[ft] bitmap_left and bitmap_top equivalent when using FT_Stroker?

[Jamie Dale]

Hey all,

We're using a mix of FreeType 2.4 and 2.6, and are seeing some artefacts
when using the bitmap_left and bitmap_top values from the glyph slot with a
glyph rendered via the FT_Stroker API.

The artefacts mostly seem to affect small and light fonts (which is likely
when FreeType gets a bit clever with its hinting), and they result in a
glyph appearing slightly above the baseline.

Historically our font rendering just used FT_Render_Glyph, however we
recently added outline font support so started using the FT_Stroker API to
handle that, before calling FT_Render_Glyph to populate the bitmap metrics.

If I memcmp the result of using FT_Render_Glyph vs. using the FT_Stroker
API I do see that the resultant bitmaps are different, which is likely why
the bitmap metrics produce incorrect results when used with the result from
FT_Stroker API? (or perhaps we're using the FT_Stroker API wrong and
they're supposed to be equivalent?).

Given the difference, I would assume there's some kind of bitmap_left and
bitmap_top equivalent when using the FT_Stroker API? If so I've not been
able to find it from the docs or API.

Thanks,
Jamie.