[developers] [spr30362] bug using climxm

[Andreas Fuchs]

Hi Ben,

Thank you very much for the additional information. With it, I think I
have located the root cause of the problem:

The default Xt language procedure looks only at the value of the
LC_ALL unix environment variable which isn't changed when changing
excl:*locale*. I am working on a fix that will do the right thing when
excl:*locale* is rebound.

Cheers,
Andreas.

On 2006-09-13, Ben Waldron <bmw20 at cl.cam.ac.uk> wrote:
> Hi Andreas -
>
> Thanks for this email. Unfortunately, the patch is not working for
> us:
>
> 1. I compile and load the patch, then load/compile our application
> 2. I open the "Parse Sentence" window. This is the window in which
>    it
> should be possible to enter text for the sentence to be parsed. Upon 
> opening this window I receive the following message:
>
> Warning: Missing charsets: ISO10646-1, GB2312.1980-0,
> KSC5601.1987-0, creating fontset for
>
> -misc-fixed-medium-r-normal--14-130-75-75-c-70-jisx0201.1976-0,-misc-fixed-medium-r-normal--14-130-75-75-c-140-jisx0208.1983-0,-adobe-helvetica-medium-r-normal--10-100-75-75-p-56-iso8859-1,
> 
> 3. I am able to enter non-latin characters if I run a command in the
> Emacs Lisp buffer. Eg.
>
> (do-parse-tty "læreren kommer på lørdag")
>
> 4. I try to paste the following into the Parse Sentence window:
>
> læreren kommer på lørdag
>
> All that appears in the window is a single lower case 'l'. I receive
> the following messages:
>
> Warning: Xt:
> Name: xmTextField
> Class: XmTextField
> Character '\346' not supported in font.  Discarded.
>
> Warning: Xt:
> Name: xmTextField
> Class: XmTextField
> Character '\345' not supported in font.  Discarded.
>
> Warning: Xt:
> Name: xmTextField
> Class: XmTextField
> Character '\370' not supported in font.  Discarded.
>
> HOWEVER when I click OK to send the contents of this window to our
> application the full text "læreren kommer på lørdag" is sent. The
> parse window for this sentence appears. This window displays
> correctly (the characters æåø are visible) except for the title. In
> the title "lreren kommer p lrdag" these non-latin characters are
> missing.
>
> 5. I open the Parse Sentence window again. I again receive the above
> 'Character ... nor supported' messages. I try to type the following 
> directly into the Parse Sentence window:
>
> læreren kommer
>
> The 'æ' character is not displayed (only "lreren kommer"). I click
> OK and receive the following message:
>
> Warning: Xt:
> Name: xmTextField
> Class: XmTextField
> Character '\346' not supported in font.  Discarded.
>
> The input received by our application is "lreren kommeræ".
>
>
> The initial locale when I start the Lisp is normally.
>
> CL-USER(2): excl:*locale*
> #<locale "C" [:LATIN1-BASE] @ #x441b482a>
>
> This is later reset to the following by our application:
>
> LKB(6): excl:*locale*
> #<locale "C" [:UTF8-BASE] @ #x44f2cc5a>
>
> However I get the same behaviour (described above) no matter whether
> the locale is set to "C" [:UTF8-BASE] after loading out application,
> before loading our application, or at Lisp startup (via the -locale
> command-line argument). I'm running Ubuntu Dapper linux.
>
> Thanks for the help,
> - Ben
>
> afuchs at franz.com wrote:
>> Hi Ben,
>>
>> I've been working on spr30362 (problems when entering non-ascii
>> text into text fields). I think I have come up with a fix that
>> allows me to enter at least german umlauts, cyrillic, and japanese
>> (Kanji, I believe) characters into text fields in a UTF-8
>> locale. Other characters (norwegian, spanish, ...) should work,
>> too.
>>
>> As this spr is related to spr30570, characters from Korean
>> encodings won't work yet. I'm working on this SPR, too.
>>
>> To test the fix, please compile & load the attached file before
>> running your application.
>>
>>
>> ------------------------------------------------------------------------
>>
>>
>> Regards, Andreas.