When the tests are run with `python setup.py test` (e.g. from the ./run-tests.sh
script), we can't have setuptools install pre-compiled wheels for extra test
requirements (e.g. Brotli, unicodedata2).
Instead of failing, it's better that we skip the tests that have such extra
requirements.
NOTE: The recommended way to run the full test suite is by using tox
(like we do on the CI), or by installing pytest and the rests of requirements.txt,
and then run `pytest` command directly.
These originated from TTFs that were built from the test UFO masters using fontmake v1.2.0 with the command `fontmake -m BuildTest.designspace -o variable`
* TupleVariationTest.test_compileDeltaValues(): also test floats
* TupleVariation: round deltas before encoding
Python 3 was raising 'struct.error: required argument is not an integer' and Python 2 was truncating when deltas are floats
Specifically, contextual positioning statements with in-line single
positioning rules, in horizontal and in vertical contexts. For these
test cases, feaLib and makeotf produce the exact same output.
https://github.com/fonttools/fonttools/issues/633
If a zero value appears in a SinglePos statement, feaLib continues to
produce ValueRecords of format 0. But if a zero value appears in a
PairPos statement inside a horizontal compilation context, feaLib now
produces ValueRecords of format 4. Likewise, if a zero value appears
in a PairPos statement inside a vertical compilation context, feaLib
now produces ValueRecords of format 8.
The OpenType Feature Syntax specification is completely silent about this,
but the new behavior matches that of makeotf.
https://github.com/fonttools/fonttools/issues/633
In context of SinglePos, makeotf compiles zero values to value records
of format 0. feaLib already does the same, but it's good to have tests
to make sure that this doesn't accidentally change.
https://github.com/fonttools/fonttools/issues/633
We now correctly handle nameid statements with surrogate pairs and
old-style macOS-encoded names (provided that fonttools supports the
specified encoding).
Resolves https://github.com/fonttools/fonttools/issues/842.
Also fixes a bug where glyph alternates in MTI feature files were
wrongly sorted by glyph name. After this change, the output is using
the same ordering as in the input MTI feature file.
Fixes https://github.com/fonttools/fonttools/issues/833.
The following test cases are expected to fail because the current implementation of drop_hints
also strips the charstrings' advance widths:
Tests/subset/subset_test.py::SubsetTest::test_no_hinting_CFF
Tests/subset/subset_test.py::SubsetTest::test_no_hinting_desubroutinize_CFF
