The current parameter name "index" in
CursorAnchorInfo.Builder#setComposingText is a bit hard
to understand.
This CL names it with "composingTextStart" to clarify its usage.
Change-Id: I504939b8d2b48875d8baa9630b5a4f04a5d66dbe
Replace locale.substring(0, 2) with a function that always
returns the first component of the locale (assumed to be the
language).
bug: 10090157
Change-Id: Ibb9eb9f65323795d139b16d76b7e7e36a4e0568c
In previous change we use setForceUse(FORCE_MEDIA, XXX) method
in order to set specific audio output type, such as LINE, HDMI_ARC,
and SPDIF. But it turns out that it conflicts with bluetooth
a2dp which uses the same setForceUse(FORCE_MEDIA, yyy).
This change is based on several conditions.
1. When other non-speaker devices are on, prevent system audio
from turing on.
2. In order to keep track of other devices' connectivity and to
turn off system audio if other device like bluetooth or
headphone preempts current output, register OnAudioPortChangeListner
to audio manager.
3. All possible system audio outputs can be merged with other
outputs without priority.
Change-Id: Id4e47d99db64b9f77a17c2c28c47787ab8980bf7
The BCP47 tag we use for selecting Simplified vs Traditional Chinese is
Hans and Hant. This patch uses those tags in fallback_fonts.xml rather
than country names, which weren't getting matched.
Change-Id: I97e1da7b86e8fe8f026365146393304a5ab74443
A VPN declared bypassable allows apps to use the new multinetwork APIs to
send/receive traffic directly over the underlying network, whereas without it,
traffic from those apps would be forced to go via the VPN.
Apps still need the right permissions to access the underlying network. For
example, if the underlying network is "untrusted", only apps with
CHANGE_NETWORK_STATE (or such permission) can actually use it directly.
New API with stub implementation to be filled out later.
Bug: 15347374
Change-Id: I8794715e024e08380a43f7a090613c5897611c5b
Instead of producing Ranking objects, RankingMap just populates
them now, allowing developers to re-use objects and avoid
unnecessary allocations.
Also rename isInterceptedByDnd() to meetsInterruptionFilter(),
since DND is not a concept anymore.
Bug: 15415840
Bug: 16099064
Change-Id: If9861cbdf14593e641a4d4ffd1b967647eb8e2b8
Will resubmit tomorrow for Build Breakage Friday as this change needs new SDK prebuilts for unbundled app branches.
This reverts commit aa746b27610680a2c0fbdf7d81a6455c4597f05e.
Change-Id: I28ba7e1b09234a5eb1b725aed043b9de98e1bc16
HDMI CEC/MHL related settings. All the settings are hidden, to be used
by the service and settings UI.
Change-Id: I1113aa5f77189db73acfa5fe36670a3cd621b1c1
So we put path and group into a big list, then use reflectiont to handle them
differently.
bug:16162141
Change-Id: I4c84dc952d5d9b49412301dcd6459395e21d3b31
Added another parameter deviceType for HdmiControlService to
run the specified local device for sendKeyEvent.
Bug: 15844076
Change-Id: I598ef320ae94bba0ace38701ae0ca12fd0625559
Over time, we've unpacked native libraries at various places with
respect to their source APK. Persisting this path in PackageSettings
has caused more pain recently with the switch to supporting multiArch
and cluster installs.
This change switches us to always derive the native library paths at
runtime based on the type of install. This also ensures that
transitioning between a bundled system app and an upgraded system
app will always build the right path.
We still persist the last generated path into PackageSettings to make
cleanup at uninstall time easier.
Bug: 16208505, 16206748, 16212206
Change-Id: Ieb82a424ca4a92b5674983453c50ba4b695abfb0
Separate definition of public keys and keysets in the manifest to better
represent their relationship. The 'key-set' tags should have nested additional
'public-key' tags that indicate which of the defined 'public-key' tags are
associated with them. The first use of a given 'public-key' name should define
its value; subsequent uses may refer to it only by name. 'key-set' and
'public-key' names may not intersect.
Also, change 'keys' tag to 'key-sets' to avoid issues with previous keysets
implementation.
Bug: 6967056
Change-Id: I7534e4a42326e97b67b55509187c0d3c21a2bb32
Historically, we were allowing an app that prints to specify that
the printed document has zero pages. While this does not make any
sense we should keep the behavior as people may have apps that do
that. This change fixes this issue and now we treat zero the same
way as undefined page count and ask the app to write all pages to
check the written PDF for the page count.
bug:16199127
Change-Id: I4e7de66b669e9f783db0252244a6c1e5b24ffe28
