String: reimplement replace(CharSequence, CharSequence).
We're not using regexes any longer. Performance is now on par
with Android M.
This change also adds basic unit-tests, which we were lacking so far.
Android M
---------
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=EMPTY}}
runtime(ns): min=33.49, 1st qu.=33.85, median=34.10, mean=34.22, 3rd qu.=34.55, max=35.09
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_16}}
runtime(ns): min=48.58, 1st qu.=48.98, median=49.20, mean=49.51, 3rd qu.=49.83, max=51.75
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_64}}
runtime(ns): min=236.04, 1st qu.=236.82, median=238.75, mean=239.42, 3rd qu.=241.72, max=243.95
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_256}}
runtime(ns): min=921.01, 1st qu.=924.97, median=932.92, mean=934.84, 3rd qu.=940.41, max=964.41
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=EMPTY}}
runtime(ns): min=33.87, 1st qu.=34.10, median=34.26, mean=34.33, 3rd qu.=34.51, max=35.00
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_16}}
runtime(ns): min=695.64, 1st qu.=747.74, median=780.07, mean=792.92, 3rd qu.=860.18, max=860.55
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_64}}
runtime(ns): min=2719.39, 1st qu.=2758.88, median=2778.37, mean=2901.49, 3rd qu.=2858.23, max=3782.71
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_256}}
runtime(ns): min=12814.17, 1st qu.=12831.58, median=12938.24, mean=12973.31, 3rd qu.=13051.15, max=13359.51
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=EMPTY}}
runtime(ns): min=33.50, 1st qu.=33.65, median=34.00, mean=34.20, 3rd qu.=34.74, max=35.20
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_16}}
runtime(ns): min=48.51, 1st qu.=49.01, median=49.37, mean=49.41, 3rd qu.=49.86, max=50.25
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_64}}
runtime(ns): min=2536.89, 1st qu.=2687.66, median=2726.48, mean=2726.00, 3rd qu.=2774.11, max=2898.73
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_256}}
runtime(ns): min=9524.14, 1st qu.=9673.23, median=10163.81, mean=10062.96, 3rd qu.=10375.81, max=10594.24
Android N
---------
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=EMPTY}}
runtime(ns): min=32.09, 1st qu.=32.33, median=32.53, mean=32.63, 3rd qu.=33.01, max=33.22
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_16}}
runtime(ns): min=48.92, 1st qu.=49.12, median=50.11, mean=50.52, 3rd qu.=52.44, max=52.98
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_64}}
runtime(ns): min=251.74, 1st qu.=253.67, median=255.06, mean=255.81, 3rd qu.=257.76, max=262.20
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceNonExistent, vm=default, parameters={s=L_256}}
runtime(ns): min=915.08, 1st qu.=923.41, median=926.69, mean=926.91, 3rd qu.=932.22, max=935.61
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=EMPTY}}
runtime(ns): min=32.09, 1st qu.=32.53, median=32.64, mean=32.73, 3rd qu.=33.03, max=33.16
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_16}}
runtime(ns): min=49.25, 1st qu.=51.35, median=51.91, mean=51.78, 3rd qu.=52.60, max=53.11
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_64}}
runtime(ns): min=2422.00, 1st qu.=2457.65, median=2536.93, mean=2534.74, 3rd qu.=2600.31, max=2662.56
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSequenceRepeated, vm=default, parameters={s=L_256}}
runtime(ns): min=9604.40, 1st qu.=9690.33, median=9935.53, mean=10176.73, 3rd qu.=10554.54, max=11774.79
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=EMPTY}}
runtime(ns): min=31.97, 1st qu.=32.20, median=32.54, mean=32.58, 3rd qu.=32.90, max=33.40
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_16}}
runtime(ns): min=657.99, 1st qu.=673.30, median=707.95, mean=717.83, 3rd qu.=752.64, max=794.76
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_64}}
runtime(ns): min=2553.89, 1st qu.=2583.49, median=2675.01, mean=2698.93, 3rd qu.=2724.23, max=3105.93
Experiment {instrument=runtime, benchmarkMethod=timeReplaceSingleSequence, vm=default, parameters={s=L_256}}
runtime(ns): min=9669.46, 1st qu.=9958.64, median=10113.85, mean=10363.71, 3rd qu.=10331.31, max=12721.61
bug: 28060800
Change-Id: I98b73231af3df575840754fad68cbc5542f1b9ab
2 files changed
