Resolves https://github.com/behdad/fonttools/issues/355
For making sure that `pyftsubset` still works after this change,
I have done the following steps:
* invoked Adobe's `makeotf` tool to build a custom font with a
MultipleSubst lookup. This lookup decomposes two two ligatures,
`c_t` and `f_f_i`, into their respective components.
* invoked the `pyftsubset` tool to produce a subset font with just
the `c_t` ligature;
* checked with `ttx` that the newly produced subset font contains
the requested `c_t` ligature and its components `c` and `t`,
but does not contain not any of `f_f_i`, `f`, or `i`.
Before this change, the XML output for GSUB lookups of type 3
was not deterministic; multiple runs of TTX on the same font
could therefore emit the alternates in a different order.
Since alternate glyphs are sets, this change makes no semantic
difference to the output. However, a deterministic ordering
is needed for tests that compare GSUB tables in TTX format.
I had broken this in e06166b83aa82582dab2e011520b1a77f72aa9f9.
Fix that, and other issues. Since I made that last commit, it has
become clear that SingleSubst works by adding delta and then doing
module 65536.
HarfBuzz was updated for this in 2011:
commit 52ebdff49d13f239efc886de935d47be9860f6e5
Author: Behdad Esfahbod <behdad@behdad.org>
Date: Tue Sep 27 12:38:16 2011 -0400
Fix GSUB lookuptype 1 subtype 1 delta wrapping
The format values for those are automatically handled in
postRead/preWrite to choose optimal format. As such, don't write them
in XML. Reduces noise.
Part of https://github.com/behdad/fonttools/issues/73
The latter hits the __eq__ method and can fail because we now
do not allow comparing objects of different types.
For example, was failing subsetting Andika-R.ttf.
All types of FeatureParams are correctly handled now.
The only thing not handled is broken fonts generated with the
old AFDKO that have their FeatureParams offset computed wrong.
I don't currently plan on handling those.
Fixes https://github.com/behdad/fonttools/issues/38
Historically, Extension lookups were only decompiled upon access,
and decompiled Extension lookups were possibly copied verbatim to
output when compiling. This discrepancy with all other table types
is confusing and undeed. The possible time saving is not worth the
loss in possible optimization of table data. Remove.
Apparently the SIL Gentium fonts have wrong sorting in Coverage
tables. That's broken by the spec, but shapers still work with
such fonts. So, we warn when reading, and when writing, such
out-of-order coverage tables.
