This CL introduces performance tests for FontCollection.

To support TTC file in /system/fonts, this CL also extends FontTestUtils

Bug:29142734
Change-Id: I9d8ad24ca55f61031b85623ab7c26234239e4f41

This CL introduces performance tests for following three modules:
- Hyphenator
- WordBreaker
- GraphemeBreak

During using UnicodeUtils, need to decouple it from gtest since
perftest doesn't have gtest dependencies.

Bug:29142734
Change-Id: I700c662fa7d0a52f19d8e93150ad1a85dc28769f

This fixes following three memory leaks in test utilities.
There is no problem in production code and this CL doesn't affect
any production behaviors.

- SkTypeface leaks due to forget calling SkSafeUnref in dtor.
- MinikinFontForTest leaks during constructing FontCollection.
- FontCollection leaks due to unnecessary AddRef.

Change-Id: I22e1e0307f1b2499296acb1aacc3ef66076a36e9

Here is a new policy of the namespace of minikin.
- All components should be in minikin namespace.
- All tests are also in minikin namespace and no anonymous namespace.

Bug: 29233740
Change-Id: I71a8a35049bb8d624f7a78797231e90fed1e2b8c

To add perftests and reuse some utility classes, reconstruct
test directory structure.

- Move unit tests from minikin/tests to minikin/tests/unittests
- Extract utilitiy classes to minikin/tests/utils which will be
  used by perftests eventually.

Change-Id: I5026b177934e72ae67d362ee888302037da2f808

Bug: 28992567
Change-Id: Id14cc86761d6c5b408262fe8684e2b293a420a4f

am: b11b991f

* commit 'b11b991f':

Change-Id: I0a3d24fcffb1d0533970058b4a3498d45e794f36

am: b4bbae32

* commit 'b4bbae32':
  Do not break after Myanmar viramas

Change-Id: I69972aed4fe1c1b956c0b65af18e1a4f1c739bf4

am: 8614b042

* commit '8614b042':

Change-Id: I4437e1206de2a45593e69956d8bcdf85a64b0f32

am: 00de2cc3

* commit '00de2cc3':
  Do not break after Myanmar viramas

Change-Id: I2e240fab84a49c73165056ee5ff18dbfca0faeec

am: 74b56175

* commit '74b56175':
  Do not break after Myanmar viramas

Change-Id: Idb9303889ac87853a730cdb25fba7faaaf352b93

am: 74b56175

* commit '74b56175':
  Do not break after Myanmar viramas

Change-Id: I13297120c3c3789c03b860c9e72b3a04922ab9fa

This is to work around a bug in ICU's line breaker, which thinks
there is a valid line break between a Myanmar kinzi and a consonant.
See http://bugs.icu-project.org/trac/ticket/12561 for the ICU bug.

Bug: 28964845
Change-Id: I076ac15077e5627cbccf6732900bcc60d8596dda

am: bf94219d

* commit 'bf94219d':

Change-Id: I022dd9f9f800851fc7941dcdc4ebd0a9aca0f0bf

am: 8010ad47

* commit '8010ad47':
  Do not break before and after ZWJ.

Change-Id: I5cdc475d6b525aac29b8433e675d2ad2feb68722

am: f9783b13

* commit 'f9783b13':

Change-Id: I5a637871133d3e45c47c5b6dc8407bda946dbd62

am: 3bee494e

* commit '3bee494e':
  Do not break before and after ZWJ.

Change-Id: If1fe6c6192aae09bab68a00bc94aba3e768dd445

am: 77f48834

* commit '77f48834':
  Do not break before and after ZWJ.

Change-Id: Ic476d9048b44b84c69fd185b76108c961d8e8545

am: 77f48834

* commit '77f48834':
  Do not break before and after ZWJ.

Change-Id: I00a0e31c19dd5b2fbe2ffb389bec4001dec5d6c1

The emoji list is generated from external/unicode/emoji-data.txt

Bug: 28248662
Change-Id: Ie49b3782505665d62c24371ca23d317ae5e9c5f7

am: 34c784e5

* commit '34c784e5':
  Returns hasVariationSelector true for VS15/VS16

Change-Id: I14b0971a3ff50572e594a0814a26e10113fe7c1c

am: 994aa84f

* commit '994aa84f':
  Returns hasVariationSelector true for VS15/VS16

Change-Id: Iff3953948c8457e2a85e5859559411511a326f33

Minikin has a special font fallback for VS15/VS16, so
hasVariationSelector for emojis with VS15/VS16 should always return
true.

Bug: 27531970
Change-Id: Ieebd58f48b135b6ec50d999df68dcc09b1284606

am: 4d42baff

* commit '4d42baff':
  Clear mLineWidths in LineBreaker::finish()

Change-Id: I59606b706f78eda015da28f0ce6201b445086cbe

am: b1203b64

* commit 'b1203b64':
  Clear mLineWidths in LineBreaker::finish()

Change-Id: I968a1584b9f2d2242725b8ab7047091ca9ecd6ee

There was the possibility of stale indents from previous invocations
persisting in the mLineWidths across multiple invocations. This patch
clears them.

Bug: 28090810
Change-Id: I3621dfbe983512046289373711709aeade52eab4

am: af36413c

* commit 'af36413c':
  Fix minikin_unittests

Change-Id: Ibac5c54ae165d96811238a4e8d2e9612d1192eaf

am: 6c60831c

* commit '6c60831c':
  Fix minikin_unittests

Change-Id: I865dd495ed60b0bbace0ba77b2071dac058887e9

This CL fixes following test cases in minikin_tests
- FontFamilyTest.hasVariationSelectorTest
- HbFontCacheTest.getHbFontLockedTest
- HbFontCacheTest.purgeCacheTest

For the fix of FontFamilyTest.hasVariationSelectorTest, removing virtual
from GetUniqueId() in MinikinFont.  After [1], MinikinFont's destructor
started calling purgeHbCache() which calls virtual method,
MinikinFont::GetUniqueId().  Fortunately, the SkTypeface::uniqueID()
returns just internal value, so we can store it at the construction time
and use it instead of calling SkTypeface::uniqueID() every time.

This patch also changes purgeHbFont to purgeHbFontLocked, as all uses of
it were already under global mutex. This change avoids deadlock on
explicit unref, as when invoked by a Java finalizer from the Java object
that holds a reference to the font.

Some of the tests needed to change to using the ref counting protocol
rather than explicitly destructing font objects, as well.

[1] 9afcc6e2

Bug: 28105730
Bug: 28105688
Change-Id: Ie5983c4869147dacabdca81af1605066cd680b3f

am: bfc161d8

* commit 'bfc161d8':
  Update minikin/sample code to use new GetTable

Change-Id: Idd6063401cdb1399bac52c975a32a0a0b221652f

am: 29abb821

* commit '29abb821':
  Update minikin/sample code to use new GetTable

Change-Id: I5fcae79c42322dcc0533dbd3eb1a51007e089170

We changed the signature of the MinikinFont::GetTable method. This
patch updates the sample code, and fixes the build.

Change-Id: I1977be868bf7636986fc802915f3dd54c418a73a

am: 0b5d3b2b

* commit '0b5d3b2b':
  Avoid copying of font table data

Change-Id: Ie1e0be38e1875702414b7ca005c5837bd60b82d8

am: aaa4e347

* commit 'aaa4e347':
  Avoid copying of font table data

Change-Id: Idc4074ebf93595fa0420986814462095730f00c2

The hb_font_t object holds on to tables of font data, acquired through
the MinikinFont::GetTable interface, which is based on copying data
into caller-owned buffers. Now that we're caching lots of hb_font_t's,
the cost of these buffers is significant.

This patch moves to a different interface, inspired by HarfBuzz's
hb_reference_table API, where the font can provide a pointer to the
actual font data (which will often be mmap'ed, so it doesn't even
consume physical RAM).

Bug: 27860101
Change-Id: Id766ab16a8d342bf7322a90e076e801271d527d4

am: aa495426

* commit 'aa495426':
  Purge hb font on Minikin font destruction

Change-Id: If40c31c90bf4f12fdd882be4f2c6f9b35ba2d09c

am: 9afcc6e2

* commit '9afcc6e2':
  Purge hb font on Minikin font destruction

Change-Id: Idb2034353407e8b390cc07862803f846d5adea0c

This patch eagerly purges the corresponding hb_font_t object from
the HbFontCache when the underlying MinikinFont is destroyed. After
that, the key will no longer be accessed, so having the entry is
wastes memory.

Bug: 27251075
Bug: 27860101
Change-Id: I1b98016133fe3baf6525ac37d970a65ddccadb4f