Bug 27577177
When ArcMotion created paths in some quadrants, the control
points were placed in the wrong place, making a non-circular
arc instead of an arc of a circle.
Change-Id: Ie842524c09385038fb1f991b877b01ee5e740f20
This is 2nd attempt of I5bb1bd8cdb9096d516d60beb9936e55bf2b757ae
The motivation of this CL is enhance the default IME subtype enabling
algorithm. The new approach is done by score based algorithm. The
design of the matching score is determined as follows:
- The matching score for the each two locale is up to 3 and determined
as follows:
- Score of 3 : matches all language, script and country.
- Score of 2 : matches the language and script.
- Score of 1 : matches the language regardless of country.
- Score of 0 : doesn't match the language regardless of script and
country.
- All locales are fully expanded before matching by addLikelySubtags in
ICU.
Bug: 27129703
Bug: 27348943
Change-Id: I8fc774154f5175abff2f16e8f12a4847bf5f5b7c
updated. Also fiddle with the UI to use more standard values.
To be sure the print service state alwasy updated I changed
PrintManager.getPrintServices to return a loader which just wraps a
registerListener/getList/removeListener combo.
I also added a new function to enabled/disable a print service to be
keep all updating logic inside the PrintManagerService->UserState.
Then I changed all code to use this new interface.
Detailed comments:
PrintServiceInfo:
- I had to add the enabled state to the PrintServiceInfo as some users
of PrintManager.getPrintServices want all services but then display
different data depending on the enabled state. Of course I could have
created two PrintManager.getPrintServices-loaders to load the two
separate list of services. I think it is much easier to add this
property though. It is updated every time new data is returned to the
PrintManager.getPrintServices-loader.
AddPrinterActivity:
- This is shown as a dialog-style overlay to indicate that the user will
return to the select-printers activity. It contains of three list that
are updated via separate loaders.
- The recommended services will be added later to keep this path set
small.
PrintActivity:
- There are two small places where we have to update the data when we
get a new list of print services.
- In very, very rare conditions it can happen that the print service
of the current printer gains or looses the "advancedOptions"
activity
- If we have no enabled print services we want to show "Add printer"
instead of "All printers...".
- Also the print registry is not the only loader anymore, hence we have
to assign loader ids to it to not conflict with the other loaders in
this activity.
- Small bug in onPrintersChanged: If a printer is selected and the print
service of this printer gets disabled the holder goes into "removed"
state which disables the printer. When the print service is then
enabled again, we forgot to re-enable the holder.
PrinterRegistry:
- The registry assumed that the FusedPrinterProvider was the only loader
in the activity. This is not true anymore, hence it has to assign the
appropriate loader ids.
- The FusedPrinterProvider has an internal loader, hence we have to
forward a loader Id into it.
- The PrintRegistry is only called backed for a single loader, hence no
need to check the loader-id.
SelectPrinterActivity:
- The AddPrinterDialog was removed as we now have the
AddPrinterActivity.
- Added a loader for the enabled services to update the empty state.
- Added dedicated loader Id for the PrinterRegistry again.
- If we have no enabled services, the SelectPrinterActivity chainloads
the AddPrinterActivity as this is the only thing the user can do
anyway. "Save a click". This should only happen when the activity is
create the first time.
- Moved the "add printer" from the menu item to the list of printers as
suggested by UX and Zach.
PrintManagerService, UserState and IPrintManagerParamtersTest:
- As the only place where the print service state is updated is now the
userstate, we have no more sychronization problems. Whohoo.
- The users can now register for changes to the print services similar
as they can register for changes to the print jobs.
- UserState.getPrintServices is the only function can exposes any
knowledge of the print services to the outside world.
Change-Id: I9be2c7300431e06aaff9bdf7eb36120d869b56ac
Currently PopupDecorView temporarily gets FLAG_NOT_FOCUSABLE bit during
exit transition. When doing this we it is important that PopupDecorView
does not have FLAG_ALT_FOCUSABLE_IM bit, because having both
FLAG_NOT_FOCUSABLE and FLAG_ALT_FOCUSABLE_IM means that PopupDecorView
is IME-focusable.
To prevent IMEs from flickering, this CL ensures that PopupDecorView is
not IME-focusable during exit transition.
Bug: 27396330
Change-Id: I0cc81791b826f7d4257f0b6085c028c796fd9536
This reverts commit 9e7a1c9824.
Seems that that CL causes ArrayIndexOutOfBoundsException when
initializing InputMethodManagerService, which results in an infinite
boot animation.
Bug: 27129703
Bug: 27348943
Change-Id: I474a87876670ac018c675ac7b4608e90fbb2434b
The motivation of this CL is enhance the default IME subtype enabling
algorithm. The new approach is done by score based algorithm. The
design of the matching score is determined as follows:
- The matching score for the each two locale is up to 3 and determined
as follows:
- Score of 3 : matches all language, script and country.
- Score of 2 : matches the language and script.
- Score of 1 : matches the language regardless of country.
- Score of 0 : doesn't match the language regardless of script and
country.
- All locales are fully expanded before matching by addLikelySubtags in
ICU.
Bug: 27129703
Change-Id: I5bb1bd8cdb9096d516d60beb9936e55bf2b757ae
If the list of providers for the application being bound is empty,
don't bother trying to register them.
Bug: 27510621
Change-Id: I921fb70a4c97433fa123fa88e170745ee013ee35
Since the data returned by these calls can grow unbounded based on
various GET flags, we need to switch 'em over.
Bug: 27391893
Change-Id: Ie849ca30dbaaa91158da4c83675657715629a0ee
- Use new dynamic black level result key when available
- Write rational black level instead of integral value
- Fix incorrect thumbnail BitsPerSample metadata
- Fix incorrect date/time format string
- Fix unknown orientation
Test: dng_validate for captured DNGs is clean
Bug: 27452979
Bug: 25862995
Change-Id: I87b73f7a4f4ecc3cf6498c7db16df68dd0b0b43a
Propagate setVisible calls to current drawable. Take window visibility
into account when determining drawable visibility.
Bug 27461595
Change-Id: Ia1de1d1686f26c2cf27630c5bde70f23c2c332bd
It is possible for an activity to be in the stopped state without
setting it's visiblility to false in window manager.
For example, the home acitivty behind the lock screen. Since the
lock screen isn't an activity it doesn't affect the visiblity set
of the home activity, so AM doesn't tell WM to hide the app token.
However, AM uses another channel to detect that the device is locked
and moves the activity into stopped state. WM on the other hand also
detects that the device is locked and hides the window surfaces of
all windows behind the lock screen. So, at this point AM has also
told WM that the activity is stopped. Once you unlock the screen
AM resumes the activity but doesn't report any visiblility changes to WM
since it's internal state didn't change. So, if you go from the home
activity to another app the home activity window will be destroyed
before the activity is stopped because mAppStopped is set to true.
We now set mAppStopped to false when the activity is resumed.
Bug: 27286867
Change-Id: Ic75456d30abd582fa44f932f5aeeb449950157ee
When legacy camera device is the producer of its output surfaces,
it needs to connect to the surfaces before using the surfaces
and disconnect from the surfaces after finishing using the surfaces.
Bug: 27331537
Bug: 27383242
Change-Id: Ia3e2ca142f7973face0f60fbfc96935054aac1c4
Added a check inside PackageManagerService to make sure data for a
package with a DO or PO for the running user is not cleared. Currently,
the 'pm clear' command goes through without any such checks.
Bug: b/27243904
Change-Id: I87d4ad2db031f47946f34627a5ee465ef144f85e
Make referenced constants @SystemApi and change argument name to
match its actual meaning.
bug: 25441401
Change-Id: I4277799b95f394edb4305be05f57437717a86691
Apps/libraries/widgets that wish to display text their own way (think
WebView/Chrome) or do some resource management themselves should be
able to use LocaleList.getAdjustedDefault(), the same way we use it
in Paint and Resources.
Bug: 26988869
Change-Id: I4b6aac7adb5da15459dca92574923f33f0702433
The reason was that we're using a different textview there which uses
a SimpleLayout. This is now changed to also use the same one.
Bug: 27436418
Change-Id: I7e6dd503e09cf69ba8ed5bb6dd07fc79ee519cd3
Add a new sms filtering method to let the carrier app tell
the platform to keep the message while keep it invisible to
the end user.
BUG: 27292180
Change-Id: I306a7dbde6dcb9929e358b3d1bbda9807ef35943
When using drag resizing it is difficult to keep big surface
surfaces (e.g. main app windows) and child windows in sync
as we resize. Furthermore it's difficult to resize child windows
quick enough to achieve more than a few frames a second as we have
to propagate through the client UI thread. Our new implementation
uses window scaling.
Bug: 26454664
Change-Id: Iac96619cefc075b1412cfeba3d3c9bcd7ce22f52
We want to compute the frames for pinned like we do for freeform
as we are not constraining layout to the suggested display area by the
PhoneWindowManager. Also update applyGravityAndUpdateFrame
to not clip frames to the display for child windows. In the case
of computeFrameLw this would not be a problem as we would then
go on to overwrite mFrame anyway, but in the case of repositionChild
it could create issues (where we have applyGravityAndUpdateframe without
compute frame).
Bug: 26454664
Change-Id: I6fd4c9f37060d51003d041566368edd2b9eb7afd
