[Freetype] Problems with pcf format using freetype in win32 env

[Kalon]

Hi,

I am trying to use Freetype2 in a Win32 environment.  The font that I am  
trying to use is Roman Czyborra's unifont.  I used his utility to convert  
the font from 'hex' format to 'bdf' format.  Using this "unifont.bdf"  
works fine with Freetype2.  The file opens correctly and it renders  
correctly with the FT_Load_Char function.

But in order to conserve space I used a linux machine to convert the 'bdf'  
file to a 'pcf' file using the command:
        bdftopcf -o unifont.pcf unifont.bdf
Then I compressed the font using:
        gzip -9 unifont.pcf
I now try to use this "unifont.pcf.gz" file with freetype2 and it does not  
work correctly.  The file opens just fine but FT_Load_Char seems to return  
erronious data in the face structure.  I wondered if maybe it was the  
gzipped format so I tried again with the "unifont.pcf" file but had the  
same trouble (file loads correctly but FT_Load_Char does not return  
correct face data).  (Detailed information below.)

I don't know how much information is needed to help me figure out what is  
going wrong.  I've included more technical details below.  Seeing that the  
'pcf' and 'pcf.gz' return a different pitch for the glyph than the 'bdf'  
format, I wonder if it is just an improper use of the bdftopcf program due  
to padding or something.  If anyone can please help me out I would really  
appreciate it.

FYI, my ultimate goal is to correctly render a unicode string using a  
small, portable font that contains most unicode characters.  Roman  
Czyborra's unifont contains all of the characters I need to be able to  
render and I don't need multiple font sizes.  None of my files contain any  
combining characters.  If there are better ways to doing this then I would  
appreciate that information also.

Thank you very much,
Kalon Mills



(Technical Infomation)
OS:   WinXP
IDE:  Visual Studio .Net 2002
Freetype2 version:  2.1.9
Freetype2 Library Build Type:  Default options compiled in Single Thread  
mode

*Loading unicode character 0x062D when using 'bdf' file*  (correct version)
 (face->glyph->bitmap)
        --    bitmap       {...}       FT_Bitmap_
         |-   rows         0x00000010  int
         |-   width        0x00000008  int
         |-   pitch        0x00000001  int
         |-+  buffer       0x00a77570  unsigned char *
         |-   num_grays    0x0000      short
         |-   pixel_mode   0x01        char
         |-   palette_mode 0           char
         |-   palette      0x00000000  void *
 face->glyph->bitmap.buffer contains correct glyph information.


*Loading unicode character 0x062d when using 'pcf' file*  (INCORRECT)
 (face->glyph->bitmap)
        --    bitmap       {...}       FT_Bitmap_
         |-   rows         0x00000010  int
         |-   width        0x00000008  int
         |-   pitch        0x00000004  int
         |-+  buffer       0x0096d3e8  unsigned char *
         |-   num_grays    0x0001      short
         |-   pixel_mode   0x01        char
         |-   palette_mode 0           char
         |-   palette      0x00000000  void *
 face->glyph->bitmap.buffer contains incorrect glyph information.  It  
conatins the glyph for character 0x072B and seems to be padded to 4 times  
the size of what it should be.


*Loading unicode character 0x062D when using 'pcf.gz' file*  (INCORRECT)
 (face->glyph->bitmap)
        --    bitmap       {...}       FT_Bitmap_
         |-   rows         0x00000010  int
         |-   width        0x00000008  int
         |-   pitch        0x00000004  int
         |-+  buffer       0x0096d3e8  unsigned char *
         |-   num_grays    0x0001      short
         |-   pixel_mode   0x01        char
         |-   palette_mode 0           char
         |-   palette      0x00000000  void *
 face->glyph->bitmap.buffer contains incorrect glyph information.  It  
conatins the glyph for character 0x0F42 and seems to be padded to 4 times  
the size of what it should be.