CEGUI (Crazy Eddie's GUI) Forums

Hello again!

Here I have another issue:

I need to add not ANSI symbols (russian text) at StaticText/Text parameter:


In what format should I represent my string?

1) CEGUI::utf8 (unsigned char)
2) std::wstring?
3) ucs4?

Thank you!

P.S. Text, defined in .layout file works well.

I will assume that your editor is set to read and write the source code files in utf8. Then you just cast the string to utf8 and it should work.



CEGUI only supports utf8 and utf32, utf16 is not supported!

Kulik wrote:I will assume that your editor is set to read and write the source code files in utf8. Then you just cast the string to utf8 and it should work.

Code: Select all

_uiTalkingLog->appendText((CEGUI::utf8*)"Актер:\n");


CEGUI only supports utf8 and utf32, utf16 is not supported!

In MSVC2008 project "Characer Set" is "Unicode". Source file is encoded in utf-8. I see this ("*" is result of ):

Show the CEGUI Logger

VitaliBR wrote:Show the CEGUI Logger

P.S. I see something about glyph, but russian font works well.

wild_r wrote:

Kulik wrote:I will assume that your editor is set to read and write the source code files in utf8. Then you just cast the string to utf8 and it should work.
CEGUI only supports utf8 and utf32, utf16 is not supported!

In MSVC2008 project "Characer Set" is "Unicode". Source file is encoded in utf-8.

Wrong. If your MSVC project is set to "Unicode" then you are using UTF-16 not UTF-8. If you want to use UTF-8 you need to set your project to "Multi-Byte".

edit: see MSDN Unicode and MBCS. It seems "Unicode" does mean "UTF-16", however "Multi-Byte" actually means DBCS and not UTF-8; so you will have to use codepages if you use MBCS. I think your best bet is probably to use the "Unicode" setting, and then use WideCharToMultiByte to convert to UTF-8. I do know there have been others who have posted similar concerns on these forums, so you may want to do a little searching and see how they ended up handling this.

Jamarr wrote:

wild_r wrote:

Kulik wrote:I will assume that your editor is set to read and write the source code files in utf8. Then you just cast the string to utf8 and it should work.
CEGUI only supports utf8 and utf32, utf16 is not supported!

In MSVC2008 project "Characer Set" is "Unicode". Source file is encoded in utf-8.

Wrong. If your MSVC project is set to "Unicode" then you are using UTF-16 not UTF-8. If you want to use UTF-8 you need to set your project to "Multi-Byte".

edit: see MSDN Unicode and MBCS. It seems "Unicode" does mean "UTF-16", however "Multi-Byte" actually means DBCS and not UTF-8; so you will have to use codepages if you use MBCS. I think your best bet is probably to use the "Unicode" setting, and then use WideCharToMultiByte to convert to UTF-8. I do know there have been others who have posted similar concerns on these forums, so you may want to do a little searching and see how they ended up handling this.

I'm also used this function to convert wchar_t* strings to utf8:



It works bad. I don't know why. If somebody could help me with it, you are welcome.

P.S. Our projec tis pretty big and we couldn't set project settings to "multi byte".

P.P.S. I have searched in forum a one day, but only what I was found, was a thread, where somebody uses a WideCharToMultiByte function.

You cannot assume the size of the buffer is rWStr.size() * sizeof(wchar_t). You need to call the function twice: the first call is used to return the size of the buffer you need to contain the converted string, the second call populates the buffer. See an example here.

Jamarr wrote:You cannot assume the size of the buffer is rWStr.size() * sizeof(wchar_t). You need to call the function twice: the first call is used to return the size of the buffer you need to contain the converted string, the second call populates the buffer. See an example here.

Thank you! Very interesting site for searching. )

Okay, here is another problem:

I'm converting std::wstring into std::string (in utf8) using this code:



The code is:


The screenshot is:

As per the documentation (http://cegui.org.uk/docs/current/classC ... 9f2a281d20):

the documentation wrote:Note:
The characters of std_str are taken to be unencoded data which represent Unicode code points 0x00..0xFF. No translation of the provided data will occur.

You could take the c_str of the std::string and cast to CEGUI::utf8*.

IMO this whole approach of converting statically specified strings at runtime is completely over the top, though.

CE.

CrazyEddie wrote:As per the documentation (http://cegui.org.uk/docs/current/classC ... 9f2a281d20):

the documentation wrote:Note:
The characters of std_str are taken to be unencoded data which represent Unicode code points 0x00..0xFF. No translation of the provided data will occur.

You could take the c_str of the std::string and cast to CEGUI::utf8*.

IMO this whole approach of converting statically specified strings at runtime is completely over the top, though.

CE.

If i using non english text, I catch this exeption:



My code is:



If all this questions "is completely over the top", please, help me )

Okay, I could see that the error is appears again in ati open gl library: atioglxx.dll. One programmer in our team has driver version 9.7. And on his machine code works well. I'll try this driver and will write here my results.

Ati is insulting me.

P.S. Other comrade with latest nvidia driver had the same error. So, I think, CEGUI is not compatable with latest OpenGl extensions.

wild_r wrote:If all this questions "is completely over the top", please, help me )

The questions are not OTT, it's the 'solution' that is. For static strings, you just have to do, for example (and to re-iterate what was already said above, by others):

But the key is that the source text file must be encoded as UTF-8, if you see "ÐепиÑÑ:" when using the above code, this simply confirms that the file is encoded as latin-1 / ISO-8859-1 and not UTF-8.

I doubt GL extensions would cause a crash such as the one you're seeing, and there's a prize for someone who can demonstrate otherwise. Crashing in the driver might indicate an issue with multi-threading if you're using such a thing; you'll need to keep all your CEGUI code running on the same thread.

CE.

CrazyEddie wrote:

wild_r wrote:If all this questions "is completely over the top", please, help me )

The questions are not OTT, it's the 'solution' that is. For static strings, you just have to do, for example (and to re-iterate what was already said above, by others):

Code: Select all

_uiTalkingLog->appendText((const CEGUI::utf8*)"Непись:\n");

But the key is that the source text file must be encoded as UTF-8, if you see "ÐепиÑÑ:" when using the above code, this simply confirms that the file is encoded as latin-1 / ISO-8859-1 and not UTF-8.

I doubt GL extensions would cause a crash such as the one you're seeing, and there's a prize for someone who can demonstrate otherwise. Crashing in the driver might indicate an issue with multi-threading if you're using such a thing; you'll need to keep all your CEGUI code running on the same thread.

CE.

One guy in our team solve this problem. Yeah!

1) If in .layout file _uiTalkingLog "Text" field includes only english letters, we have "Access violation" error in function:





2) If in .layout file "Text" parameter is looks like this:



All works well!




====

I think, it means that a texture, that stores a images of letters, contents only taht characters sets, which existing in .lyaout file. Awesome!