Help needed - CreateScalableFontResource

[Shachar Shemesh]

Hi list,

We have in Israel a lawyer who has cought on to the idea of open source, 
and is helping "The Cause"(tm) here quite a lot. As part of his 
activities he is wants to demonstrate a Hebrew law and precedance lookup 
program at a conference on Dec 2nd using Wine, and he asked for my help.

When trying to run the program, I get the following error:
sun@sun:~/c/PD_OR/EXE$ LANG=he_IL wine PDWIN G
Could not stat /mnt/fd0 (No such file or directory), ignoring drive A:
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.485, setting to 
0xdeadbeef
err:fixup:NE_LoadSegment No implementation for USER.491, setting to 
0xdeadbeef
fixme:win:stub_USER_490 stub
fixme:font:CreateScalableFontResourceA 
(0,"C:\PD_OR\FONTS/AMIR____.FOT","C:\PD_OR\FONTS/AMIR____.TTF",(null)): 
stub
fixme:font:CreateScalableFontResourceA 
(0,"C:\PD_OR\FONTS/AMIRBLD_.FOT","C:\PD_OR\FONTS/AMIRBLD_.TTF",(null)): 
stub
fixme:dialog:MSGBOX_OnInit task modal msgbox ! Not modal yet.

The application then goes on to display a error message saying (In 
hebrew) "The fonts for the application could not be found, please 
reinstall...".

The fonts are indeed available in C:\PD_OR\FONTS as TTF files. It seems 
the application calls CreateScalableFontResource, and then tries to 
register the fonts. Copying the TTF fonts to the fonts directory did not 
solve this problem.

I am looking for some solution, even a workaround, that will solve this 
problem for me. I have contact with the application's developers, and 
they are very interested in getting this to work (not interested enough 
to actually install Linux, but they are interested enough to answer 
technical questions about the app).

My questions:
A. Is there any way I can patch Wine to fool the app into thinking that 
part succeeds? Or is performing this test necessary, for some strange 
reason, in order to use the TTF fonts?
B. It MAY be possible to ask the developers to actually modify their 
code to a moderate degree, in order to get this app to work. Is there 
anything in particular I can ask them?
C. Is anyone working on this API? I don't know the first thing about 
font files and resources, and while I CAN learn, is someone with already 
existing knowledge can do something about it, it would be much preferable.

Thanks,

Shachar

[Huw D M Davies]

On Mon, Nov 25, 2002 at 10:27:55PM +0200, Shachar Shemesh wrote:
> Hi list,
>
> We have in Israel a lawyer who has cought on to the idea of open source, 
> and is helping "The Cause"(tm) here quite a lot. As part of his 
> activities he is wants to demonstrate a Hebrew law and precedance lookup 
> program at a conference on Dec 2nd using Wine, and he asked for my help.
>
> fixme:font:CreateScalableFontResourceA 
> (0,"C:\PD_OR\FONTS/AMIR____.FOT","C:\PD_OR\FONTS/AMIR____.TTF",(null)): 
> stub
> fixme:font:CreateScalableFontResourceA 
> (0,"C:\PD_OR\FONTS/AMIRBLD_.FOT","C:\PD_OR\FONTS/AMIRBLD_.TTF",(null)): 
> stub
> fixme:dialog:MSGBOX_OnInit task modal msgbox ! Not modal yet.
>
> The application then goes on to display a error message saying (In 
> hebrew) "The fonts for the application could not be found, please 
> reinstall...".
>
> The fonts are indeed available in C:\PD_OR\FONTS as TTF files. It seems 
> the application calls CreateScalableFontResource, and then tries to 
> register the fonts. Copying the TTF fonts to the fonts directory did not 
> solve this problem.
>
> I am looking for some solution, even a workaround, that will solve this 
> problem for me. I have contact with the application's developers, and 
> they are very interested in getting this to work (not interested enough 
> to actually install Linux, but they are interested enough to answer 
> technical questions about the app).
>
> My questions:
> A. Is there any way I can patch Wine to fool the app into thinking that 
> part succeeds? Or is performing this test necessary, for some strange 
> reason, in order to use the TTF fonts?
> B. It MAY be possible to ask the developers to actually modify their 
> code to a moderate degree, in order to get this app to work. Is there 
> anything in particular I can ask them?
> C. Is anyone working on this API? I don't know the first thing about 
> font files and resources, and while I CAN learn, is someone with already 
> existing knowledge can do something about it, it would be much preferable.

You could try hacking CreateScalableFontResource to return TRUE.  The
next thing the app will do is to call AddFontResource using the .FOT
filename, this will also fail so you might have to fix the return
value for that function too.  However if you have the TTF files in
c:\windows\fonts then hopefully the app will be happy.

CreateScalableFontResource is on my todo list.

Note that the app is also failing to load some 16bit user functions,
so even with the fonts fixed you may have more problems later on...

Huw.

[Shachar Shemesh]

That did not work. Both CreateScalableFontResource and AddFontResource 
work. The application then goes on to enumerate the fonts, and 
X11Drv.EnumDeviceFonts fails to detect any "Amir" fonts. Output of +all 
of relevant part is at the end.

Any idea why?

I will just mention that I am using fonts installed in the windows/Fonts 
directory.

Also, when investigating the font in pfaedit, it seems all text is 
defined in the user defined area. I'm not sure what it means. I tried 
converting the font to a "standard" one, but that did not solve the 
problem either.

Shachar

Huw D M Davies wrote:
> You could try hacking CreateScalableFontResource to return TRUE.  The
> next thing the app will do is to call AddFontResource using the .FOT
> filename, this will also fail so you might have to fix the return
> value for that function too.  However if you have the TTF files in
> c:\windows\fonts then hopefully the app will be happy.
>
> CreateScalableFontResource is on my todo list.

Not the problem so far.
> Note that the app is also failing to load some 16bit user functions,
> so even with the fonts fixed you may have more problems later on...
>
> Huw.

080b6888:trace:global:GlobalHandle16 01af
080b6888:Call GDI.330: ENUMFONTFAMILIES(0860,030f40da 
"Amir",0aaf0008,06b7d86c) ret=030f:070e ds=06b7
080b6888:Call kernel32.lstrcpynA(41472b92,403dba6a "Amir",00000020) 
ret=409ee486
080b6888:trace:string:lstrcpynA (0x41472b92, "Amir", 32)
080b6888:Ret  kernel32.lstrcpynA() retval=41472b92 ret=409ee486
080b6888:Call kernel32._EnterSysLevel(40a1b9a4) ret=409f2e47
080b6888:trace:win32:_EnterSysLevel (0x40a1b9a4, level 3): thread 
80b6888 (fs 0287, pid 1061) count before 0
080b6888:trace:win32:_EnterSysLevel (0x40a1b9a4, level 3): thread 
80b6888 (fs 0287, pid 1061) count after  1
080b6888:Ret  kernel32._EnterSysLevel() retval=0000006d ret=409f2e47
080b6888:trace:gdi:GDI_GetObjPtr (0x860): enter 1
080b6888:Call kernel32.MultiByteToWideChar(00000000,00000000,41472b92 
"Amir",ffffffff,414729fc,00000020) ret=409ed821
080b6888:Ret  kernel32.MultiByteToWideChar() retval=00000005 ret=409ed821
080b6888:Call kernel32.MapLS(41472af0) ret=409ee256
080b6888:Ret  kernel32.MapLS() retval=09d72af0 ret=409ee256
080b6888:Call kernel32.MapLS(41472a3c) ret=409ee25f
080b6888:Ret  kernel32.MapLS() retval=09d72a3c ret=409ee25f
080b6888:Call 
x11drv.EnumDeviceFonts(4149f1b0,414729e0,409ee02c,41472b34) ret=409ee27f
080b6888:Ret  x11drv.EnumDeviceFonts() retval=00000000 ret=409ee27f
080b6888:Call kernel32.UnMapLS(09d72af0) ret=409ee28e
080b6888:Ret  kernel32.UnMapLS() retval=00000001 ret=409ee28e
080b6888:Call kernel32.UnMapLS(09d72a3c) ret=409ee296
080b6888:Ret  kernel32.UnMapLS() retval=00000001 ret=409ee296
080b6888:trace:gdi:GDI_ReleaseObj (0x860): leave 1
080b6888:Call kernel32._LeaveSysLevel(40a1b9a4) ret=409f2fe3
080b6888:trace:win32:_LeaveSysLevel (0x40a1b9a4, level 3): thread 
80b6888 (fs 0287, pid 1061) count before 1
080b6888:trace:win32:_LeaveSysLevel (0x40a1b9a4, level 3): thread 
80b6888 (fs 0287, pid 1061) count after  0
080b6888:Ret  kernel32._LeaveSysLevel() retval=0000006d ret=409f2fe3
080b6888:trace:global:GlobalHandle16 01af
080b6888:Ret  GDI.330: ENUMFONTFAMILIES() retval=0000 ret=030f:070e ds=06b7
080b6888:trace:global:GlobalHandle16 0137

[Huw D M Davies]

On Tue, Nov 26, 2002 at 08:28:57PM +0200, Shachar Shemesh wrote:
> That did not work. Both CreateScalableFontResource and AddFontResource 
> work. The application then goes on to enumerate the fonts, and 
> X11Drv.EnumDeviceFonts fails to detect any "Amir" fonts. Output of +all 
> of relevant part is at the end.
>
> Any idea why?

Yup, EnumFonts16 is broken.  Could you give this untested patch a go?

Huw.

[Zsolt Rizsanyi]

On Tuesday 26 November 2002 21:15, Huw D M Davies wrote:
> On Tue, Nov 26, 2002 at 08:28:57PM +0200, Shachar Shemesh wrote:
> > That did not work. Both CreateScalableFontResource and AddFontResource
> > work. The application then goes on to enumerate the fonts, and
> > X11Drv.EnumDeviceFonts fails to detect any "Amir" fonts. Output of +all
> > of relevant part is at the end.
> >
> > Any idea why?
>
> Yup, EnumFonts16 is broken.  Could you give this untested patch a go?

I have tested the patch and it does fixes a problem in a win16 app. It always 
popped up dialog boxes to report that they do not exist. But since it has 
used them, I never bothered enough to report.

Regards
Zsolt

[Shachar Shemesh]

Huw D M Davies wrote:
> On Tue, Nov 26, 2002 at 08:28:57PM +0200, Shachar Shemesh wrote:
> > That did not work. Both CreateScalableFontResource and AddFontResource 
> > work. The application then goes on to enumerate the fonts, and 
> > X11Drv.EnumDeviceFonts fails to detect any "Amir" fonts. Output of +all 
> > of relevant part is at the end.
> >
> > Any idea why?
>
> Yup, EnumFonts16 is broken.  Could you give this untested patch a go?
>
> Huw.

Works great for me. Thanks!!

Is there any chance you'll submit the official fix to be included by the 
time we have to present the software (Monday - Sunday evening if you are 
in the US, as we are ahead of the US). I would like to be able to say 
that "this is the out of the box Wine".

Shachar
> Index: objects/font.c
>
> RCS file: /home/wine/wine/objects/font.c,v
> retrieving revision 1.93
> diff -u -r1.93 font.c
> --- objects/font.c	25 Nov 2002 21:09:49 -0000	1.93
> +++ objects/font.c	26 Nov 2002 20:11:51 -0000
> @@ -71,6 +71,7 @@
>   LPENUMLOGFONTEX16     lpLogFont;
>   SEGPTR                segTextMetric;
>   SEGPTR                segLogFont;
> - DWORD                 dwFlags;
> HDC                   hdc;
> DC                   *dc;
> PHYSDEV               physDev;
>
> @@ -524,6 +525,7 @@
>     {
>         FONT_EnumLogFontExWTo16(plf, pfe->lpLogFont);
>    FONT_NewTextMetricExWTo16(ptm, pfe->lpTextMetric);
> -
>    pfe->dwFlags |= ENUM_CALLED;
>   GDI_ReleaseObj( pfe->hdc );  /* release the GDI lock */
>
>   ret = FONT_CallTo16_word_llwl( pfe->lpEnumFunc, pfe->segLogFont, pfe->segTextMetric,
>
> @@ -590,36 +592,49 @@
>                                    DWORD dwFlags)
> {
>     fontEnum16 fe16;
> - INT16	retVal = 0;
> -
>
> INT16	ret = 1, ret2;
> DC* 	dc = DC_GetDCPtr( HDC_32(hDC) );
> -
>
> NEWTEXTMETRICEX16 tm16;
> -
>
> ENUMLOGFONTEX16 lf16;
> -
>
> LOGFONTW lfW;
> -
>
> BOOL enum_gdi_fonts;
>
> if (!dc) return 0;
> -
>
> FONT_LogFont16ToW(plf, &lfW);
> -
>
> fe16.hdc = HDC_32(hDC);
>   fe16.dc = dc;
>   fe16.physDev = dc->physDev;
> -
>
> fe16.lpLogFontParam = plf;
> -
>
> fe16.lpEnumFunc = efproc;
> -
>
> fe16.lpData = lParam;
> -
>
> fe16.lpTextMetric = &tm16;
> -
>
> fe16.lpLogFont = &lf16;
> -
>
> fe16.segTextMetric = MapLS( &tm16 );
> -
>
> fe16.segLogFont = MapLS( &lf16 );
> -
>
> fe16.dwFlags = 0;
> -
> -
>
> enum_gdi_fonts = GetDeviceCaps16(hDC, TEXTCAPS) & TC_VA_ABLE;
> - if (dc->funcs->pEnumDeviceFonts)
> - if (!dc->funcs->pEnumDeviceFonts && !enum_gdi_fonts)
> {
> -
>    NEWTEXTMETRICEX16 tm16;
> -
>    ENUMLOGFONTEX16 lf16;
> -
>    LOGFONTW lfW;
> -
>    FONT_LogFont16ToW(plf, &lfW);
> -
> -
>    fe16.lpLogFontParam = plf;
> -
>    fe16.lpEnumFunc = efproc;
> -
>    fe16.lpData = lParam;
> -
>    fe16.lpTextMetric = &tm16;
> -
>    fe16.lpLogFont = &lf16;
> -
>    fe16.segTextMetric = MapLS( &tm16 );
> -
>    fe16.segLogFont = MapLS( &lf16 );
> -
> -
>    retVal = dc->funcs->pEnumDeviceFonts( dc->physDev, &lfW,
> -
>                                          FONT_EnumInstance16, (LPARAM)&fe16 );
> -
>    UnMapLS( fe16.segTextMetric );
> -
>    UnMapLS( fe16.segLogFont );
> -
>    ret = 0;
> -
>    goto done;
> }
> -
> - if (enum_gdi_fonts)
> -
>    ret = WineEngEnumFonts( &lfW, FONT_EnumInstance16, (LPARAM)&fe16 );
> - fe16.dwFlags &= ~ENUM_CALLED;
> - if (ret && dc->funcs->pEnumDeviceFonts) {
> - ret2 = dc->funcs->pEnumDeviceFonts( dc->physDev, &lfW, FONT_EnumInstance16, (LPARAM)&fe16 );
> - if(fe16.dwFlags & ENUM_CALLED) /* update ret iff a font gets enumed */
> -
>    ret = ret2;
> - }
>
> +done:
> - UnMapLS( fe16.segTextMetric );
> - UnMapLS( fe16.segLogFont );
> if (fe16.hdc) GDI_ReleaseObj( fe16.hdc );
> - return retVal;
> - return ret;
>
> }
>
> /***********************************************************************

[Huw D M Davies]

On Wed, Nov 27, 2002 at 07:08:08PM +0200, Shachar Shemesh wrote:
> Works great for me. Thanks!!
>
> Is there any chance you'll submit the official fix to be included by the 
> time we have to present the software (Monday - Sunday evening if you are 
> in the US, as we are ahead of the US). I would like to be able to say 
> that "this is the out of the box Wine".

I've sent the EnumFont16 fix off to wine-patches.

Huw.

[Shachar Shemesh]

Huw D M Davies wrote:
> On Wed, Nov 27, 2002 at 07:08:08PM +0200, Shachar Shemesh wrote:
> > Works great for me. Thanks!!
> >
> > Is there any chance you'll submit the official fix to be included by the 
> > time we have to present the software (Monday - Sunday evening if you are 
> > in the US, as we are ahead of the US). I would like to be able to say 
> > that "this is the out of the box Wine".
>
> I've sent the EnumFont16 fix off to wine-patches.
>
> Huw.

Thanks.

[Shachar Shemesh]

I will test the patch tonight.

I will just mention that I got permission to give the fonts in question 
to the subscribers of this list for the purpose of making this app work. 
It's useful when the person most interested in this app on Linux is the 
copyright holder's lawer..... :-)

Anyone who thinks having the actual Font will help is encouraged to 
contact me via email, and I'll send it.

On a related note - though I couldn't use the font from Wine in an 
unrelated app, I could use it in Windows. There is one, possibly 
important, difference, however. While in Wine I installed the font, on 
Windows I opened my test app while the other app was running (and thus - 
exporting the font).

Shachar

Shachar Shemesh wrote:
> That did not work. Both CreateScalableFontResource and AddFontResource 
> work. The application then goes on to enumerate the fonts, and 
> X11Drv.EnumDeviceFonts fails to detect any "Amir" fonts. Output of 
> +all of relevant part is at the end.
>
> Any idea why?
>
> I will just mention that I am using fonts installed in the 
> windows/Fonts directory.
>
> Also, when investigating the font in pfaedit, it seems all text is 
> defined in the user defined area. I'm not sure what it means. I tried 
> converting the font to a "standard" one, but that did not solve the 
> problem either.
>        Shachar
>
>
> Huw D M Davies wrote:
> > You could try hacking CreateScalableFontResource to return TRUE.  The
> > next thing the app will do is to call AddFontResource using the .FOT
> > filename, this will also fail so you might have to fix the return
> > value for that function too.  However if you have the TTF files in
> > c:\windows\fonts then hopefully the app will be happy.
> >
> > CreateScalableFontResource is on my todo list.
>
> Not the problem so far.
> > Note that the app is also failing to load some 16bit user functions,
> > so even with the fonts fixed you may have more problems later on...
> >
> > Huw.
>
> 080b6888:trace:global:GlobalHandle16 01af
> 080b6888:Call GDI.330: ENUMFONTFAMILIES(0860,030f40da 
> "Amir",0aaf0008,06b7d86c) ret=030f:070e ds=06b7
> 080b6888:Call kernel32.lstrcpynA(41472b92,403dba6a "Amir",00000020) 
> ret=409ee486
> 080b6888:trace:string:lstrcpynA (0x41472b92, "Amir", 32)
> 080b6888:Ret  kernel32.lstrcpynA() retval=41472b92 ret=409ee486
> 080b6888:Call kernel32._EnterSysLevel(40a1b9a4) ret=409f2e47
> 080b6888:trace:win32:_EnterSysLevel (0x40a1b9a4, level 3): thread 
> 80b6888 (fs 0287, pid 1061) count before 0
> 080b6888:trace:win32:_EnterSysLevel (0x40a1b9a4, level 3): thread 
> 80b6888 (fs 0287, pid 1061) count after  1
> 080b6888:Ret  kernel32._EnterSysLevel() retval=0000006d ret=409f2e47
> 080b6888:trace:gdi:GDI_GetObjPtr (0x860): enter 1
> 080b6888:Call kernel32.MultiByteToWideChar(00000000,00000000,41472b92 
> "Amir",ffffffff,414729fc,00000020) ret=409ed821
> 080b6888:Ret  kernel32.MultiByteToWideChar() retval=00000005 ret=409ed821
> 080b6888:Call kernel32.MapLS(41472af0) ret=409ee256
> 080b6888:Ret  kernel32.MapLS() retval=09d72af0 ret=409ee256
> 080b6888:Call kernel32.MapLS(41472a3c) ret=409ee25f
> 080b6888:Ret  kernel32.MapLS() retval=09d72a3c ret=409ee25f
> 080b6888:Call 
> x11drv.EnumDeviceFonts(4149f1b0,414729e0,409ee02c,41472b34) ret=409ee27f
> 080b6888:Ret  x11drv.EnumDeviceFonts() retval=00000000 ret=409ee27f
> 080b6888:Call kernel32.UnMapLS(09d72af0) ret=409ee28e
> 080b6888:Ret  kernel32.UnMapLS() retval=00000001 ret=409ee28e
> 080b6888:Call kernel32.UnMapLS(09d72a3c) ret=409ee296
> 080b6888:Ret  kernel32.UnMapLS() retval=00000001 ret=409ee296
> 080b6888:trace:gdi:GDI_ReleaseObj (0x860): leave 1
> 080b6888:Call kernel32._LeaveSysLevel(40a1b9a4) ret=409f2fe3
> 080b6888:trace:win32:_LeaveSysLevel (0x40a1b9a4, level 3): thread 
> 80b6888 (fs 0287, pid 1061) count before 1
> 080b6888:trace:win32:_LeaveSysLevel (0x40a1b9a4, level 3): thread 
> 80b6888 (fs 0287, pid 1061) count after  0
> 080b6888:Ret  kernel32._LeaveSysLevel() retval=0000006d ret=409f2fe3
> 080b6888:trace:global:GlobalHandle16 01af
> 080b6888:Ret  GDI.330: ENUMFONTFAMILIES() retval=0000 ret=030f:070e 
> ds=06b7
> 080b6888:trace:global:GlobalHandle16 0137