bug#71411: 30.0.50; Show Unicode names when describing translations

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

'C-x 8 RET' will show Unicode names when it shows completions, which
helps people find the complete name of the character they want to use.

I have a patch to do something similar for 'C-h k' and 'C-h b' (for
translated keys only), so youʼd get

'C-h k C-x 8 E' =>

    € 'EURO SIGN' (translated from C-x 8 E)

Since the purpose of this is to provide more information by default in
situations where people are searching for answers, I
donʼt think this needs a configuration variable, unlike the analogous
feaure in 'what-cursor-position'. (yes, people can hit 'C-u C-x =' in
the *Help* buffer, but first they have to know about that).

Patch to follow when I have the bug number.

Robert
-- 

bug#71411: 30.0.50; Show Unicode names when describing translations

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Date: Fri, 07 Jun 2024 12:32:58 +0200
> 
> 
> 'C-x 8 RET' will show Unicode names when it shows completions, which
> helps people find the complete name of the character they want to use.
> 
> I have a patch to do something similar for 'C-h k' and 'C-h b' (for
> translated keys only), so youʼd get
> 
> 'C-h k C-x 8 E' =>
> 
>     € 'EURO SIGN' (translated from C-x 8 E)
> 
> Since the purpose of this is to provide more information by default in
> situations where people are searching for answers, I
> donʼt think this needs a configuration variable, unlike the analogous
> feaure in 'what-cursor-position'. (yes, people can hit 'C-u C-x =' in
> the *Help* buffer, but first they have to know about that).

Does this mean that if I type "C-h k a" I will be told that a is LATIN
SMALL LETTER A?  That'd be an annoyance, I think.

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

[-- Attachment #1: Type: text/plain, Size: 310 bytes --]

>>>>> On Fri, 07 Jun 2024 14:08:33 +0300, Eli Zaretskii <eliz@gnu.org> said:

    Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
    Eli> SMALL LETTER A?  That'd be an annoyance, I think.

Thatʼs not a translated binding, so no.

Let me attach the patches.

Robert
-- 

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Add-char-to-name-convenience-function.patch --]
[-- Type: text/x-diff, Size: 2554 bytes --]

From 847d582f89110f8c2f0e88c1533ced0606970fba Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Fri, 7 Jun 2024 12:05:34 +0200
Subject: [PATCH 1/3] Add 'char-to-name' convenience function
To: emacs-devel@gnu.org

* lisp/international/mule-cmds.el (char-to-name): New function.  This is
the inverse of 'char-from-name'.
* doc/lispref/nonascii.texi (Character Codes): Document it.
* etc/NEWS: Announce it.
---
 doc/lispref/nonascii.texi       | 5 +++++
 etc/NEWS                        | 4 ++++
 lisp/international/mule-cmds.el | 7 +++++++
 3 files changed, 16 insertions(+)

diff --git a/doc/lispref/nonascii.texi b/doc/lispref/nonascii.texi
index b33082e2b24..195b818c0bd 100644
--- a/doc/lispref/nonascii.texi
+++ b/doc/lispref/nonascii.texi
@@ -435,6 +435,11 @@ Character Codes
 @end example
 @end defun
 
+@defun char-to-name char
+This function returns the Unicode name of @var{char}.  Returns
+@code{nil} if @var{char} is not a character or has no Unicode name.
+@end defun
+
 @defun get-byte &optional pos string
 This function returns the byte at character position @var{pos} in the
 current buffer.  If the current buffer is unibyte, this is literally
diff --git a/etc/NEWS b/etc/NEWS
index 5fb1625a76c..c63e231f44f 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -2186,6 +2186,10 @@ it returns a short "unique" string that identifies the function.
 In either case, the string is propertized so clicking on it gives
 further details.
 
+** New function 'char-to-name'.
+This is a convenience function to return the Unicode name of a char (if
+it has one).
+
 ** New function 'cl-type-of'.
 This function is like 'type-of' except that it sometimes returns
 a more precise type.  For example, for nil and t it returns 'null'
diff --git a/lisp/international/mule-cmds.el b/lisp/international/mule-cmds.el
index e80c42f523a..7d784ef3b1b 100644
--- a/lisp/international/mule-cmds.el
+++ b/lisp/international/mule-cmds.el
@@ -3183,6 +3183,13 @@ mule--ucs-names-group
            (script (and char (aref char-script-table char))))
       (if script (symbol-name script) "ungrouped"))))
 
+(defun char-to-name (char)
+  "Return the Unicode name for CHAR, if it has one, else nil.
+Return nil if CHAR is not a character."
+  (and (characterp char)
+       (or (get-char-code-property char 'name)
+           (get-char-code-property char 'old-name))))
+
 (defun char-from-name (string &optional ignore-case)
   "Return a character as a number from its Unicode name STRING.
 If optional IGNORE-CASE is non-nil, ignore case in STRING.
-- 
2.39.2


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #3: 0002-Use-char-to-name.patch --]
[-- Type: text/x-diff, Size: 2188 bytes --]

From 5ff4c3ac95d8dfcc172ad36f0ebd9c12443c9046 Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Fri, 7 Jun 2024 12:07:50 +0200
Subject: [PATCH 2/3] Use 'char-to-name'
To: emacs-devel@gnu.org

* lisp/international/emoji.el (emoji--name): Lookup name using 'char-to-name'.
* lisp/international/textsec.el (textsec-domain-suspicious-p): And here.
* lisp/simple.el (what-cursor-position): And here.
---
 lisp/international/emoji.el   | 2 +-
 lisp/international/textsec.el | 2 +-
 lisp/simple.el                | 3 +--
 3 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/lisp/international/emoji.el b/lisp/international/emoji.el
index 4f3aab5a6be..cbf7c547f4c 100644
--- a/lisp/international/emoji.el
+++ b/lisp/international/emoji.el
@@ -245,7 +245,7 @@ emoji--fontify-glyph
 
 (defun emoji--name (glyph)
   (or (gethash glyph emoji--names)
-      (get-char-code-property (aref glyph 0) 'name)))
+      (char-to-name (aref glyph 0))))
 
 (defvar-keymap emoji-list-mode-map
   "RET" #'emoji-list-select
diff --git a/lisp/international/textsec.el b/lisp/international/textsec.el
index 4740dd81345..224363732fe 100644
--- a/lisp/international/textsec.el
+++ b/lisp/international/textsec.el
@@ -296,7 +296,7 @@ textsec-domain-suspicious-p
                             ""
                           (concat ": " (string char)))
                         char
-                        (get-char-code-property char 'name)))))
+                        (char-to-name char)))))
      domain)
     ;; Does IDNA allow it?
     (unless (puny-highly-restrictive-domain-p domain)
diff --git a/lisp/simple.el b/lisp/simple.el
index 68209eadc41..8d7e379b584 100644
--- a/lisp/simple.el
+++ b/lisp/simple.el
@@ -1827,8 +1827,7 @@ what-cursor-position
   (interactive "P")
   (let* ((char (following-char))
          (char-name (and what-cursor-show-names
-                         (or (get-char-code-property char 'name)
-                             (get-char-code-property char 'old-name))))
+                         (char-to-name char)))
          (char-name-fmt (if char-name
                             (format ", %s" char-name)
                           ""))
-- 
2.39.2


[-- Attachment #4: 0003-Show-character-names-when-describing-translations.patch --]
[-- Type: text/x-diff, Size: 2772 bytes --]

From a2adb11e15a60ad30741c6cd8c47f03b5b1a902e Mon Sep 17 00:00:00 2001
From: Robert Pluim <rpluim@gmail.com>
Date: Fri, 7 Jun 2024 12:21:11 +0200
Subject: [PATCH 3/3] Show character names when describing translations
To: emacs-devel@gnu.org

This implements Bug#71411.

* lisp/help.el (help-key-description): Use 'char-to-name' to show the
Unicode name of translated keys.
(help--describe-command): And here.

* etc/NEWS: Announce the change.
---
 etc/NEWS     | 14 ++++++++++++++
 lisp/help.el |  9 +++++++--
 2 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/etc/NEWS b/etc/NEWS
index c63e231f44f..66bc4d741ff 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -182,6 +182,20 @@ This user option controls outline visibility in the output buffer of
 *** 'C-h m' ('describe-mode') uses outlining by default.
 Set 'describe-mode-outline' to nil to get back the old behavior.
 
+*** 'C-h k' ('describe-key') shows Unicode name.
+For translated keys only, this will now show the Unicode name of the
+translation, e.g. 'C-h k C-x 8 E' will now show
+
+    € 'EURO SIGN' (translated from C-x 8 E)
+
+*** 'C-h b' ('describe-bindings') shows Unicode names.
+For translated keys only, Unicode names will now be shown, i.e.
+
+    A-!			¡	INVERTED EXCLAMATION MARK
+    A-$			¤	CURRENCY SIGN
+
+and so on.
+
 ** Outline mode
 
 *** New commands to show/hide outlines by regexp.
diff --git a/lisp/help.el b/lisp/help.el
index 2feb178ff6c..878dd404256 100644
--- a/lisp/help.el
+++ b/lisp/help.el
@@ -883,7 +883,9 @@ help-key-description
       (let ((otherstring (help--key-description-fontified untranslated)))
 	(if (equal string otherstring)
 	    string
-	  (format "%s (translated from %s)" string otherstring))))))
+          (if-let ((char-name (char-to-name (aref string 0))))
+              (format "%s '%s' (translated from %s)" string char-name otherstring)
+            (format "%s (translated from %s)" string otherstring)))))))
 
 (defun help--binding-undefined-p (defn)
   (or (null defn) (integerp defn) (equal defn #'undefined)))
@@ -1664,7 +1666,10 @@ help--shadow-lookup
 (defun help--describe-command (definition &optional translation)
   (cond ((or (stringp definition) (vectorp definition))
          (if translation
-             (insert (key-description definition nil) "\n")
+             (insert (concat (key-description definition nil)
+                             (when-let ((char-name (char-to-name (aref definition 0))))
+                               (format "\t%s" char-name))
+                             "\n"))
            ;; These should be rare nowadays, replaced by `kmacro's.
            (insert "Keyboard Macro\n")))
         ((keymapp definition)
-- 
2.39.2

bug#71411: 30.0.50; Show Unicode names when describing translations

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Cc: 71411@debbugs.gnu.org
> Date: Fri, 07 Jun 2024 14:16:38 +0200
> 
>     Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
>     Eli> SMALL LETTER A?  That'd be an annoyance, I think.
> 
> Thatʼs not a translated binding, so no.

What is a "translated binding" or "translated key"?  Without
explaining that, at least the NEWS entries are not clear enough.

bug#71411: 30.0.50; Show Unicode names when describing translations

[Stephen Berman via Bug reports for GNU Emacs, the Swiss army knife of text editors]

On Fri, 07 Jun 2024 14:16:38 +0200 Robert Pluim <rpluim@gmail.com> wrote:

>>>>>> On Fri, 07 Jun 2024 14:08:33 +0300, Eli Zaretskii <eliz@gnu.org> said:
>
>     Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
>     Eli> SMALL LETTER A?  That'd be an annoyance, I think.
>
> Thatʼs not a translated binding, so no.
>
> Let me attach the patches.
[...]
> diff --git a/etc/NEWS b/etc/NEWS
> index c63e231f44f..66bc4d741ff 100644
> --- a/etc/NEWS
> +++ b/etc/NEWS
> @@ -182,6 +182,20 @@ This user option controls outline visibility in the output buffer of
>  *** 'C-h m' ('describe-mode') uses outlining by default.
>  Set 'describe-mode-outline' to nil to get back the old behavior.
>  
> +*** 'C-h k' ('describe-key') shows Unicode name.
> +For translated keys only, this will now show the Unicode name of the
> +translation, e.g. 'C-h k C-x 8 E' will now show
> +
> +    € 'EURO SIGN' (translated from C-x 8 E)

How do you input 'C-h k C-x 8 E'?  I can only input 'C-h k C-x 8 e'.
(Also with 'C-x 8' typing 'E' (i.e. 'S-e') inputs 'e').

Steve Berman

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

>>>>> On Fri, 07 Jun 2024 16:42:01 +0300, Eli Zaretskii <eliz@gnu.org> said:

    >> From: Robert Pluim <rpluim@gmail.com>
    >> Cc: 71411@debbugs.gnu.org
    >> Date: Fri, 07 Jun 2024 14:16:38 +0200
    >> 
    Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
    Eli> SMALL LETTER A?  That'd be an annoyance, I think.
    >> 
    >> Thatʼs not a translated binding, so no.

    Eli> What is a "translated binding" or "translated key"?  Without
    Eli> explaining that, at least the NEWS entries are not clear enough.

Itʼs anything that goes through `key-translation-map'. How
about:

-- >8 --
    *** 'C-h k' ('describe-key') shows Unicode name.
    For keybindings which produce single characters via translation (such as
    those using the 'C-x 8' prefix), 'C-h k' now shows the Unicode name of
    the produced character in addition to the character itself, e.g.

    'C-h k C-x 8 E' =>

        € 'EURO SIGN' (translated from C-x 8 E)

    *** 'C-h b' ('describe-bindings') shows Unicode names.
    For keybindings which produce single characters via translation, the
    Unicode names will now be shown in addition to the character itself,
    i.e.

        A-!			¡	INVERTED EXCLAMATION MARK
        A-$			¤	CURRENCY SIGN

    and so on.
-- >8 --

Robert
-- 

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

>>>>> On Fri, 07 Jun 2024 16:01:41 +0200, Stephen Berman <stephen.berman@gmx.net> said:

    Stephen> On Fri, 07 Jun 2024 14:16:38 +0200 Robert Pluim <rpluim@gmail.com> wrote:
    >>>>>>> On Fri, 07 Jun 2024 14:08:33 +0300, Eli Zaretskii <eliz@gnu.org> said:
    >> 
    Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
    Eli> SMALL LETTER A?  That'd be an annoyance, I think.
    >> 
    >> Thatʼs not a translated binding, so no.
    >> 
    >> Let me attach the patches.
    Stephen> [...]
    >> diff --git a/etc/NEWS b/etc/NEWS
    >> index c63e231f44f..66bc4d741ff 100644
    >> --- a/etc/NEWS
    >> +++ b/etc/NEWS
    >> @@ -182,6 +182,20 @@ This user option controls outline visibility in the output buffer of
    >> *** 'C-h m' ('describe-mode') uses outlining by default.
    >> Set 'describe-mode-outline' to nil to get back the old behavior.
    >> 
    >> +*** 'C-h k' ('describe-key') shows Unicode name.
    >> +For translated keys only, this will now show the Unicode name of the
    >> +translation, e.g. 'C-h k C-x 8 E' will now show
    >> +
    >> +    € 'EURO SIGN' (translated from C-x 8 E)

    Stephen> How do you input 'C-h k C-x 8 E'?  I can only input 'C-h k C-x 8 e'.
    Stephen> (Also with 'C-x 8' typing 'E' (i.e. 'S-e') inputs 'e').

Itʼs a binding I added to master on Monday. You can use

'C-h k C-x 8 * E'

instead.

Robert
-- 

bug#71411: 30.0.50; Show Unicode names when describing translations

[Stephen Berman]

On Fri, 07 Jun 2024 16:25:12 +0200 Robert Pluim <rpluim@gmail.com> wrote:

>>>>>> On Fri, 07 Jun 2024 16:01:41 +0200, Stephen Berman
> <stephen.berman@gmx.net> said:
>
>     Stephen> On Fri, 07 Jun 2024 14:16:38 +0200 Robert Pluim
>     Stephen> <rpluim@gmail.com> wrote:
>     >>>>>>> On Fri, 07 Jun 2024 14:08:33 +0300, Eli Zaretskii <eliz@gnu.org> said:
>     >> 
>     Eli> Does this mean that if I type "C-h k a" I will be told that a is LATIN
>     Eli> SMALL LETTER A?  That'd be an annoyance, I think.
>     >> 
>     >> Thatʼs not a translated binding, so no.
>     >> 
>     >> Let me attach the patches.
>     Stephen> [...]
>     >> diff --git a/etc/NEWS b/etc/NEWS
>     >> index c63e231f44f..66bc4d741ff 100644
>     >> --- a/etc/NEWS
>     >> +++ b/etc/NEWS
>     >> @@ -182,6 +182,20 @@ This user option controls outline visibility in
>     >> the output buffer of
>     >> *** 'C-h m' ('describe-mode') uses outlining by default.
>     >> Set 'describe-mode-outline' to nil to get back the old behavior.
>     >> 
>     >> +*** 'C-h k' ('describe-key') shows Unicode name.
>     >> +For translated keys only, this will now show the Unicode name of the
>     >> +translation, e.g. 'C-h k C-x 8 E' will now show
>     >> +
>     >> +    € 'EURO SIGN' (translated from C-x 8 E)
>
>     Stephen> How do you input 'C-h k C-x 8 E'?  I can only input 'C-h k C-x 8 e'.
>     Stephen> (Also with 'C-x 8' typing 'E' (i.e. 'S-e') inputs 'e').
>
> Itʼs a binding I added to master on Monday.

Ah, time to git pull again.

>                                             You can use
>
> 'C-h k C-x 8 * E'
>
> instead.

Oh, right, I forgot about that.  Thanks.

Steve Berman

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

>>>>> On Fri, 07 Jun 2024 16:29:38 +0200, Stephen Berman <stephen.berman@gmx.net> said:

    >> You can use
    >> 
    >> 'C-h k C-x 8 * E'
    >> 
    >> instead.

    Stephen> Oh, right, I forgot about that.  Thanks.

Thatʼs why I added 'C-x 8 E' 😀

Robert
-- 

bug#71411: 30.0.50; Show Unicode names when describing translations

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Cc: 71411@debbugs.gnu.org
> Date: Fri, 07 Jun 2024 16:06:22 +0200
> 
> >>>>> On Fri, 07 Jun 2024 16:42:01 +0300, Eli Zaretskii <eliz@gnu.org> said:
> 
>     Eli> What is a "translated binding" or "translated key"?  Without
>     Eli> explaining that, at least the NEWS entries are not clear enough.
> 
> Itʼs anything that goes through `key-translation-map'. How
> about:
> 
> -- >8 --
>     *** 'C-h k' ('describe-key') shows Unicode name.
>     For keybindings which produce single characters via translation (such as
>     those using the 'C-x 8' prefix), 'C-h k' now shows the Unicode name of
>     the produced character in addition to the character itself, e.g.

Are there any other cases of "translated characters", besides "C-x 8"?
If so, please mention one or two more.  If there are no other
examples, I'd remove "such as".

Otherwise, LGTM; please install.

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

>>>>> On Sat, 08 Jun 2024 16:24:08 +0300, Eli Zaretskii <eliz@gnu.org> said:

    >> *** 'C-h k' ('describe-key') shows Unicode name.
    >> For keybindings which produce single characters via translation (such as
    >> those using the 'C-x 8' prefix), 'C-h k' now shows the Unicode name of
    >> the produced character in addition to the character itself, e.g.

    Eli> Are there any other cases of "translated characters", besides "C-x 8"?
    Eli> If so, please mention one or two more.  If there are no other
    Eli> examples, I'd remove "such as".

iso-transl.el is (almost) the only thing that uses
`key-translation-map' (the other is 'double.el', but thatʼs really
niche). I can mention the other prefixes.

-- >8 --
*** 'C-h k' ('describe-key') shows Unicode name.
For keybindings which produce single characters via translation (such as
those using the 'C-x 8' or 'A-' prefix, or 'dead-acute', 'dead-grave',
etc), 'C-h k' now shows the Unicode name of the produced character in
addition to the character itself, e.g.

'C-h k C-x 8 E' =>

    € 'EURO SIGN' (translated from C-x 8 E)

*** 'C-h b' ('describe-bindings') shows Unicode names.
For keybindings which produce single characters via translation, the
Unicode names will now be shown in addition to the character itself,
i.e.

    A-!			¡	INVERTED EXCLAMATION MARK
    A-$			¤	CURRENCY SIGN

and so on.
-- >8 --
Robert
-- 

bug#71411: 30.0.50; Show Unicode names when describing translations

[Eli Zaretskii]

> From: Robert Pluim <rpluim@gmail.com>
> Cc: 71411@debbugs.gnu.org
> Date: Mon, 10 Jun 2024 11:03:07 +0200
> 
> >>>>> On Sat, 08 Jun 2024 16:24:08 +0300, Eli Zaretskii <eliz@gnu.org> said:
> 
>     Eli> Are there any other cases of "translated characters", besides "C-x 8"?
>     Eli> If so, please mention one or two more.  If there are no other
>     Eli> examples, I'd remove "such as".
> 
> iso-transl.el is (almost) the only thing that uses
> `key-translation-map' (the other is 'double.el', but thatʼs really
> niche). I can mention the other prefixes.
> 
> -- >8 --
> *** 'C-h k' ('describe-key') shows Unicode name.
> For keybindings which produce single characters via translation (such as
> those using the 'C-x 8' or 'A-' prefix, or 'dead-acute', 'dead-grave',
> etc), 'C-h k' now shows the Unicode name of the produced character in
> addition to the character itself, e.g.
> 
> 'C-h k C-x 8 E' =>
> 
>     € 'EURO SIGN' (translated from C-x 8 E)
> 
> *** 'C-h b' ('describe-bindings') shows Unicode names.
> For keybindings which produce single characters via translation, the
> Unicode names will now be shown in addition to the character itself,
> i.e.
> 
>     A-!			¡	INVERTED EXCLAMATION MARK
>     A-$			¤	CURRENCY SIGN
> 
> and so on.

What about characters produced by input methods?  For example:

  C-u C-\ cyrillic-translit RET
  C-h k z h =>

    ж (translated from h ж) runs the command self-insert-command

(which, btw, uncovers a bug, since "h ж" is bogus and should be "z h"
instead).

bug#71411: 30.0.50; Show Unicode names when describing translations

[Robert Pluim]

>>>>> On Mon, 10 Jun 2024 15:34:35 +0300, Eli Zaretskii <eliz@gnu.org> said:

    Eli> What about characters produced by input methods?  For example:

    Eli>   C-u C-\ cyrillic-translit RET
    Eli>   C-h k z h =>

    Eli>     ж (translated from h ж) runs the command self-insert-command

I completely forgot to check input methods, it works there as well:

    ж 'CYRILLIC SMALL LETTER ZHE' (translated from h ж) runs the command
    self-insert-command (found in global-map), which is an interactive
    primitive-function in ‘C source code’.

Iʼll put that in the NEWS entry.

    Eli> (which, btw, uncovers a bug, since "h ж" is bogus and should be "z h"
    Eli> instead).

Right, but thatʼs a separate issue.

Robert
--