what-cursor-position vs. Unicode

what-cursor-position vs. Unicode

[Dan Jacobson]

Today we shall discuss what-cursor-position when given an argument of ^U.
We see that it gives Unicode information:
  character: Z (90, #o132, #x5a, U+005A)
Except when you really need it:
  character: 丹 (107109, #o321145, #x1a265)
It should mention U+4E39.
emacs-version "22.0.50.1"

Re: what-cursor-position vs. Unicode

[Eli Zaretskii]

> From: Dan Jacobson <jidanni@jidanni.org>
> Date: Sat, 03 Jun 2006 10:34:59 +0800
> Cc: handa@etl.go.jp
> 
> It should mention U+4E39.
> emacs-version "22.0.50.1"

It does for me, at least when reading your mail.

Perhaps you should send the original file (as a binary attachment), or
describe how you produced the character, if it wasn't from a file.

Also, please tell when was your Emacs resync'ed with CVS, and please
try looking at the character in "emacs -Q".

Re: what-cursor-position vs. Unicode

[Dan Jacobson]

EZ> It does for me, at least when reading your mail.
Hmmm, me too, but not from a file.
EZ> Also, please tell when was your Emacs resync'ed with CVS
All I know beyond emacs-version "22.0.50.1" is I use Debian
emacs-snapshot 20060518-1.
KH> #x1a265 is a character of chinese-cns11643-1, and the
KH> current Emacs doesn't support Unicode mapping for that
KH> character set.
All I know is me and my Unicode UTF-8 char sitting in the file.
>> Just wondering: Why not?
KH> I myself want to avoid spending a time on what becomes useless in
KH> the future.
I see, there is some funny level of indirection that will be
eliminated in the future. Good.

Re: what-cursor-position vs. Unicode

[Miles Bader]

Dan Jacobson <jidanni@jidanni.org> writes:
> KH> I myself want to avoid spending a time on what becomes useless in
> KH> the future.
>
> I see, there is some funny level of indirection that will be
> eliminated in the future. Good.

In the future (well actually right now, on a CVS branch) Emacs will use
a unicode internal representation, where obviously this sort of thing
will be easier...

-Miles
-- 
"Most attacks seem to take place at night, during a rainstorm, uphill,
 where four map sheets join."   -- Anon. British Officer in WW I

Re: what-cursor-position vs. Unicode

[Kenichi Handa]

In article <87irnjklv0.fsf@jidanni.org>, Dan Jacobson <jidanni@jidanni.org> writes:

> Today we shall discuss what-cursor-position when given an argument of ^U.
> We see that it gives Unicode information:
>   character: Z (90, #o132, #x5a, U+005A)
> Except when you really need it:
>   character: 丹 (107109, #o321145, #x1a265)
> It should mention U+4E39.
> emacs-version "22.0.50.1"

#x1a265 is a character of chinese-cns11643-1, and the
current Emacs doesn't support Unicode mapping for that
character set.

---
Kenichi Handa
handa@m17n.org

Re: what-cursor-position vs. Unicode

[Werner LEMBERG]

> #x1a265 is a character of chinese-cns11643-1, and the
> current Emacs doesn't support Unicode mapping for that
> character set.

Just wondering: Why not?


    Werner

Re: what-cursor-position vs. Unicode

[Kenichi Handa]

In article <20060605.102222.112830788.wl@gnu.org>, Werner LEMBERG <wl@gnu.org> writes:

>> #x1a265 is a character of chinese-cns11643-1, and the
>> current Emacs doesn't support Unicode mapping for that
>> character set.

> Just wondering: Why not?

Because no one has implemented it.  I myself want to avoid
spending a time on what becomes useless in the future.  In
addition, in the current Emacs code, adding something like
lisp/international/subst-cns.el leads to slower startup in
CJK locales, which I want to avoid.

But, if someone implement it and Richard agrees on including
it before the release, please go ahead.

---
Kenichi Handa
handa@m17n.org

Re: what-cursor-position vs. Unicode

[Werner LEMBERG]

[-- Attachment #1: Type: Text/Plain, Size: 1209 bytes --]

> >> #x1a265 is a character of chinese-cns11643-1, and the
> >> current Emacs doesn't support Unicode mapping for that
> >> character set.
>
> > Just wondering: Why not?
>
> Because no one has implemented it.

I've sent a `subst-cns.el' file to you, Ken'ichi-san, and the
experimental diff for utf-8.el is attached.  A great deal of character
codes is larger than U+20000; this works just fine.

> I myself want to avoid spending a time on what becomes useless in
> the future.

Well, it was rather simple; I just wrote a small perl script to
extract the data from the Unihan.txt data base.  On the other hand, I
think it is *very* important to provide good conversion from and to
Unicode for all the charsets Emacs supports, thus it wasn't wasted
time IMHO.

> In addition, in the current Emacs code, adding something like
> lisp/international/subst-cns.el leads to slower startup in CJK
> locales, which I want to avoid.

Agreed -- my changes to utf-8.el don't take this into account.  What
about an additional `unicode' language environment which loads really
all mapping tables?

BTW, I suggest to set up a `Chinese-EUC-TW' language environment for
which `subst-cns.el' is loaded by default.


    Werner

[-- Attachment #2: utf-8.el.diff --]
[-- Type: Text/Plain, Size: 3445 bytes --]

--- utf-8.el.old	2005-10-15 07:43:43.000000000 +0200
+++ utf-8.el	2006-06-09 17:01:46.000000000 +0200
@@ -1,7 +1,7 @@
 ;;; utf-8.el --- UTF-8 decoding/encoding support -*- coding: iso-2022-7bit -*-
 
 ;; Copyright (C) 2001, 2002, 2003, 2004  Free Software Foundation, Inc.
-;; Copyright (C) 2001, 2002, 2003, 2004
+;; Copyright (C) 2001, 2002, 2003, 2004, 2006
 ;;   National Institute of Advanced Industrial Science and Technology (AIST)
 ;;   Registration Number H14PRO021
 
@@ -194,6 +194,10 @@
 
 (defconst utf-translate-cjk-charsets '(chinese-gb2312
 				       chinese-big5-1 chinese-big5-2
+				       chinese-cns11643-1 chinese-cns11643-2
+				       chinese-cns11643-3 chinese-cns11643-4
+				       chinese-cns11643-5 chinese-cns11643-6
+				       chinese-cns11643-7
 				       japanese-jisx0208 japanese-jisx0212
 				       katakana-jisx0201
 				       korean-ksc5601)
@@ -267,7 +271,9 @@
 	ucs-unicode-to-mule-cjk (make-hash-table :test 'eq)))
 
 (defcustom utf-translate-cjk-unicode-range '((#x2e80 . #xd7a3)
-					     (#xff00 . #xffef))
+					     (#xff00 . #xffef)
+					     (#x20000 . #x2a6df)
+					     (#x2f800 . #x2fa1f))
   "List of Unicode code ranges supported by `utf-translate-cjk-mode'.
 Setting this variable directly does not take effect;
 use either \\[customize] or the function
@@ -314,22 +320,26 @@
 	     (load "subst-jis")
 	     (load "subst-big5")
 	     (load "subst-gb2312")
-	     (load "subst-ksc"))
+	     (load "subst-ksc")
+	     (load "subst-cns"))
 	    ((string= "Chinese-BIG5" current-language-environment)
 	     (load "subst-jis")
 	     (load "subst-ksc")
 	     (load "subst-gb2312")
-	     (load "subst-big5"))
+	     (load "subst-big5")
+	     (load "subst-cns"))
 	    ((string= "Chinese-GB" current-language-environment)
 	     (load "subst-jis")
 	     (load "subst-ksc")
 	     (load "subst-big5")
-	     (load "subst-gb2312"))
+	     (load "subst-gb2312")
+	     (load "subst-cns"))
 	    (t
 	     (load "subst-ksc")
 	     (load "subst-gb2312")
 	     (load "subst-big5")
-	     (load "subst-jis")))) ; jis covers as much as big5, gb2312
+	     (load "subst-jis")
+	     (load "subst-cns")))) ; jis covers as much as big5, gb2312
 
     (when redefined
       (define-translation-hash-table 'utf-subst-table-for-decode
@@ -365,14 +375,22 @@
 zero or negative.  This is a minor mode.
 Enabling this allows the coding systems mule-utf-8,
 mule-utf-16le and mule-utf-16be to encode characters in the charsets
-`korean-ksc5601', `chinese-gb2312', `chinese-big5-1',
-`chinese-big5-2', `japanese-jisx0208' and `japanese-jisx0212', and to
-decode the corresponding unicodes into such characters.
+
+  korean-ksc5601
+  chinese-gb2312
+  chinese-big5-1 chinese-big5-2
+  chinese-cns11643-1 chinese-cns11643-2 chinese-cns11643-3
+  chinese-cns11643-4 chinese-cns11643-5 chinese-cns11643-6
+  chinese-cns11643-7
+  japanese-jisx0208 japanese-jisx0212
+
+and to decode the corresponding unicodes into such characters.
 
 Where the charsets overlap, the one preferred for decoding is chosen
 according to the language environment in effect when this option is
 turned on: ksc5601 for Korean, gb2312 for Chinese-GB, big5 for
-Chinese-Big5 and jisx for other environments.
+Chinese-Big5 and jisx for other environments.  The CNS charsets
+are always loaded last.
 
 This mode is on by default.  If you are not interested in CJK
 characters and want to avoid some overhead on encoding/decoding

[-- Attachment #3: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel