Re: [2/2] gdi32/tests: Added tests for GetFontResourceInfoW

[Dmitry Timoshkov]

Stefan Leichter  wrote:
> - pCopyFileW = (void *)GetProcAddress(hkernel32, "CopyFileW");
> - pExpandEnvironmentStringsW = (void *)GetProcAddress(hkernel32, "ExpandEnvironmentStringsW");
> - pGetCurrentDirectoryW = (void *)GetProcAddress(hkernel32, "GetCurrentDirectoryW");

There is no reason to import these APIs using GetProcAddress. Also testing
GetFontResourceInfo without checking actual buffer contents is not very
useful.

Google points to http://www.undocprint.org/winspool/getfontresourceinfo which
seems to have some description of this API, with symbolic names of the type
parameter, so testing something besides type == 1 would be a good idea as well.

[2/2] gdi32/tests: Added tests for GetFontResourceInfoW

[Stefan Leichter]

Friday 14 December 2012 Dmitry Timoshkov
> Stefan Leichter  wrote:
> > - pCopyFileW = (void *)GetProcAddress(hkernel32, "CopyFileW");
> > - pExpandEnvironmentStringsW = (void *)GetProcAddress(hkernel32,
> >
> > "ExpandEnvironmentStringsW"); +    pGetCurrentDirectoryW = (void
> > *)GetProcAddress(hkernel32, "GetCurrentDirectoryW");
>
> There is no reason to import these APIs using GetProcAddress. Also testing
> GetFontResourceInfo without checking actual buffer contents is not very
> useful.

According to MSDN this APIs are WIn2k++ and WinXP++ only
> Google points to http://www.undocprint.org/winspool/getfontresourceinfo
> which seems to have some description of this API, with symbolic names of
> the type parameter, so testing something besides type == 1 would be a good
> idea as well.

The tests show what needs to be implemented to fix bug #30514. Feel free to 
send tests for other query types ;-)

[2/2] gdi32/tests: Added tests for GetFontResourceInfoW

[Dmitry Timoshkov]

Stefan Leichter  wrote:
> > > - pCopyFileW = (void *)GetProcAddress(hkernel32, "CopyFileW");
> > > - pExpandEnvironmentStringsW = (void *)GetProcAddress(hkernel32,
> > >
> > > "ExpandEnvironmentStringsW"); +    pGetCurrentDirectoryW = (void
> > > *)GetProcAddress(hkernel32, "GetCurrentDirectoryW");
> >
> > There is no reason to import these APIs using GetProcAddress. Also testing
> > GetFontResourceInfo without checking actual buffer contents is not very
> > useful.
>
> According to MSDN this APIs are WIn2k++ and WinXP++ only

That's not true. You can always send a test to testbot and see what happens.
> > Google points to http://www.undocprint.org/winspool/getfontresourceinfo
> > which seems to have some description of this API, with symbolic names of
> > the type parameter, so testing something besides type == 1 would be a good
> > idea as well.
>
> The tests show what needs to be implemented to fix bug #30514. Feel free to 
> send tests for other query types ;-)

Adding such tests isn't too hard, but anyway please check at least contents of
the buffer you get, and use symbolic names for query types.

[2/2] gdi32/tests: Added tests for GetFontResourceInfoW

[Stefan Leichter]

Friday 14 December 2012 Dmitry Timoshkov
> Stefan Leichter  wrote:
> > > > - pCopyFileW = (void *)GetProcAddress(hkernel32, "CopyFileW");
> > > > - pExpandEnvironmentStringsW = (void *)GetProcAddress(hkernel32,
> > > >
> > > > "ExpandEnvironmentStringsW"); +    pGetCurrentDirectoryW = (void
> > > > *)GetProcAddress(hkernel32, "GetCurrentDirectoryW");
> > >
> > > There is no reason to import these APIs using GetProcAddress. Also
> > > testing GetFontResourceInfo without checking actual buffer contents is
> > > not very useful.
> >
> > According to MSDN this APIs are WIn2k++ and WinXP++ only
>
> That's not true. You can always send a test to testbot and see what
> happens.
> > > Google points to http://www.undocprint.org/winspool/getfontresourceinfo
> > > which seems to have some description of this API, with symbolic names
> > > of the type parameter, so testing something besides type == 1 would be
> > > a good idea as well.
> >
> > The tests show what needs to be implemented to fix bug #30514. Feel free
> > to send tests for other query types ;-)
>
> Adding such tests isn't too hard, but anyway please check at least contents
> of the buffer you get, and use symbolic names for query types.

Do you know an offical header file containing the symbolic names? So fare i did 
not find any. Therefore the symbolic names are just fantasy names for me till 
now.

If there is no header file for the symbolic names do i have to define the 
symbolic names twice (dll and test) or do we have a common way to include a 
private header file of a dll into the tests ot the dll?

[2/2] gdi32/tests: Added tests for GetFontResourceInfoW

[Dmitry Timoshkov]

Stefan Leichter  wrote:
> > > > Google points to http://www.undocprint.org/winspool/getfontresourceinfo
> > > > which seems to have some description of this API, with symbolic names
> > > > of the type parameter, so testing something besides type == 1 would be
> > > > a good idea as well.
> > >
> > > The tests show what needs to be implemented to fix bug #30514. Feel free
> > > to send tests for other query types ;-)
> >
> > Adding such tests isn't too hard, but anyway please check at least contents
> > of the buffer you get, and use symbolic names for query types.
>
> Do you know an offical header file containing the symbolic names? So fare i did 
> not find any. Therefore the symbolic names are just fantasy names for me till 
> now.

I have no additional information except what google points to.
> If there is no header file for the symbolic names do i have to define the 
> symbolic names twice (dll and test) or do we have a common way to include a 
> private header file of a dll into the tests ot the dll?

Since it's a gdi32 API I'd guess that both the API declaration and related
defines should be in wingdi.h, but taking into account that it's undocumented
(yet) probably it's better to put defines in the same file that uses the API.