Hi Fly_space,
Glad to see that you do not give up!
Sorry! My explanations were not enough explicit and the available documentation that I pointed is rather short.
I'll try to give some precisions on how to do that step by step.
This is is true but it is a problem only if you work with linux.
Most probably you have Windows and in this case the important link is the following:
1) download this archive and extract the files in a directory of your choice. Say "txffonts".
- This directory contains now 5 files. (have a look at the readme.txt)
2) choose the korean font that contains all the characters you need, including the missing characters.
The easiest is to take the font used by your text editor to show the text of the script. This way you are sure that the required characters are present.
I call this font "font.ttf". You have to replace "font.ttf" by the actual name of that font.
3) copy "font.ttf" from the fonts directory of Windows to your newly created "txffont" directory
4) build the font
The font to build is sansbold20_ko.txf (the _LL means the 2 letters of the locale font)
Don't bother with sans12 and sans14. These fonts are not used to display scripts and have the required characters to display what is worth.
The command line
ttf2txf -w 1048 -h 1048 -f codepoints_LL.txt -s 20 -o sansbold20_LL.txf font_bold.ttf
becomes a little more explicit:
- ttf2txf is the conversion program
- sansbold20_LL.txf must be replaced by sansbold20_ko.txf
- font_bold.ttf must be replaced by your own korean ttf (true type) font
- maybe the given values for the font size(1048, 1048) are too small for an asian font and we'll have to find better values
But the codepoints_LL.txt remains mysterious. It is only a txt file that contains the list of the caracters' codes to include in the font.
codepoints_ko.txt contains "only" the 487 characters that are necessary to ensure the translation of Celestia. If your script uses other characters they will be missing.
The problem is to find a replacement file for codepoints_ko.txt. This is why I proposed the "coreen2.txt" in my previous post. This file contains 11456 codes of characters. Maybe it is too much but I cannot read korean and I cannot say more on this. Any help is welcome.
Anyway if you agree to use my coreen2.txt file instead of codepoints_ko.txt:
- copy coreen2.txt into the "txffonts" directory
- open a command window in this directory
- try to type the command line by replacing the parameters. In my example I use arialuni.ttf as korean source font and I get this:
buildfont.png
As you can see, the font is generated but the error message indicates that the texture is too small and then only 2486 characters (instead of 115456 as expected) are taken into account.
The solution is to increase the size of the texture. With -w 2480 and -h 2048 I get:
buildfont2.png
This time there is no error and the font is created correctly.
The correct values depend on the font you have chosen.
You have to experiment a bit but I suppose that you can take these value as a starting point.
- copy the newly created sansbold20_ko.txf in the "fonts" directory of celestia and you are done!
PS
This is a kind of tutorial on how to build a txf font but not a definitive solution.
? propos... What is the original font used to generate the Celestia's korean font?
Who knows? Chris? Vincent? an idea for an effective solution?