Re: Problems with linearHoriAdvance

[David Turner]

Hello Thomas,

> 
> I am trying to calculate escapements for each glyph
> as a floating point value that when multiplied by
> the point size will return the width in pixels.
> 
> It seems that linearHoriAdvance is not correct:
> 
> (gdb) p *mFace->glyph
> $2 = {library = 0x8601e10, face = 0x86aa4e8, next = 0x0, flags = 0,
> generic = {data = 0x0, finalizer = 0}, metrics = {width = 0, height = 0,
> horiBearingX = 0, horiBearingY = 0, horiAdvance = 180, vertBearingX = 0,
> vertBearingY = 0, vertAdvance = 2379}, linearHoriAdvance = 481,
> linearVertAdvance = 2379, advance = {x = 180, y = 0},
> 
> The docs say that this value should be in 16.16 format, but the
> value of 481 above doesn't seem meaningful in this context.
> 
> I did notice the line in ahhint.c:
> ahhint.c:      /* XXX: TO DO - slot->linearHoriAdvance */
> 
> Is there something missing here or am I doing something wrong?
> 

This was a bug. The value returned was always in font units and
never scaled to 16.16 fixed floats. It has been fixed in the
current snapshot. Please feel free to upgrade..

Thanks,

- David


> Thanks,
> Thomas