Home » U++ Library support » U++ Libraries and TheIDE: i18n, Unicode and Internationalization » Locale troubles with UPP
Locale troubles with UPP [message #1416] |
Fri, 03 March 2006 04:26 |
lundman
Messages: 175 Registered: March 2006 Location: Tokyo
|
Experienced Member |
|
|
We have come across some locale issues when trying to use UPP.
On Windows XP, runnning in Japanese locale, we can not compile the Core due to:
1. warnings from msc 8 about wrong codepage for .iml and .t files
2. codepage 1252 strings interpreted as containing newlines by codepage 932, resulting in errors
In particular, IMAGE_SCAN part.
You can change the locale to English to compile, but that is not a solution for us as many of our Japanese apps will then fail. Alas, make clean also cleans Core!
These issues do not occur if you use mingw.
C# Compiler has a /codepage= switch, but vc8 does not.
3. (low severity) NetBSD with LC_CTYPE=ja_JP.UTF-8
All applications (exmples) with a layout, seems to double the width of all widgets, making the frame double in width beyond my desktop. Can not be Zoomed smaller (in width). Unsetting locale result in normal behavior.
|
|
|
|
Re: Locale troubles with UPP [message #1421 is a reply to message #1416] |
Fri, 03 March 2006 10:07 |
|
mirek
Messages: 14039 Registered: November 2005
|
Ultimate Member |
|
|
[quote title=lundman wrote on Thu, 02 March 2006 22:26]
We have come across some locale issues when trying to use UPP.
On Windows XP, runnning in Japanese locale, we can not compile the Core due to:
1. warnings from msc 8 about wrong codepage for .iml and .t files
2. codepage 1252 strings interpreted as containing newlines by codepage 932, resulting in errors
In particular, IMAGE_SCAN part.
You can change the locale to English to compile, but that is not a solution for us as many of our Japanese apps will then fail. Alas, make clean also cleans Core!
These issues do not occur if you use mingw.
C# Compiler has a /codepage= switch, but vc8 does not.
[/quite]
In fact, this is one thing I am totally puzzled about MSC8. I have encountered this problem in another context already, seems like MSC8 decided to do some quite strange games with local and character literals (while all other compilers known to me simply take the content of character literal as it is and as expected).
Please, I would need somebody to investigate this strange issue!
Quote: |
3. (low severity) NetBSD with LC_CTYPE=ja_JP.UTF-8
All applications (exmples) with a layout, seems to double the width of all widgets, making the frame double in width beyond my desktop. Can not be Zoomed smaller (in width). Unsetting locale result in normal behavior.
|
Hm, interesting moment - U++ meets far east...
Mirek
|
|
|
|
|
|
Re: Locale troubles with UPP [message #1481 is a reply to message #1478] |
Mon, 06 March 2006 07:53 |
|
mirek
Messages: 14039 Registered: November 2005
|
Ultimate Member |
|
|
lundman wrote on Mon, 06 March 2006 00:28 |
I went back to this. They recommended applocale, which might work, if you run XP, but is not supported on Win2k.
I tried various methods of wrapping IDE, etc, with .bat files that first set the codepage, #pragma for codepage, and setlocale() but nothing lets me compile upp.
We are forced to stay 7.1 until MS can solve it.
Lund
|
After some research... have you tried
#pragma setlocale("C")
?
Place it somewhere into Core/Core.h or Core/Defs.h...
This info indicates it should help...:
http://www.openoffice.org/issues/show_bug.cgi?id=36782
I would gladly place it into Core.h....
Mirek
[Updated on: Mon, 06 March 2006 07:57] Report message to a moderator
|
|
|
Goto Forum:
Current Time: Fri Sep 20 13:51:30 CEST 2024
Total time taken to generate the page: 0.03492 seconds
|