Re: introduction / first question and special characters
Hello Alain,
On 20:58 Wed 14 Sep , Alain Bench wrote:
> On Wednesday, September 14, 2005 at 5:56:18 PM +0200, Daniel Hertrich wrote:
>
> >> <URL:http://www.in-ulm.de/~mascheck/locale/checklocale.c>
> > I will test that.
>
> Real success of each locale category setting will be indicated,
> together with printability of each character.
>
Here is the output
[Latin1/9] If there's no real copyrightsymbol at the end of this sentence,
then your terminal/terminalemulator/font is not ISO8859-1/15 ready: ©
(Comment: The copyright sign looks good.)
- Current environment settings:
LANG = "de_DE.ISO-8859-1"
- Implicitly setting all locale categories with LANG failed.
You might want to unset/fix it now and/or set supported categories instead.
- Implicitly setting LC_CTYPE by LANG failed.
- Implicitly setting LC_NUMERIC by LANG failed.
- Implicitly setting LC_TIME by LANG failed.
- Implicitly setting LC_COLLATE by LANG failed.
- Implicitly setting LC_MONETARY by LANG failed.
- Implicitly setting LC_MESSAGES by LANG failed.
And setting LANG to "C" again, as it was set initially:
[Latin1/9] If there's no real copyrightsymbol at the end of this sentence,
then your terminal/terminalemulator/font is not ISO8859-1/15 ready: ©
- Current environment settings:
LANG = "C"
- Implicitly setting all locale categories with LANG succeeded.
Testing LC_CTYPE with isprint():
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
! " # $ % & ' ( ) * + , - . /
0 1 2 3 4 5 6 7 8 9 : ; < = > ?
@ A B C D E F G H I J K L M N O
P Q R S T U V W X Y Z [ \ ] ^ _
` a b c d e f g h i j k l m n o
p q r s t u v w x y z { | } ~ #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # #
- Testing LC_MESSAGES with perror(), but it's a libc message.
- Implicitly setting LC_CTYPE by LANG succeeded.
- Implicitly setting LC_NUMERIC by LANG succeeded.
- Implicitly setting LC_TIME by LANG succeeded.
- Implicitly setting LC_COLLATE by LANG succeeded.
- Implicitly setting LC_MONETARY by LANG succeeded.
- Implicitly setting LC_MESSAGES by LANG succeeded.
(Additionally, on stderr I get:
Error message for "EAGAIN": Resource temporarily
unavailable)
I have asked in the Zaurus forums for locale issues and
solutions. Hopefully someone will give a good hint.
> Libiconv is really needed by Mutt only when the libc doesn't provide
> iconv (very old libc), or when the provided iconv is ahum... suboptimal
> (say on HP/UX).
>
> Now, it is possible to install and use libiconv on all platforms,
> you just have to use some linker magic to force its usage without
> conflict with libc iconv. Libiconv's INSTALL gives hints.
The libiconv package I installed does not contain an
INSTALL. It only contains the following files:
//
//usr/bin/iconv
//usr/lib/libiconv.la
//usr/lib/libiconv.so.2.1.0
//usr/lib/libcharset.la
//usr/lib/libcharset.so.1.0.0
Can I consult the INSTALL of any available source package of
libiconv or are there differences?
> >> build Mutt with the same +LOCALES_HACK after you renamed out
> >> /hdd2/usr/include/iconv.h?
> > No obvious difference to before.
>
> ????!? OK: Now same renaming out, but rebuild with both options
> --enable-locales-fix and --without-wc-funcs.
Tried, but unsuccessfully (sorry, copy and paste messed up
line breaks):
In file included from ../protos.h:24, from
../mutt.h:889, from auth.c:27:
../mbyte.h:21: conflicting types for `wcwidth'
/mnt/card/devel/include/wchar.h:328: previous declaration of
`wcwidth' make[2]: *** [auth.o] Error 1 make[2]: Leaving
directory `/usr/mnt.rom/card/devel/projects/mutt-1.5.9/imap'
make[1]: *** [all-recursive] Error 1 make[1]: Leaving
directory `/usr/mnt.rom/card/devel/projects/mutt-1.5.9'
make: *** [all] Error 2
Any idea what causes that error?
> Of course you need to
> change muttrc to set charset=iso-8859-1 to test octalcodization.
>
> Describe and quote what you see here: ??
Currently (without the wcfuncs modiication), with us-asii I
see two question marks, with
ISO-8859-1 I see "backslash 3 4 4"
> BTW about your web setup:
>
> - Exporting TERM=linux for qkonsole is well suited? Grepping "sharp",
> "zaurus", or "qkons" in latest terminfo unsuccesfulled... TED?
I do not have any file named "terminfo" on my system.
I have a termcap where I found the "linux" console type in,
but none of the strings you suggested is inside that file.
> - export set TERM=linux
> ^^^
works. But "set" is superfluous you mean?
> - AFAICS your umlaut problem can't be an /etc/termcap issue.
okay.
> - Your subscribe regexps are too lax.
I have not understood completely what they are good for.
I'll reread the docs. How should they look like?
> - The $send_charset is broken, perhaps by HTMLization. Anyway let me
> suggest one in infosig.
The HTML export of the program I wrote that documentation in
is not entirely clean. I heard from someone that single
quotes appear as backquotes in my HTML. I think I'll some
day write my own HTML export routine for that program...
Well, rechecked on the web page... what is wrong with the
send_charset? The syntax follows the documentation of that
variable, doesn't it?
Please let me know.
I thought about another possible workaround for the Umlaut
problem: I could probably let all mails pass a filter which
converts the Umlauts to something like ae, oe, ue etc.
Any idea where such a filter (a simple sed command maybe)
could be placed? in frint of procmail, or better behind
procmail?...
Well, it is probably not that easy, because the filter
program had to deal with diffeent encodings in the mails.
Anyone here who uses such a filter already and could give
advice?
Thanks!
daniel
--
Dipl.-Ing. Daniel Hertrich
Reichertshofen, Germany
http://www.daniel-hertrich.de