[PATCH] (Draft3) Make X11 server-side fonts and Mule suck less.

[Aidan Kehoe]

 Ar an séiú lá de mí na Samhain, scríobh stephen at xemacs.org: 

 >  > They’re already commented out--they’re just commentary on GNU Emacs’
 >  > approach to implementing Unicode. I’ve replaced them with a clearer
 >  > description of GNU’s approach and how ours differs.
 > 
 > I know they're commented out, but as I said, I've seen code that uses
 > those charsets in the wild.  If and when we get a bug report on that,
 > I'd rather have that code in an obvious place.

Then checking the below into FSF-Compat would be a better approach, I
think. Though the correspondence with their unicode code points is not
exact--that’s something I need to look into and to solve. 

 > Something's weird, then, because in the patch you posted there were a
 > lot of repeated braces at the same indent level, as
 > 
 >     code;
 >   }
 >   }
 > 
 > Are you saying you've since reformatted that to
 >
 >         code;
 >       }
 >   }
 > 
 > ?

Well, I didn’t see the symptom, despite looking at the original, and I
didn’t see the cure, despite looking at the diff, but indent-region should
have taken care of it. Or maybe it’s a question of tab stops. 

Anyway, here’s that code mentioned above: 

;;; BEGIN compat-mule-unicode-charsets.el

(define-ccl-program fsf-compat-ccl-encode-to-ucs-2
  `(1
    ((r1 = (r1 << 8))
     (r1 = (r1 | r2))
     (mule-to-unicode r0 r1)
     (r1 = (r0 >> 8))
     (r2 = (r0 & 255))))
  "CCL program to transform Mule characters to UCS-2.")

(defun fsf-compat-init-mule-unicode-charsets ()
  "Make some Mule character sets that the FSF uses available in XEmacs.

These character sets cover some Unicode code space explicitly; we use a
different solution to the same problem, so you should only need these
character sets if you're editing FSF source.  "
  (let (ku ten charset-symbol)
    (loop
      for (first-ucs last-ucs final) in '((#x0100 #x24FF ?1)
					  (#x2500 #x33ff ?2)
					  (#xE000 #xFFFF ?3))
      do 
      (setq charset-symbol 
	    (intern (format "mule-unicode-%04x-%04x"
			    first-ucs last-ucs)))
      (make-charset charset-symbol
		    (format 
		     "Unicode subset (U+%04X..U+%04X) for FSF compatibility."
		     first-ucs last-ucs)
		    (list 'dimension 2 
			  'registries ["iso10646-1"]
			  'chars 96
			  'columns 1
			  'direction 'l2r
			  'final final
			  'graphic 0
			  'short-name (format "Unicode subset %c" final)
			  'long-name (format "Unicode subset (U+%04X..U+%04X)"
					     first-ucs last-ucs)
			  'ccl-program 'fsf-compat-ccl-encode-to-ucs-2))
      (setq ten 32)

      ;; The names of the character sets lie, at least as of GNU Emacs
      ;; 22.0.50.3. The difference appears to be that they keep assigning
      ;; code points until the end of the 96x96 space of the character sets.
      (loop for ku from 32 to 127 do
	(set-unicode-conversion (make-char charset-symbol ku ten) first-ucs)
	(incf first-ucs)
	(incf ten)
	(when (= ten 128) (setq ten 32))))))

;;; BEGIN compat-mule-unicode-charsets.el

-- 
Santa Maradona, priez pour moi!