History log of /frameworks/minikin/tests/unittest/MeasurementTests.cpp
Revision Date Author Comments (<<< Hide modified files) (Show modified files >>>)
1d461589869ee5b7102f96271b0ef0a776ab513c 20-Oct-2017 Seigo Nonaka <nona@google.com> Style cleanup.

This CL contains following cleanups:
- Rename the following classes and corresponding file names:
FontLanguage -> Locale
FontLanguages -> LocaleList
FontLanguageListCache -> LocaleListCache
- Remove unnecessary #include lines.
- Define the Minikin's header include ordering and follow that style.
See doc/minikin_style.md for details.
- Always use std::string and std::vector instead of string and vector
with using-declaration.
- Replace LOG_ALWAYS_FATAL_IF with MINIKIN_ASSERT so that the assertion
is removed on build configuration for user.

This CL does not change any behavior.

Bug: 65024629
Test: make checkbuild
Change-Id: If41bb3c3389303053bbe95e349eb2dd4b619257c
/frameworks/minikin/tests/unittest/MeasurementTests.cpp
b01028d1d7bc3906ef71c72ad985919f79304b5e 14-Mar-2017 Roozbeh Pournader <roozbeh@google.com> Break grapheme clusters after viramas if they end a cluster

Previously, we stayed on the conservative side and disallowed any
grapheme breaks (and thus cursoring) where a virama was followed by a
letter, since we did not know if the virama would be forming a
cluster with the letter or not. This created problems with Indic
languages with infrequent conjuncts, such as Tamil.

Now we use the information in calculated advances to find if a
cluster is formed. If there is no cluster, we break the grapheme and
allow cursoring after the virama.

Test: Unit tests added to GraphemeBreakTests and MeasurementTests.
Test: Also manually tested Tamil sequences.
Bug: 35721792
Change-Id: Ib159edb94b3ad6f693f0d3dad016b332b2cef447
/frameworks/minikin/tests/unittest/MeasurementTests.cpp