At any position, if state is not zero, mark unsafe-to-break before,
unless we can reason it safe.

At any position, if there's an action entry for end-of-text, mark
unsafe to break.

Also changes buffer diff impl to allow for flag differences as long
as the buffer glyph flags are superset of reference glyph flags.

With this, all MORX tests pass.

This only counts recursions right now.  Good start.

Hopefully...
Fixes https://github.com/behdad/harfbuzz/issues/429

If buffers are empty, content type should be ignored.

This fixes last of the failing tests: fuzzed.tests.  Green again!

I like to have a mode where CONTAINS_NOTDEF and CONTAINS_DOTTEDCIRCLE are not
returned.  Abused a value of -1 for that.  hb-shape now uses it.  Fixes two
of the six tests failing with --verify in test/shaping/run-tests.sh.

New API.

Change hb_buffer_diff to explicitly cast result of abs to unsigned when
comparing with position_fuzz to avoid unsafe signed/unsigned comparions
warnings on windows.

Based on patch from Jonathan Kew.

Needs more cleaning up and documentation.

New API:
hb_buffer_diff_flags_t
hb_buffer_diff()

We break and shape fragments and reconstruct shape result from them.
Remains to compare to original buffer.  Going to add some buffer
comparison API and use here, instead of open-coding.

Ouch!

Makes consumption easier.

We were relying on cluster merges not requiring unsafe flagging because
they get merged.  If cluster level requests no merging, then we flag
unsafe when merge would have happened.

This shouldn't matter.

Always use the algorithm for non-monotone case.  It's more robust.

Not all shapers code is updated to set this properly.
GSUB and Arabic shaper are updated.
GPOS and other shapers are NOT.

Fixes https://github.com/behdad/harfbuzz/issues/224

Fixes https://bugs.chromium.org/p/chromium/issues/detail?id=614647
If allocation fails, we might be leaving junk behind.  At least
clear it up.

Part of https://github.com/behdad/harfbuzz/issues/277

This makes defining HB_NDEBUG much less relevant, to the
point of irrelevance.  Sorry about all the fuss in previous
release!

API changes:
- If NDEBUG is defined, define HB_NDEBUG
- Disable costlier sanity checks if HB_NDEBUG is defined.

In 1.2.3 introduced some code to disable costly sanity checks if
NDEBUG is defined.  NDEBUG, however, disables all assert()s as
well.  With HB_NDEBUG, one can disable costlier checks but keep
assert()s.

I'll probably add a way to define HB_NDEBUG automatically in
release tarballs.  But for now, production systems that do NOT
define NDEBUG, are encouraged to define HB_NDEBUG for our build.

Saves some sweet time and binary size!

Fixes https://github.com/behdad/harfbuzz/issues/223

Right now we cannot test this because it has to be tested using hb-fuzzer.
We should move all fuzzing tests from test/shaping/tests/fuzzed.tests to
test/fuzzing/ and have its own test runner.  At that point, should add
test from this issue as well.

Some of it (create, reference, destroy) are adapted from Cairo docs.

Currently just announces lookup applications.  Message-API *will* change.
hb-shape / hb-view are updated to print-out messages to stder if --debug
is specified.

This reverts commit 68b507a3.

Commit abadc171 provides a better
fix for this.

https://github.com/behdad/harfbuzz/issues/161

See discussion here:
https://github.com/behdad/harfbuzz/commit/81ef4f407d9c7bd98cf62cef951dc538b13442eb

There's no way to disable this fallback, but I don't think it would
be needed.  Let's hope for the best!

Fixes https://github.com/behdad/harfbuzz/issues/153