| b07812780bdb1aee50fae61cfa5f42be10dc8cfb |
|
17-Feb-2017 |
Vladimir Marko <vmarko@google.com> |
Do not use String.setCharAt() in CaseMapper.
The internal API String.setCharAt() breaks the assumption
that strings are really immutable. That in turn breaks
string compression invariants - compressible strings must
be compressed. This is a first step of removing this API.
We now do an extra allocation for those paths where we
previously used a modifiable String but on the other hand
we avoid repeated JNI calls when mapping longer strings.
Relevant cases from StringCaseMappingBenchmark:
(Lower is better.)
timeToLowerCase_US/UPPER2: 443.46 -> 376.47 (-15%)
timeToLowerCase_US/MIXED2: 441.58 -> 358.37 (-19%)
timeToLowerCase_US/UPPER8: 1062.30 -> 548.56 (-48%)
timeToLowerCase_US/MIXED8: 896.00 -> 550.03 (-39%)
timeToLowerCase_US/UPPER32: 5175.08 -> 1502.36 (-71%)
timeToLowerCase_US/MIXED32: 2743.20 -> 1387.55 (-49%)
timeToLowerCase_US/UPPER512: 68071.45 -> 18709.82 (-73%)
timeToLowerCase_US/MIXED512: 46786.96 -> 18748.79 (-59%)
timeToUpperCase_US/LOWER2: 602.21 -> 706.88 (+17%)
timeToUpperCase_US/MIXED2: 511.35 -> 773.32 (+51%)
timeToUpperCase_US/LOWER8: 1194.41 -> 989.79 (-20%)
timeToUpperCase_US/MIXED8: 1113.68 -> 1033.80 (-13%)
timeToUpperCase_US/LOWER32: 4899.31 -> 2246.14 (-54%)
timeToUpperCase_US/MIXED32: 3469.12 -> 2455.02 (-30%)
timeToUpperCase_US/LOWER512: 76607.06 -> 25412.57 (-67%)
timeToUpperCase_US/MIXED512: 55735.11 -> 26115.87 (-56%)
There was a lot of noise, especially in the short string
benchmarks, for example the timeToLowerCase_US/LOWER2 which
should have been unaffected (not listed above) had +35%.
The timeToUpperCase_US/LOWER2 from this report seems to be
on the low end of the range while timeToUpperCase_US/MIXED2
hit the high end of the range. The timings for long strings
are much more consistent.
Test: testrunner.sh --host
Test: run-libcore-tests.sh --mode=host
Bug: 31040547
Change-Id: Id107428c000de1ff3db94b4824f70bf3ae5001dc
|
| 728029941ba340aeb8e4098b7ad12d3a30e38aa9 |
|
05-Oct-2015 |
Neil Fuller <nfuller@google.com> |
Switch from jarjar to generated source for ICU4J
Future changes will include ICU4J classes as public APIs
(still beneath android.icu). To generate these docs source
code will be needed under the expected package (android.icu),
as well as the ability to apply changes to the javadoc
comments programatically so ICU can be upgraded easily.
A consequence of this change is that libcore code can now
depend on icu in the android.icu package and not the
com.ibm.icu package as was the case previously. A nice
side effect is that the source code is available for use
during debugging as well.
Bug: 22023363
Change-Id: Icba5668fae53e2dc98ccc74e2f1ad1ca226e5feb
|
| 17552b6592aeaaa564725a07e10e8c4b1c8163cb |
|
14-Jul-2015 |
Rayhaan Jaufeerally <rayhaan@google.com> |
Remove Transliterator wrapper class.
This change removes the wrapper class in Java which was used for
proxying requests to ICU4c. Now ICU4J is called directly at the
call sites. See: https://android-review.googlesource.com/#/c/162002/
This is done by replacing instantiations of the old Transliterator
class with calls to a static factory in ICU4J which returns a cached
transliterator if available, otherwise creates it on demand.
This change requires that the transliterator resource directories
and source files are added to our java libraries, which increases
the size of core-libart.jar from 2.289 MB to 2.332 MB. Maybe it's
possible to wait until dependencies on ICU4c have been removed to
see if we can save space there.
Change-Id: I2b2f84a051ba9815ba6d346dfe7fb764e39032c4
|
| 83c7414449bc406b581f0cb81ae06e7bce91403c |
|
15-Jan-2014 |
Jeff Hao <jeffhao@google.com> |
Removed offset and value from String and added StringFactory.
Change-Id: I55314ceb906d0bf7e78545dcd9bc3489a5baf03f
|