desktop and encodings

desktop and encodings

[Mads Jensen]

Hi NG

I wish to have desktop-mode store the encodings, so I won't have to set
them manually after desktop opened with "C-x RET f RET encoding". 

I once had this enabled, but lost the configuration. I'd be very
greatful if anyone knows how to do this.

-- 
Mads Jensen - mail sent to address ends in /dev/null
	      s/spam/madsj for emailing me

Research is what I'm doing when I don't know what I'm doing.
                -- Wernher von Braun

Re: desktop and encodings

[Kevin Rodgers]

Mads Jensen wrote:
> I wish to have desktop-mode store the encodings, so I won't have to set
> them manually after desktop opened with "C-x RET f RET encoding". 

(setq desktop-locals-to-save
       (cons 'buffer-file-coding-system desktop-locals-to-save))

-- 
Kevin Rodgers

Re: desktop and encodings

[Mads Jensen]

Kevin Rodgers <ihs_4664@yahoo.com> writes:
> (setq desktop-locals-to-save
>        (cons 'buffer-file-coding-system desktop-locals-to-save))

This leaves me with one major problem, though. I occassionally use the
Danish special characters æ, ø and å (ae, oe and aa, if they come out as
weird symbols) in unicode, and when I then open emacs, they come out
incorrectly, so I sometimes have to spend some time replacing those
symbols into the right characters, if I briefly forgot about the bug and
saved the file. 

I use emacs GNU Emacs 21.3.2 on Slackware 10.1 with the default LANG=
variables set to "en_US" in /etc/profile.d/lang.sh

I'd be very happy if anyone has a solution to this problem.
-- 
Mads Jensen - mail sent to address ends in /dev/null
	      s/spam/madsj for emailing me

Research is what I'm doing when I don't know what I'm doing.
                -- Wernher von Braun

Re: desktop and encodings

[Peter Dyballa]

Am 21.05.2005 um 20:33 schrieb Mads Jensen:

> I use emacs GNU Emacs 21.3.2 on Slackware 10.1 with the default LANG=
> variables set to "en_US" in /etc/profile.d/lang.sh
>
> I'd be very happy if anyone has a solution to this problem.
>

That's probably the reason why you have these problems! Set LANG to 
Danish in some ISO 8859 encoding or in UTF-8! Maybe da_DA.UTF-8 ...

The other thing you can do: give all your files with non-US ASCII 
content in the header a remark like:

;;; -*- mode: Text; coding: iso-8859-4; -*-

GNU Emacs will learn from that this text files contents is Northern ISO 
Latin.

--
Greetings
                                  <]
    Pete      o        __o         |__    o           recumbo
     ___o    /I       -\<,         |o \  -\),-%       ergo sum!
___/\ /\___./ \___...O/ O____.....`-O-'-()--o_________________

Re: desktop and encodings

[Mads Jensen]

Peter Dyballa <Peter_Dyballa@Web.DE> writes:
> That's probably the reason why you have these problems! Set LANG to
> Danish in some ISO 8859 encoding or in UTF-8! Maybe da_DA.UTF-8 ...

I tried setting the LANG variable to both da_DA, da_DA.utf-8,
da_DA.iso-8859-1 plus en_US... (the same LANG variables as mentioned
before, just with a different language), but that did not help :-/ 
 
> The other thing you can do: give all your files with non-US ASCII
> content in the header a remark like:
> 
> ;;; -*- mode: Text; coding: iso-8859-4; -*-
> GNU Emacs will learn from that this text files contents is Northern ISO
> Latin.

Did not help either. I tried writing "æøå" in my ~/.bashrc, which I've
set to UTF-8 with C-x RET f RET utf-8 RET and tested with all the
aforementioned locales, but same result, æøå gets turned into something
like ¥...

I tried searching groups.google.com, but no help to find there :-(

Thanks for the help.
-- 
Mads Jensen - mail sent to address ends in /dev/null
	      s/spam/madsj for emailing me

Research is what I'm doing when I don't know what I'm doing.
                -- Wernher von Braun

Re: desktop and encodings

[Peter Petersen]

* Mads Jensen <spam@raptus.dk> schrieb:

> I tried setting the LANG variable to both da_DA, da_DA.utf-8,
> da_DA.iso-8859-1 plus en_US... (the same LANG variables as mentioned
> before, just with a different language), but that did not help :-/ 
>  
<...>
>
> Did not help either. I tried writing "æøå" in my ~/.bashrc, which I've
> set to UTF-8 with C-x RET f RET utf-8 RET and tested with all the
> aforementioned locales, but same result, æøå gets turned into something
> like ¥...


Hi,

>From a text console I launch Emacs like this:

LC_CTYPE=de_DE.utf8 exec filterm - dynafont emacs -nw --eval
"(set-keyboard-coding-system 'latin-9)"

(LC_CTYPE doesn't even seem to be needed...)

filterm is part of the "konwert" and "konwert-filters" packages, and
"dynafont" is one of its filters allowing me to read and reply to utf-8
texts (for instance when using the newsreader Gnus).

I described some problems with filterm recently, but was able to solve
them (now control-g works again!). filterm and dynafont switch your
console to utf-8 mode and also take care to chose a utf-8 capable screen
font etc.


This is just something that crossed my mind - I don't know if this fits
to your situation.


Let's see:

æ ø å

€

äöü

ß

éèê

ñ


;-)



Best regards from Germany
Peter Petersen

Re: desktop and encodings

[Peter Dyballa]

Am 23.05.2005 um 16:13 schrieb Mads Jensen:

> æøå gets turned into something like Â¥...
>

What see is the 'translation' of some ISO Latin encoding into UTF-8 and 
then displaying these double byte values as unibytes!

This could explain a bit:

;   oct   dec   hex    UCS2    UTF-8
;=====================================
   = 240 = 160 = A0 = U+00A0 =    C2 A0 : NO-BREAK SPACE
Ą = 241 = 161 = A1 = U+0104 =    C4 84 : LATIN CAPITAL LETTER A WITH 
OGONEK
ĸ = 242 = 162 = A2 = U+0138 =    C4 B8 : LATIN SMALL LETTER KRA
Ŗ = 243 = 163 = A3 = U+0156 =    C5 96 : LATIN CAPITAL LETTER R WITH 
CEDILLA
¤ = 244 = 164 = A4 = U+00A4 =    C2 A4 : CURRENCY SIGN
Ĩ = 245 = 165 = A5 = U+0128 =    C4 A8 : LATIN CAPITAL LETTER I WITH 
TILDE
Ļ = 246 = 166 = A6 = U+013B =    C4 BB : LATIN CAPITAL LETTER L WITH 
CEDILLA
§ = 247 = 167 = A7 = U+00A7 =    C2 A7 : SECTION SIGN
¨ = 250 = 168 = A8 = U+00A8 =    C2 A8 : DIAERESIS
Š = 251 = 169 = A9 = U+0160 =    C5 A0 : LATIN CAPITAL LETTER S WITH 
CARON
Ē = 252 = 170 = AA = U+0112 =    C4 92 : LATIN CAPITAL LETTER E WITH 
MACRON
Ģ = 253 = 171 = AB = U+0122 =    C4 A2 : LATIN CAPITAL LETTER G WITH 
CEDILLA
Ŧ = 254 = 172 = AC = U+0166 =    C5 A6 : LATIN CAPITAL LETTER T WITH 
STROKE
­ = 255 = 173 = AD = U+00AD =    C2 AD : HYPHEN-MINUS
Ž = 256 = 174 = AE = U+017D =    C5 BD : LATIN CAPITAL LETTER Z WITH 
CARON

Á = 301 = 193 = C1 = U+00C1 =    C3 81 : LATIN CAPITAL LETTER A WITH 
ACUTE
 = 302 = 194 = C2 = U+00C2 =    C3 82 : LATIN CAPITAL LETTER A WITH 
CIRCUMFLEX
à = 303 = 195 = C3 = U+00C3 =    C3 83 : LATIN CAPITAL LETTER A WITH 
TILDE
Ä = 304 = 196 = C4 = U+00C4 =    C3 84 : LATIN CAPITAL LETTER A WITH 
DIAERESIS
Å = 305 = 197 = C5 = U+00C5 =    C3 85 : LATIN CAPITAL LETTER A WITH 
RING ABOVE
Æ = 306 = 198 = C6 = U+00C6 =    C3 86 : LATIN CAPITAL LETTER AE

æ = 346 = 230 = E6 = U+00E6 =    C3 A6 : LATIN SMALL LETTER AE


First column contains the glyphs as they are, next columns have the 
glyph's byte value expressed as octal, decimal, or hexadecimal 
numerals. Next column, UCS2, show the slot number (ASCII code) of that 
glyph in Unicode (which, I think, is too the internal representation in 
GNU Emacs). The next column now shows into which bytes the glyphs from 
column 1 are translated as UTF-8. As you can see you can 'see' the 
UTF-8 bytes as 'normal' characters, a UTF-8 encoded æ is just 'ÄĻ' if 
displayed in ISO Latin-4, 'Ħ' in ISO Latin-1 ...

So, to conclude: your Emacs obviously saves your input as UTF-8, and 
you have to make the buffer display in UTF-8 too! The correct headers 
would look like

	;;; -*- mode: Text; coding: utf-8; -*-

Once you have the file opened in the wrong encoding you can change that 
with revert-buffer-with-coding-system, C-x RET r utf-8 RET.

Have you thought of

(prefer-coding-system     'utf-8-unix)

Could be it cures a lot. There is too (set-language-environment  
'Danish) ...


--
Mit friedvollen Grüßen

   Pete

In a world without walls and fences, who needs gates and windows?

Re: desktop and encodings

[Mads Jensen]

Peter Dyballa <Peter_Dyballa@Web.DE> writes:
[snip]

> So, to conclude: your Emacs obviously saves your input as UTF-8, and you
> have to make the buffer display in UTF-8 too! The correct headers would
> look like
> 
> 	;;; -*- mode: Text; coding: utf-8; -*-

Yeah, just found out about that; it really makes everything a whole lot
easier. 

> Once you have the file opened in the wrong encoding you can change that
> with revert-buffer-with-coding-system, C-x RET r utf-8 RET.

Thanks. There are so many things about Emacs, that I'm unfamiliar with,
so it's great learning about new stuff :-)

> (prefer-coding-system     'utf-8-unix)

I have both a (prefer-coding-system ...) for latin-1 and unicode, and
they appear to help a lot. Anyways, thanks!

-- 
Mads Jensen - mail sent to address ends in /dev/null
              s/spam/madsj for emailing me 
              gpg: 7E775BDA

It is impossible to make anything foolproof because fools are so
ingenious.