[cairo] Simple cairo-trace printing cleanups

[Chris Wilson]

On Thu, 2009-01-22 at 11:52 +1100, Carl Worth wrote:
> Hi Chris,
> 
> Here are two tiny patches to adjust what cairo-trace prints to
> stdout/stderr. It removes some prints about FT_Open_Face and
> adds a print of the filename actually logged to.
> 
> I would have just pushed without asking, but I wanted to check
> if the FT_Open_Face messages were anything more than leftover
> debug code that's just not needed anymore.

Thanks Carl,

I've applied these to my tree with minor modifications - just to be
awkward. ;-)

The FT_Open_Face() wrapping is currently not entirely satisfactory
(hence why I had left the debugging spew in there). At the moment, we
can capture the data for a FT_Face via FT_New_Face() and
FT_New_Memory_Face() but not for FT_Open_Face(). FT_Open_Face() is also
exposed, but we also (currently) intercept internal library calls to
FT_Open_Face(). This is just a complication that needs to worked-around,
as is capturing the stream data. However, no application I've traced so
far has called FT_Open_Face() directly (if it did it would generate a
segfault when cairo-trace tried to lookup the font) so I've not been
motivated to fix it. So I'll happily just comment out the wrapping and
leave stderr unmolested.
-ickle