesday, June 14, 2005 3:08 PM
    To: Drew Adams
    Cc: Emacs-Pretest-Bug
    Subject: Re: buff-menu.el in CVS today has utf-8 characters?
    
    
    > In the minibuffer, it says "Select coding system (default 
    mule-utf-8):"
    > Point is on this in buff-menu.el: –-Stef
    > Questions:
    > Is it normal to have these characters in the file?
    
    I don't see them here.  Are they present in the buff-menu.el 
    buffer as well?
    If not, how do you copy&paste from buff-menu.el to foo.el?
    
    > Does this file need to be utf-8? Should it be?
    
    Well, there are a few other coding systems possible, but it contains
    a non-ASCII non-latin-1 char (more specifically an EM DASH, U+2014).
    
    > What coding system should I use to save this file? I used utf-8.
    
    Good choice.
    
    > The first line of the buffer has this - shouldn't it force utf-8
    > encoding automatically?  -*- coding:utf-8 -*-
    
    I agree it should force utf-8 automatically.  I use the patch below for
    that precise purpose.
    
    
            Stefan
    
    
    --- orig/lisp/international/mule-cmds.el
    +++ mod/lisp/international/mule-cmds.el
    @@ -854,6 +855,24 @@
          (mapcar (function (lambda (x) (cons x (coding-system-base x))))
                  default-coding-system))
     
    +    (unless no-other-defaults
    +      ;; If the file has a coding cookie, try to use it before anything
    +      ;; else (i.e. before default-coding-system which will 
    typically come
    +      ;; from file-coding-system-alist).
    +      (unless (or (stringp from) find-file-literally)
    +        (let* ((auto-cs (save-excursion
    +                          (save-restriction
    +                            (widen)
    +                            (narrow-to-region from to)
    +                            (goto-char (point-min))
    +                            (set-auto-coding (or file 
    buffer-file-name "")
    +                                             (buffer-size)))))
    +               (base (if auto-cs (coding-system-base auto-cs))))
    +          (or (memq base '(nil undecided))
    +              (rassq base default-coding-system)
    +              (push (cons auto-cs base)
    +                    default-coding-system)))))
    +
         ;; From now on, the list of defaults is reversed.
         (setq default-coding-system (nreverse default-coding-system))