Fixes #1054.

New API:
+hb_set_is_subset()

Some of the build bots, still fail on this.  I suppose mine pass because
I enable C++11...

Anyway, remove these again.

This reverts commit ff92de76.

Revert that and remove ASSERT_POD. Let's see which bots are
unhappy with this configuration...

Towards possibly using Null pool for some nil objects.

llvm-gcc-4.2 bot had this problem:
hb-private.hh:812: error: initializer specified for non-virtual method 'void hb_vector_t<Type, StaticSize>::operator=(const hb_vector_t<Type, StaticSize>&) [with Type = hb_user_data_array_t::hb_user_data_item_t, unsigned int StaticSize = 1u]'

Removing the delete didn't work with a constructor. So, remove constructor.
Just disallow assignment.  Still better than nothing.

We'll see if this sticks to the bots.

Part of https://github.com/harfbuzz/harfbuzz/issues/1017

Part of https://github.com/harfbuzz/harfbuzz/issues/1017

For consistency.

Part of https://github.com/harfbuzz/harfbuzz/issues/1017

Was totally b0rked. Ouch!

Fixes https://github.com/harfbuzz/harfbuzz/issues/863

In preparation to fix https://github.com/harfbuzz/harfbuzz/issues/863

Binary operations are sets are completely broken. Ouch.

Switch from a 128+64 hierarchy, to 32+32. New page size is 1024.
This should make set iteration faster, without affecting other op
performances much.

New API:
- hb_set_previous()
- hb_set_previous_range()

Ouch!

Fixes https://github.com/harfbuzz/harfbuzz/issues/657

Specificaly, when a range or sorted array has unexpected order, we take that as
font data being garbage and bail out.  This fixes significant slowdown on a bad
version of Chandas font which has a 600KB GPOS with garbage inside.

Later on, I like to add a maximum-work counter for collect_glyphs to protect
against malicious fonts as well.

Fixes https://bugs.chromium.org/p/chromium/issues/detail?id=794896

Does page lookup as needed.

The three "XXXXX"'s should be switched to false.  Doing that separately for ease
of bisecting...

Not optimized to use sortedness yet.  Also start putting in place infra
to faster reject bad data.

A version of Chandas.ttf found on some Chrome bots has 660kb of GPOS,
mostly junk.  That is causing 48 million of set->add() calls in
collect_glyphs(), which is insane.

In the upcoming commits, I'll be speeding that up by optimizing
add_sorted_array(), while also reducing work by rejecting out-of-sort
arrays quickly and propagate the rejection.

Part of https://bugs.chromium.org/p/chromium/issues/detail?id=794896

Now that pagesize is 8192, this feels better.

Fixes perf regression on some heavy fonts in Chrome's FT+HB
interaction.

See:
https://bugs.chromium.org/p/chromium/issues/detail?id=782220

More work to be done:
https://bugs.chromium.org/p/chromium/issues/detail?id=794896

To be used to optimize adding a whole bunch of (sorted) items at the same time,
as in CoverageFormat1.

It's as good as it gets, and seems to be on par with previous set implementation
in my benchmark.

Would be great if someone can double-check my bitops.