list charset stuff needs buffer name other than *Help*

list charset stuff needs buffer name other than *Help*

[Dan Jacobson]

It is very bad for the user to have gone thru the trouble of going
thru the help menus till he finally gets to seeing the table
"Characters in the charset chinese-big5-1."  and then have that all be
wiped away when he executes the next help command.  Therefore that
character set stuff should be put in a different buffer than *Help*.
Indeed, the buffer that starts with

Use <mouse-2> or C-c C-c:
  on a column title to sort by that title,
  on a charset name to list characters.

should already have a different name than *Help*

By the way, (after changing the name of the buffer, of course), I did
C-u C-x = on a certain char. and found that none of the many numbers
below was like the one on that row, "217x", in the "Characters in the
charset chinese-big5-1" table.

  character: [i took it out of this mail] (0450362, 151794, 0x250f2)
    charset: chinese-big5-1 (Frequently used part (A141-C67F) of Big5 (Chinese traditional))
 code point: 33 114
buffer code: 0x98 0xA1 0xF2
  file code: 0xA1 0xB3 (encoded by coding system chinese-big5-unix)
-- 
http://jidanni.org/ Taiwan(04)25854780

Re: list charset stuff needs buffer name other than *Help*

[Kenichi Handa]

In article <87k7ldtt8f.fsf@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
> It is very bad for the user to have gone thru the trouble of going
> thru the help menus till he finally gets to seeing the table
> "Characters in the charset chinese-big5-1."  and then have that all be
> wiped away when he executes the next help command.  Therefore that
> character set stuff should be put in a different buffer than *Help*.
> Indeed, the buffer that starts with

> Use <mouse-2> or C-c C-c:
>   on a column title to sort by that title,
>   on a charset name to list characters.

> should already have a different name than *Help*

Thank you for a good suggestion.  I've just installed the
attached changes.

> By the way, (after changing the name of the buffer, of course), I did
> C-u C-x = on a certain char. and found that none of the many numbers
> below was like the one on that row, "217x", in the "Characters in the
> charset chinese-big5-1" table.

chinese-big5-1/2 are the charsets invented by Emacs to
support Big5.  Big5 characters are mapped to these charsets
in Emacs.  Thus, 217x are the code-points of chinese-big5-1/2,
not those of Big5 itself.

If you want to see the real code-points of Big5, please do:
M-x list-charset-chars RET big5 RET.

By the way, I made M-x list-character-sets to list
also indirectly supported charsets such as Big5.

---
Ken'ichi HANDA
handa@etl.go.jp

2002-09-25  Kenichi Handa  <handa@etl.go.jp>

	* international/mule-diag.el (list-character-sets): Use the buffer
	name "*Character Set List*", not "*Help*".  List also indirectly
	supported character sets.
	(list-charset-chars): Use the buffer name "*Character List*", not
	"*Help*".  Display the current charset name in the modeline.
	(non-iso-charset-alist): Add mapped charset list for `mac-roman'.
	(sort-listed-character-sets): Don't alter the region showing
	indirectly supported charsets.

Index: mule-diag.el
===================================================================
RCS file: /cvs/emacs/lisp/international/mule-diag.el,v
retrieving revision 1.79
diff -u -c -r1.79 mule-diag.el
cvs server: conflicting specifications of output style
*** mule-diag.el	19 Sep 2002 05:06:16 -0000	1.79
--- mule-diag.el	25 Sep 2002 08:50:43 -0000
***************
*** 84,94 ****
  but still shows the full information."
    (interactive "P")
    (help-setup-xref (list #'list-character-sets arg) (interactive-p))
!   (with-output-to-temp-buffer (help-buffer)
      (with-current-buffer standard-output
        (if arg
  	  (list-character-sets-2)
  	;; Insert header.
  	(insert
  	 (substitute-command-keys
  	  (concat "Use "
--- 84,95 ----
  but still shows the full information."
    (interactive "P")
    (help-setup-xref (list #'list-character-sets arg) (interactive-p))
!   (with-output-to-temp-buffer "*Character Set List*"
      (with-current-buffer standard-output
        (if arg
  	  (list-character-sets-2)
  	;; Insert header.
+ 	(insert "Indirectly supported character sets are shown below.\n")
  	(insert
  	 (substitute-command-keys
  	  (concat "Use "
***************
*** 117,123 ****
  	(insert "------\t------------\t\t\t--------------\t- -- ----------\n")
  
  	;; Insert body sorted by charset IDs.
! 	(list-character-sets-1 'id)))))
  
  (defun sort-listed-character-sets (sort-key)
    (if sort-key
--- 118,145 ----
  	(insert "------\t------------\t\t\t--------------\t- -- ----------\n")
  
  	;; Insert body sorted by charset IDs.
! 	(list-character-sets-1 'id)
! 
! 	;; Insert non-directly-supported charsets.
! 	(insert-char ?- 72)
! 	(insert "\n\nINDIRECTLY SUPPORTED CHARSETS SETS:\n\n"
! 		(propertize "CHARSET NAME\tMAPPED TO" 'face 'bold)
! 		"\n------------\t---------\n")
! 	(dolist (elt non-iso-charset-alist)
! 	  (insert-text-button (symbol-name (car elt))
! 			      :type 'list-charset-chars
! 			      'help-args (list (car elt)))
! 	  (indent-to 16)
! 	  (dolist (e (nth 1 elt))
! 	    (when (>= (+ (current-column) 1 (string-width (symbol-name e)))
! 		      ;; This is an approximate value.  We don't know
! 		      ;; the correct window width of this buffer yet.
! 		      78)
! 	      (insert "\n")
! 	      (indent-to 16))
! 
! 	    (insert (format "%s " e)))
! 	  (insert "\n"))))))
  
  (defun sort-listed-character-sets (sort-key)
    (if sort-key
***************
*** 127,134 ****
  	  (goto-char (point-min))
  	  (re-search-forward "[0-9][0-9][0-9]")
  	  (beginning-of-line)
! 	  (delete-region (point) (point-max))
! 	  (list-character-sets-1 sort-key)))))
  
  (defun charset-multibyte-form-string (charset)
    (let ((info (charset-info charset)))
--- 149,161 ----
  	  (goto-char (point-min))
  	  (re-search-forward "[0-9][0-9][0-9]")
  	  (beginning-of-line)
! 	  (let ((pos (point)))
! 	    (search-forward "----------")
! 	    (beginning-of-line)
! 	    (save-restriction
! 	      (narrow-to-region pos (point))
! 	      (delete-region (point-min) (point-max))
! 	      (list-character-sets-1 sort-key)))))))
  
  (defun charset-multibyte-form-string (charset)
    (let ((info (charset-info charset)))
***************
*** 249,255 ****
  
  (defvar non-iso-charset-alist
    `((mac-roman
!      nil
       mac-roman-decoder
       ((0 255)))
      (viscii
--- 276,283 ----
  
  (defvar non-iso-charset-alist
    `((mac-roman
!      (ascii latin-iso8859-1 mule-unicode-2500-33ff
! 	    mule-unicode-0100-24ff mule-unicode-e000-ffff)
       mac-roman-decoder
       ((0 255)))
      (viscii
***************
*** 487,494 ****
  characters encoded by various Emacs coding systems which correspond to
  PC `codepages' and other coded character sets.  See `non-iso-charset-alist'."
    (interactive (list (read-charset "Character set: ")))
!   (with-output-to-temp-buffer "*Help*"
      (with-current-buffer standard-output
        (setq indent-tabs-mode nil)
        (set-buffer-multibyte t)
        (cond ((charsetp charset)
--- 515,528 ----
  characters encoded by various Emacs coding systems which correspond to
  PC `codepages' and other coded character sets.  See `non-iso-charset-alist'."
    (interactive (list (read-charset "Character set: ")))
!   (with-output-to-temp-buffer "*Character List*"
      (with-current-buffer standard-output
+       (setq mode-line-format (copy-sequence mode-line-format))
+       (let ((slot (memq 'mode-line-buffer-identification mode-line-format)))
+ 	(if slot
+ 	    (setcdr slot
+ 		    (cons (format " (%s)" charset)
+ 			  (cdr slot)))))
        (setq indent-tabs-mode nil)
        (set-buffer-multibyte t)
        (cond ((charsetp charset)

Re: list charset stuff needs buffer name other than *Help*

[Dan Jacobson]

>>>>> "K" == Kenichi Handa <handa@m17n.org> writes:

>> By the way, (after changing the name of the buffer, of course), I did
>> C-u C-x = on a certain char. and found that none of the many numbers
>> below was like the one on that row, "217x", in the "Characters in the
>> charset chinese-big5-1" table.

K> chinese-big5-1/2 are the charsets invented by Emacs to
K> support Big5.  Big5 characters are mapped to these charsets
K> in Emacs.  Thus, 217x are the code-points of chinese-big5-1/2,
K> not those of Big5 itself.

K> If you want to see the real code-points of Big5, please do:
K> M-x list-charset-chars RET big5 RET.

Hmmm, one would also expect this to be available in the Mule
menu... seems even more often used by users than the thing that gave
me 217x...
-- 
http://jidanni.org/ Taiwan(04)25854780

Re: list charset stuff needs buffer name other than *Help*

[Kenichi Handa]

In article <878z1pvjh1.fsf@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
K>  If you want to see the real code-points of Big5, please do:
K>  M-x list-charset-chars RET big5 RET.

> Hmmm, one would also expect this to be available in the Mule
> menu... seems even more often used by users than the thing that gave
> me 217x...

Currently we have this menu:
  menu-bar -> options -> mule -> list-character-sets
Do you mean you also want something like this?
  menu-bar -> options -> mule -> list-charset-chars -> ascii
						    -> latin-iso8859-1
						    ...
						    -> Big5
						    ...

---
Ken'ichi HANDA
handa@m17n.org

can't get native charset codes from mule menu

[Dan Jacobson]

>>>>> "K" == Kenichi Handa <handa@m17n.org> writes:

K> In article <878z1pvjh1.fsf@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
K> If you want to see the real code-points of Big5, please do:
K> M-x list-charset-chars RET big5 RET.

>> Hmmm, one would also expect this to be available in the Mule
>> menu... seems even more often used by users than the thing that gave
>> me 217x...

K> Currently we have this menu:
K>   menu-bar -> options -> mule -> list-character-sets
K> Do you mean you also want something like this?
K>   menu-bar -> options -> mule -> list-charset-chars -> ascii
-> latin-iso8859-1
K> 						    ...
-> Big5
K> 						    ...

No. I felt bad because anything in
menu-bar -> options -> mule -> list-charset-chars
will just show you the emacs internal codes.
Therefore I want an additional tree:
menu-bar -> options -> mule -> list-charset-chars-with-their-native-codes
or something like that, because:
1. don't most folks want to see a list with the native codes and not
emacs special codes?  But the menu just has the latter and not the
former.
2. Isn't it weird that M-x list-charset-chars will show you native
codes whereas the menu item of the same name shows you emacs codes?
Oh, the menu item is in fact "List Character Sets", but sounds close
enough, and there is no alternative on the menu to get you native
codes.

By the way, I suppose your subtrees idea isn't bad, but remember to
take care of the bigger problem first.
-- 
http://jidanni.org/ Taiwan(04)25854780

Re: can't get native charset codes from mule menu

[Kenichi Handa]

In article <87adlwvoq7.fsf_-_@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
> No. I felt bad because anything in
> menu-bar -> options -> mule -> list-charset-chars
> will just show you the emacs internal codes.

If you simply select `big5' instead of `chinese-big5-1/2',
you can see Big5 native code.  Why is that not enough?

> Therefore I want an additional tree:
> menu-bar -> options -> mule -> list-charset-chars-with-their-native-codes
> or something like that, because:

For that to work, it should accept only indirectly supported
charsets such as cpXXX, big5, viscii.  Otherwise, Emacs
don't know which native charsets a user is expecting.
Then, the menu
should be ...-> mule -> list-indirectly-supported-charset-chars.

But, if we provide this menu:
  ...-> mule -> list-charset-chars
and make it also accepts big5, isn't it enough?

> 1. don't most folks want to see a list with the native codes and not
> emacs special codes?  But the menu just has the latter and not the
> former.
> 2. Isn't it weird that M-x list-charset-chars will show you native
> codes whereas the menu item of the same name shows you emacs codes?

You are misunderstanding what the current menu does.

> Oh, the menu item is in fact "List Character Sets", but sounds close
> enough, and there is no alternative on the menu to get you native
> codes.

As I wrote before, in the current code, the menu "List
Character Sets" (i.e. M-x list-character-sets) shows also
Big5 in the section "INDIRECTLY SUPPORTED CHARACTER SETS".
From there, you can click on `big5' to get a native code
list.

---
Ken'ichi HANDA
handa@m17n.org

Re: can't get native charset codes from mule menu

[Dan Jacobson]

>>>>> "K" == Kenichi Handa <handa@m17n.org> writes:

K> In article <87adlwvoq7.fsf_-_@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
>> No. I felt bad because anything in
>> menu-bar -> options -> mule -> list-charset-chars
>> will just show you the emacs internal codes.

K> If you simply select `big5' instead of `chinese-big5-1/2',
K> you can see Big5 native code.  Why is that not enough?

Huh, I don't see "big5", just the latter:
2 lines matching "big" in buffer *Help*.
     30:152(98)	chinese-big5-1			98 XX XX	2 94 0
     31:153(99)	chinese-big5-2			99 XX XX	2 94 1

>> Therefore I want an additional tree:
>> menu-bar -> options -> mule -> list-charset-chars-with-their-native-codes
>> or something like that, because:

K> For that to work, it should accept only indirectly supported
K> charsets such as cpXXX, big5, viscii.  Otherwise, Emacs
K> don't know which native charsets a user is expecting.
K> Then, the menu
K> should be ...-> mule -> list-indirectly-supported-charset-chars.

K> But, if we provide this menu:
...-> mule -> list-charset-chars
K> and make it also accepts big5, isn't it enough?

oh, I see this hasn't been added yet.  OK, but you must somehow
separate the two flavors of output makers in the list... 

>> 1. don't most folks want to see a list with the native codes and not
>> emacs special codes?  But the menu just has the latter and not the
>> former.
>> 2. Isn't it weird that M-x list-charset-chars will show you native
>> codes whereas the menu item of the same name shows you emacs codes?

K> You are misunderstanding what the current menu does.

>> Oh, the menu item is in fact "List Character Sets", but sounds close
>> enough, and there is no alternative on the menu to get you native
>> codes.

K> As I wrote before, in the current code, the menu "List
K> Character Sets" (i.e. M-x list-character-sets) shows also
K> Big5 in the section "INDIRECTLY SUPPORTED CHARACTER SETS".
K> From there, you can click on `big5' to get a native code
K> list.

Oh, I'm still using emacs-version"21.2.1" so I don't see that I guess.

Re: can't get native charset codes from mule menu

[Kenichi Handa]

In article <87smzi3lqh.fsf@jidanni.org>, Dan Jacobson <jidanni@dman.ddts.net> writes:
K>  But, if we provide this menu:
...->  mule -> list-charset-chars
K>  and make it also accepts big5, isn't it enough?

> oh, I see this hasn't been added yet.  OK, but you must somehow
> separate the two flavors of output makers in the list... 

I don't understand the meaning of "separate the two flavors ...".
Could you explain it?

> Oh, I'm still using emacs-version"21.2.1" so I don't see that I guess.

In 21.2, M-x list-charset-chars RET already accepts "big5".
But, in this version, M-x list-character-sets RET doesn't
list "INDIRECTLY SUPPORTED CHARACTER SETS".

---
Ken'ichi HANDA
handa@m17n.org