StringPool:entry* are not trivially moveable, as they contain Vector<>
objects. The data in Vector<> is trivially moveable, but Vector<>
itself is a dynamic type with a vtable, which is not trivially
moveable.
Impact on aapt for frameworks-res is negligible, ~3%.
Bug: 31595853
Test: m -j framework-res
Change-Id: I562a46ef8934aed7aab5c5377db53d9cf64a08b9
Permissions such as android.permission.ACCESS_FINE_LOCATION
and android.permission.ACCESS_COARSE_LOCATION assumed the underlying feature
technology and aapt dump badging would emit an implied feature.
This is no longer accurate, so the only implied feature is on the general
android.hardware.location when the target SDK of the app is 21+.
Bug:25218770
Change-Id: Ica9ebf7d4f360b137af54a8c8fd6c0358cd9be6a
- Prevent third party apps from inadvertently changing internal SystemUI
flags through a call to setSystemUiVisibility(). These flags are only
set in the individual SystemUI components and can be updated in WMS
directly.
Bug: 29875297
Change-Id: I5ea238c8fb16a0eccd6e993d95a912acb359cee6
* changes:
WindowManager: Include sigil in RemoteSurfaceTrace format.
Pass SurfaceTrace through WM command in binary format.
Log transaction state to RemoteSurfaceTrace.
WindowManager RemoteSurfaceTrace infrastructure
Add "wm surface-trace" command which enables tracing of surface
commands to be switched on at runtime. Primarily intended for use
by WM CTS tests. First target in CTS will be to use show/hide
events to eliminate polling in WM tests and increase speed. Next up
looking at things like verifying various transitions and relaunch
scenarios are flicker free. Later we may want to look at a smarter
or more structured format...but it's really not much hassle to parse
the commands off a pipe so I wanted to get us started.
Test: cts-tradefed run singleCommand cts -o --module CtsWindowManagerHostTestCases --test android.server.cts.SurfaceViewMovementTests#testSurfaceMovesWithParent
Change-Id: I1ff912c405a6cb9996ee9b6e2c465d57706191ba
ZD and ZD_TYPE are set by utils/Compat.h, remove them from
StringPool.cpp and Resource.cpp.
Test: m -j native
Bug: 31492149
Change-Id: I6154e4598dd1f279e348ef6c0b32d756464bda8a
If the language tag "und-Zsym" is applied to a TextView, the default
presentation should be text-style. However, it didn't work before this
patch, due to a lack of "und-Zsym" language tag in fonts.xml.
Bug: 31297127
Change-Id: Ieed8fe1532806dd1e75c58aa0d7a724bfeed7796
Styled strings use spans to denote which part
is styled (<b>, <i>, etc). Spans are simply a range
of indices into the original string.
In Java, we use String and its internal representation, meaning
we must encode the indices using UTF16 lengths.
When the internal AAPT2 representation of strings switched to UTF8,
the indices also began to index into the UTF8 string.
This change reverts the indices to use UTF16 lengths.
Bug:31170115
Change-Id: I07b8b5b67d2542c7e0a855b601cdbd3ac4ebffb0
Update version to 2.2 and update readme to reflect added
support for inline complex XML resources.
Change-Id: If825c96541952e95860d34390c6c36704a24129b
A compiled ResourceTable may contain multiple packages.
When referencing a framework attribute, for instance,
the compile phase has no idea if that attribute is a USE
or a DECLARATION, like so:
<declare-styleable name="Foo">
<item name="android:id" />
</declare-styleable>
If this file is compiled with the package 'android', then this
is a declaration. If it is compiled with the package 'com.foo',
this is a reference that must resolve at link time.
Since this is common, warning that the package 'android' is being
ignored during the merge phase is incorrect behaviour.
The warning was antroduced in ag/1363992
Change-Id: I09254d79ac379439face177eb724df882d1d0ce1
Styleables should only be merged when processing overlays.
This moves the styleable merging code out of ResourceTable
and into TableMerger.
Change-Id: I3aae05cf4dd875cd25ac2ac744b61194409b2fee
Restrict saved surface to launcher start (ACTION_MAIN&CATEGORY_
LAUNCHER), or there is no intent at all (eg. task being brought to
front). If the intent is something else, likely the app is going
to show some specific page or view, instead of what's left last time.
This solves problems like the launcher shortcuts on DeckClock,
each of them is a different intent and will show one specific
view regardless of last states. Another example is Chrome tab
opened directly by action VIEW to open some URL.
(Note that this doesn't solve the problem with Chrome homescreen
shortcuts, it will still start with saved surface (if Chrome
is already open). This is because the shortcut is a trampoline
activity that starts the real chrome tab activity, but when
the trampoline is started, the whole task is already brought
to front, and ChromeTab could become visible with the task
before we actually start it.)
bug: 31055479
bug: 27747315
Change-Id: Id3e61c61ef516b0edc1f174320f02661222f226b
(cherry picked from commit ad24f96def)
libziparchive-host is no longer necessary, and mixing libziparchive-host
and libziparchive can cause ODR violations.
Change-Id: I623c0c5579e28f26a6e1db7bd953f33ed1ca68cd
Restrict saved surface to launcher start (ACTION_MAIN&CATEGORY_
LAUNCHER), or there is no intent at all (eg. task being brought to
front). If the intent is something else, likely the app is going
to show some specific page or view, instead of what's left last time.
This solves problems like the launcher shortcuts on DeckClock,
each of them is a different intent and will show one specific
view regardless of last states. Another example is Chrome tab
opened directly by action VIEW to open some URL.
(Note that this doesn't solve the problem with Chrome homescreen
shortcuts, it will still start with saved surface (if Chrome
is already open). This is because the shortcut is a trampoline
activity that starts the real chrome tab activity, but when
the trampoline is started, the whole task is already brought
to front, and ChromeTab could become visible with the task
before we actually start it.)
bug:27747315
Change-Id: Id3e61c61ef516b0edc1f174320f02661222f226b
These are not yet ready for updating in the unicode data, so we patch
them in code until they are.
Bug: 30594617
Change-Id: Ib4c8568aa4840341474eb0c0c95674d640473f4d
(cherry picked from commit 2b8b819fee)
