bug#34375: Problem with the 'to input' information in what-cursor-position

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Andy Smith
I ran what-cursor-position on a unicode char. Get the following ...

position: 12 of 16 (69%), column: 0
            character: Ő (displayed as Ő) (codepoint 336, #o520, #x150)
    preferred charset: unicode (Unicode (ISO10646))
code point in charset: 0x0150
               script: latin
               syntax: w which means: word
             category: .:Base, L:Left-to-right (strong), j:Japanese, l:Latin
             to input: type "C-x 8 RET 150" or "C-x 8 RET LATIN CAPITAL LETTER O WITH DOUBLE ACUTE"
          buffer code: #xC5 #x90
            file code: #xC5 #x90 (encoded by coding system utf-8-unix)
              display: by this font (glyph code)
    xft:-DAMA-Ubuntu Mono-normal-normal-normal-*-72-*-*-*-m-0-iso10646-1
            (#x12E)

If I follow the 'to input' instructions ... i.e "C-x 8 RET 150" this
doesn't work. i.e. I can roundtrip and insert a copy of the character I
initially looked up. Needed to do "C-x 8 RET 0150" instead.

The to-input field maybe needs to display an octal string properly
prefixed with a zero maybe?

Rgds,

A.

Reply | Threaded
Open this post in threaded view
|

bug#34375: Correction....

Andy Smith
Sorry - I said I 'can' roundtrip between the instructions and the character. I mean't to say "*can't* roundtrip". Sorry for the confusion.

Rgds,

A.

Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Robert Pluim
In reply to this post by Andy Smith
Andy Smith <[hidden email]> writes:

> I ran what-cursor-position on a unicode char. Get the following ...
>
> position: 12 of 16 (69%), column: 0
>             character: Ő (displayed as Ő) (codepoint 336, #o520, #x150)
>     preferred charset: unicode (Unicode (ISO10646))
> code point in charset: 0x0150
>                script: latin
>                syntax: w which means: word
>              category: .:Base, L:Left-to-right (strong), j:Japanese, l:Latin
>              to input: type "C-x 8 RET 150" or "C-x 8 RET LATIN CAPITAL
> LETTER O WITH DOUBLE ACUTE"
>           buffer code: #xC5 #x90
>             file code: #xC5 #x90 (encoded by coding system utf-8-unix)
>               display: by this font (glyph code)
>     xft:-DAMA-Ubuntu Mono-normal-normal-normal-*-72-*-*-*-m-0-iso10646-1
>             (#x12E)
>
> If I follow the 'to input' instructions ... i.e "C-x 8 RET 150" this
> doesn't work. i.e. I can roundtrip and insert a copy of the character I
> initially looked up. Needed to do "C-x 8 RET 0150" instead.
>
> The to-input field maybe needs to display an octal string properly
> prefixed with a zero maybe?

(you didnʼt specify which version of Emacs youʼre using)

The number you need to enter is in hex, and C-x 8 ret implicitly
assumes that numbers entered are in hexadecimal format, at least in
26.1 and master, so I suspect youʼre not running either of those.

Robert



Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Eli Zaretskii
In reply to this post by Andy Smith
> From: Andy Smith <[hidden email]>
> Date: Thu, 7 Feb 2019 23:00:01 +0000
>
> I ran what-cursor-position on a unicode char. Get the following ...
>
> position: 12 of 16 (69%), column: 0
>             character: Ő (displayed as Ő) (codepoint 336, #o520, #x150)
>     preferred charset: unicode (Unicode (ISO10646))
> code point in charset: 0x0150
>                script: latin
>                syntax: w which means: word
>              category: .:Base, L:Left-to-right (strong), j:Japanese, l:Latin
>              to input: type "C-x 8 RET 150" or "C-x 8 RET LATIN CAPITAL LETTER O WITH DOUBLE ACUTE"
>           buffer code: #xC5 #x90
>             file code: #xC5 #x90 (encoded by coding system utf-8-unix)
>               display: by this font (glyph code)
>     xft:-DAMA-Ubuntu Mono-normal-normal-normal-*-72-*-*-*-m-0-iso10646-1
>             (#x12E)
>
> If I follow the 'to input' instructions ... i.e "C-x 8 RET 150" this
> doesn't work.

I cannot reproduce this.  It works for me.  What do you get if you
type "C-x 8 RET 150", as indicated in what-cursor-position's output?

> The to-input field maybe needs to display an octal string properly
> prefixed with a zero maybe?

No, the code is interpreted as hex by "C-x 8 RET".  So I wonder what
went wrong in your case.  Can you show the result of "C-h l" (the
letter ell, not the digit one) after typing "C-x 8 RET 150"?



Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Andy Smith
Thanks for getting back.

(version)
"GNU Emacs 26.1 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.22.30)
 of 2018-07-05"

Just to explain what I'm seeing. Apologies for having to use screenshot but the I'm not sure if the unicode stuff will translate well via email :-(

My point is that 'what-char-position' gives an incorrect instruction to the user for character 150. I think it should really say ....

to input:  type "C-8 RET 0150 ...."

instead of

to input:  type "C-8 RET 150 ...."

2c60 works as expected so maybe it's a problem for characters who have a hex representation less than four letters.

Cheers,

A.



2019-02-08_10-11.png




On Fri, 8 Feb 2019 at 07:11, Eli Zaretskii <[hidden email]> wrote:
> From: Andy Smith <[hidden email]>
> Date: Thu, 7 Feb 2019 23:00:01 +0000
>
> I ran what-cursor-position on a unicode char. Get the following ...
>
> position: 12 of 16 (69%), column: 0
>             character: Ő (displayed as Ő) (codepoint 336, #o520, #x150)
>     preferred charset: unicode (Unicode (ISO10646))
> code point in charset: 0x0150
>                script: latin
>                syntax: w which means: word
>              category: .:Base, L:Left-to-right (strong), j:Japanese, l:Latin
>              to input: type "C-x 8 RET 150" or "C-x 8 RET LATIN CAPITAL LETTER O WITH DOUBLE ACUTE"
>           buffer code: #xC5 #x90
>             file code: #xC5 #x90 (encoded by coding system utf-8-unix)
>               display: by this font (glyph code)
>     xft:-DAMA-Ubuntu Mono-normal-normal-normal-*-72-*-*-*-m-0-iso10646-1
>             (#x12E)
>
> If I follow the 'to input' instructions ... i.e "C-x 8 RET 150" this
> doesn't work.

I cannot reproduce this.  It works for me.  What do you get if you
type "C-x 8 RET 150", as indicated in what-cursor-position's output?

> The to-input field maybe needs to display an octal string properly
> prefixed with a zero maybe?

No, the code is interpreted as hex by "C-x 8 RET".  So I wonder what
went wrong in your case.  Can you show the result of "C-h l" (the
letter ell, not the digit one) after typing "C-x 8 RET 150"?
Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Eli Zaretskii
> From: Andy Smith <[hidden email]>
> Date: Fri, 8 Feb 2019 10:18:21 +0000
>
> (version)
> "GNU Emacs 26.1 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.22.30)
>  of 2018-07-05"
>
> Just to explain what I'm seeing. Apologies for having to use screenshot but the I'm not sure if the unicode stuff
> will translate well via email :-(
>
> My point is that 'what-char-position' gives an incorrect instruction to the user for character 150. I think it should
> really say ....
>
> to input:  type "C-8 RET 0150 ...."
>
> instead of
>
> to input:  type "C-8 RET 150 ...."

Are you seeing this problem in "emacs -Q"?  (It's "C-x 8 RET", btw,
not "C-8 RER", right?)

The character that you get is U+1009F LINEAR B IDEOGRAM B150.  Its
name includes "150" as a substring, so perhaps you did something that
caused Emacs to auto-complete the name?  Like typing TAB or using some
optional completion package?  That's why I ask about "emacs -Q".



Reply | Threaded
Open this post in threaded view
|

bug#34375: Problem with the 'to input' information in what-cursor-position

Andy Smith
<facepalm>

yes you're right. Helm is the culprit :-(. Sorry for wasting your time, my intentions were honest!

I've just double checked with 'emacs -Q'. All works as expected.

Thanks for your help!

Cheers,

A.


On Fri, 8 Feb 2019 at 10:40, Eli Zaretskii <[hidden email]> wrote:
> From: Andy Smith <[hidden email]>
> Date: Fri, 8 Feb 2019 10:18:21 +0000
>
> (version)
> "GNU Emacs 26.1 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.22.30)
>  of 2018-07-05"
>
> Just to explain what I'm seeing. Apologies for having to use screenshot but the I'm not sure if the unicode stuff
> will translate well via email :-(
>
> My point is that 'what-char-position' gives an incorrect instruction to the user for character 150. I think it should
> really say ....
>
> to input:  type "C-8 RET 0150 ...."
>
> instead of
>
> to input:  type "C-8 RET 150 ...."

Are you seeing this problem in "emacs -Q"?  (It's "C-x 8 RET", btw,
not "C-8 RER", right?)

The character that you get is U+1009F LINEAR B IDEOGRAM B150.  Its
name includes "150" as a substring, so perhaps you did something that
caused Emacs to auto-complete the name?  Like typing TAB or using some
optional completion package?  That's why I ask about "emacs -Q".