Gives me a good 10% speedup for the Devanagari test case.  Less so
for less lookup-intensive tests.

For the Devanagari test case, the false positive rate of the GSUB digest
is 4%.

Implement two set digests, and one that combines the two.

Fallback mark positioning works now...  With hb-ft and hb-view /
hb-shape at least.

If there is no GPOS, zero mark advances.

If there *is* GPOS and the shaper requests so, zero mark advances for
attached marks.

Fixes regression with Tibetan, where the font has GPOS, and marks a
glyph as mark where it shouldn't get zero advance.

Enabled for all shapers except for Indic.

Still working on it.

Based on patch from Philip Withnall.
https://bugs.freedesktop.org/show_bug.cgi?id=41095

Add a shaper class struct.

When we removed the separate Hangul shaper, the specific normalization
preference of Hangul was lost.  Fix that.  Also, the Thai shaper was
copied from Hangul, so had the fully-composed normalization behavior,
which was unnecessary.  So, fix that too.