If SpanSet.init() is called several times in a row with different
values, it is possible to change "numberOfSpans" in a way that
will prevent SpanSet.recycle() from nulling out all the spans.
This can lead to memory leaks of large objects through spans
references. User @piwai reported this leak:
com.squareup.marketfont.MarketSpan
`-[1] of array android.text.style.CharacterStyle[]
`-spans of object android.text.SpanSet
`-mCharacterStyleSpanSet of object android.text.TextLine
`-[1] of array android.text.TextLine[]
`-sCached of class android.text.TextLine
The MarketSpan instance is kept alive through a recycled TextLine
which itself contains a SpanSet.
Change-Id: Idfb2233ca16895dbe735c312662eaf0b4a2ecd65
The comparator's equal implementation doesn't satisfy the constraints
of an equals method, namely being reflexive. Use the standard Object
implementation instead.
Bug: 19797138
Change-Id: I74f888e99533e1945aab7ab10fe8ee3ded6388f4
Rewrite the DateUtils' relative time formatting APIs
(getRelativeTimeSpanString, getRelativeDateTimeString) to use ICU ones.
Two APIs that take withPreposition parameter are not changed. Because
(a) ICU doesn't provide functionality to format preposition; (b) They
are not really computing relative time but instead calling
formatDateRange() to get the absolute time/date string.
Bug: 19146457
Bug: 5252772
Change-Id: Iea8d699d63cc4438513910da66d038912e44fb8d
This works around a bug in standalone (e.g. non-Zygote)
runtimes when a device is attached to a host that is running
DDM.
There is a race condition:
When the runtime receives a HELLO from DDM it calls
TextUtils.isEmpty().
Calling any TextUtils methods statically initializes
Layout. Layout has dependencies on other classes, which in
turn have dependencies on native methods that are not always
registered when the call takes place. Registration and DDM
handling are done in separate threads.
This is not a fix, merely a workaround until the race can
be resolved.
Bug: 18081539
(cherry-picked from commit d29bdb266d)
Change-Id: Id0d8578eab9e59d479a7c1b2e7ea1890ac0c8de6
This works around a bug in standalone (e.g. non-Zygote)
runtimes when a device is attached to a host that is running
DDM.
There is a race condition:
When the runtime receives a HELLO from DDM it calls
TextUtils.isEmpty().
Calling any TextUtils methods statically initializes
Layout. Layout has dependencies on other classes, which in
turn have dependencies on native methods that are not always
registered when the call takes place. Registration and DDM
handling are done in separate threads.
This is not a fix, merely a workaround until the race can
be resolved.
Bug: 18081539
Change-Id: If1bd3de6597bc93da381c8f86dacf40156449561
All supported locales use only U+2025 and U+2026 to represent
ellipses, and it will unlikely change in future. Given translated
resources are inconsistent and often use three dots it is safer
to use constants instead of resources.
(cherry-pick of ed0daa93e48d38e54a7ad1c99c461510a4c07599.)
Bug: 18542179
Change-Id: I51a6cb903f62f739fbadd6b78e5765c0028d641a
TextLine was not fully cleared before recycling it which leads to
activity leaks if the activity happens: to register a text watcher.
bug:19045507
Change-Id: Ife0f7ce29865bd30ca2dfe8795023f51f275d659
This reverts commit 9dfe86d410.
That change moved the lenience from POSSIBLE to VALID, which eliminated
false positive links, especially 4 digit phone numbers, but caused
significant false negatives, leading to CTS test failures
(android.text.util.cts.LinkifyTest#testAddLinks7 in particular).
The true fix requires new functionality to validate phone numbers in
a mobile context. In the meantime, the best solution is to revert.
Bug: 18708556
A generated quick link to telephone number gets clicked and
crashes if corresponding activity doesn't exist in device.
It attempts to open up an activity to view content which will
in turn generate an uncaught ActivityNotFoundException.
Solved by catching exception when launching activity for the
specified content.
Change-Id: I47364519f1eceb5b978b29382107deae1891c7da
The linkify logic used POSSIBLE as its leniency setting, which resulted
in false positives such as 4-digit years being interpreted as phone
numbers. Changing to VALID as per recommendation of libphonenumber
people, which fixes this problem.
Bug: 18489494
Change-Id: I77d330285de46de2fdda22daed41392106ec6ddd
In the Truncate.MIDDLE case, when the line is less than half the layout
width, the computeEllipsis logic could go past the left edge of the
string. This patch fixes the off-by-one and avoids the resulting index
out of bounds crash, and also changes the behavior so that when
ellipsizing at the middle, the string to the end of the paragraph is
taken into account.
Bug: 18508627
Change-Id: I24be09c23a5aa158791a9717419307613b8a22e8
Also, move over DateFormat.getInstance(SHORT) instead of
LocaleData.shortDateFormat4 (which forces 4 digit years).
Based on an analysis of git history, there seems to be no
real motiviation for using 4 digit years.
If anyone in the platform complains loudly, we could look at
using a skeleton pattern instead of just replacing 'y' with 'yy'.
If third party apps are affected, we could add a targetSdkversion
based check and revert to the old behaviour for older targetSdks.
bug: 18388178
bug: 18322220
Change-Id: Ieab2b41691958a2e668d382c5a6b3d9aaf741ed2
The 24 hour setting was not respected correctly. Also
fixed a bug where the next alarm would not display itself
in the QS panel.
Bug: 16239208
Change-Id: I89734f783912dead5831db49db53fba04dbf54ee
The "moreChars" test in StaticLayout's generate method would evaluate to
false when the last character in a word caused the break. This in turn
suppressed the ellipsis in this case. The proposed fix is always to set
moreChars true in the code path where the line is broken because more
text wouldn't fit.
Bug: 17738112
Change-Id: Ifa1a69841ca952da4d1937dc8326778179b026b3
libcore now offers a wider variety of 12-/24-hour time formats,
so be more specific about which one we want here.
(cherry-pick of 85f60d3a03b5b5d9a0e8b8a138eb85a6b53a1eca.)
Bug: 10361358
Change-Id: I846ab7a6f84cd49e876ad21e9366aff1600e0530
The motivation is an API change: FloatMath is going to be
deprecated and/or removed. Performance is not the goal of
this change.
That said...
Math is faster than FloatMath with AOT compilation.
While making the change, occurances of:
{Float}Math.sqrt(x * x + y * y) and
{Float}Math.sqrt({Float}Math.pow(x, 2) + {Float}Math.pow(y, 2))
have been replaced with:
{(float)} Math.hypot(x, y)
Right now there is no runtime intrinsic for hypot so is not faster
in all cases for AOT compilation:
Math.sqrt(x * x + y * y) is faster than Math.hypot(x, y) with
AOT, but all other combinations of FloatMath, use of pow() etc.
are slower than hypot().
hypot() has the advantage of being self documenting and
could be optimized in future. None of the behavior differences
around NaN and rounding appear to be important for the cases
looked at: they all assume results and arguments are in range
and usually the results are cast to float.
Different implementations measured on hammerhead / L:
AOT compiled:
[FloatMath.hypot(x, y)]
benchmark=Hypot_FloatMathHypot} 633.85 ns; σ=0.32 ns @ 3 trials
[FloatMath.sqrt(x*x + y*y)]
benchmark=Hypot_FloatMathSqrtMult} 684.17 ns; σ=4.83 ns @ 3 trials
[FloatMath.sqrt(FloatMath.pow(x, 2) + FloatMath.pow(y, 2))]
benchmark=Hypot_FloatMathSqrtPow} 1270.65 ns; σ=12.20 ns @ 6 trials
[(float) Math.hypot(x, y)]
benchmark=Hypot_MathHypot} 96.80 ns; σ=0.05 ns @ 3 trials
[(float) Math.sqrt(x*x + y*y)]
benchmark=Hypot_MathSqrtMult} 23.97 ns; σ=0.01 ns @ 3 trials
[(float) Math.sqrt(Math.pow(x, 2) + Math.pow(y, 2))]
benchmark=Hypot_MathSqrtPow} 156.19 ns; σ=0.12 ns @ 3 trials
Interpreter:
benchmark=Hypot_FloatMathHypot} 1180.54 ns; σ=5.13 ns @ 3 trials
benchmark=Hypot_FloatMathSqrtMult} 1121.05 ns; σ=3.80 ns @ 3 trials
benchmark=Hypot_FloatMathSqrtPow} 3327.14 ns; σ=7.33 ns @ 3 trials
benchmark=Hypot_MathHypot} 856.57 ns; σ=1.41 ns @ 3 trials
benchmark=Hypot_MathSqrtMult} 1028.92 ns; σ=9.11 ns @ 3 trials
benchmark=Hypot_MathSqrtPow} 2539.47 ns; σ=24.44 ns @ 3 trials
Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: I06c91f682095e627cb547d60d936ef87941be692
All supported locales use only U+2025 and U+2026 to represent
ellipses, and it will unlikely change in future. Given translated
resources are inconsistent and often use three dots it is safer
to use constants instead of resources.
Change-Id: I51a6cb903f62f739fbadd6b78e5765c0028d641a
