Merge commit '9ff71d38a060476ba19deacf12ddaeef3cbac00d'
* commit '9ff71d38a060476ba19deacf12ddaeef3cbac00d':
whoops, still need to dismiss when launching an intent for in-app search
When a restore agent throws an exception back into the framework, assume that it
may have left the app's data in an inconsistent state and clear the data for
that app (again -- it was cleared just prior to the attempted restore, too).
Now that the search manager service handles hiding (not dismissing) and reshowing it
when the user hits back after launching a result, search manager can't cache
"mShowing". Also noticed a few other minor problems that was hosing the handling
of pause / resume to reshow the dialog, like moving some logic to onHide instead
of onDismiss.
When we parse the metadata we check that there is enough data in the parcel by checking
the byte availables and the size in the header.
Since the size is in the header has been read, we should make sure than dataAvailable() >= size - 4
This bug was hidden by some test code which has been removed.
Instead of calling remove(), which would hide the soft
keyboard, remove the WebTextView manually, so the soft
keyboard remains up (if it was up to begin with). WebView
will properly handle the keys now that the focus has
changed.
* changes:
Fix 1986313. WebKit assumes things happened in a certain order. In this case, we can't call LoadListener's nativeError until BrowserFrame's startLoadingResource returns. Otherwise, the failed request will not be removed from the DocLoader's request list. So the page will never finish loading.
Before, if dropDownAlwaysVisible was false,
the soft keyboard height was always taken into
account when calculating the available vertical space for
the AutoCompleteTextView drop-down height.
This meant that the drop-down would never pop over the soft
keyboard when navigating in the drop-down list.
This change makes the drop-down pop over the soft keyboard
when the keyboard is not needed, regardless of the value of
dropDownAlwaysVisible.
Fixes http://b/issue?id=1978935
That way we don't have to wait for clearsilver to write the thousands
of html files during every build.
(this is the frameworks/base change, there is also a build change)
A subclass of SurfaceView is created in order to appear like a normal
SurfaceView to jni but can be managed in Java easily. This class implements the
SurfaceHolder.Callback interface in order to notify the native PluginSurface
when the surface changes.
This does not change the actual thread priority but if the default priority
changes, this will have the desired result of being slightly less favorable.
I tried messing with the priorities to see if a higher or lower priority helped
but it seemed to do more harm than good. It might be worth more investigation to
find out where all the thread time is being spent so we can determine the
correct priority.
Merge commit '8b9baf2dfb4dd1bd6379d923a3aa61a082783368'
* commit '8b9baf2dfb4dd1bd6379d923a3aa61a082783368':
Remove exception throw when a view has a self dependency inside a
Merge commit '841f13c8e9ff3f7695b6c18a8abcec3c947983ff'
* commit '841f13c8e9ff3f7695b6c18a8abcec3c947983ff':
* Reverted the change in PackageParser that I checked by accident
Merge commit '5c536e9162721c460699a041959a0d67de1d20db'
* commit '5c536e9162721c460699a041959a0d67de1d20db':
Fix issue where scaled bitmap sizes could be wrong.
Merge commit '4b897aa99bf71f933f5b202c1531858382122d6f'
* commit '4b897aa99bf71f933f5b202c1531858382122d6f':
Making sure that there are no calls made to the native TTS
Merge commit '22eef5da75604cd14c0f87332cf957c9948a6e33'
* commit '22eef5da75604cd14c0f87332cf957c9948a6e33':
Add a delay before collecting memory information
RelativeLayout.
There are unfortunately successful 3rd party apps declaring self dependencies.
There's no reason to crash the app so just ignore this issue.
* More surface view fix.
- correct event translation on surface view.
- use compatible window
* removed FLAG_NO_COMPATIBILITY_SCALE. It was my misunderstanding of how SurfaceView works, and this was not necessary.
* Added compatibility related info to package dumpsys
The Bitmap functions to get the scaled width/height couldn't actually
do the right thing because they didn't know the destination they would
be drawing to. Now there are two forms of them, taking an explicit
parameter specifying the destination.
