Background:
The applications with the granted INTERNAL_SYSTEM_WINDOW and
INTERACT_ACROSS_USERS_FULL means that it could show the same
window for all of users. i.e. to use user 0 presents all of
UI things to all of users.
INTERNAL_SYSTEM_WINDOW usually comes with INTERACT_ACROSS_USERS_FULL
because it will serve all of users to know the information that
comes from framework and system server.
Solution:
Because SystemUI never restarts after the user changing,
ClipboardService can't tell if the callingUid has the the same userId
with the current user or not. The solution is to use the permission
check. Especially, INTERACT_ACROSS_USERS_FULL and
INTERNAL_SYSTEM_WINDOW. To check INTERACT_ACROSS_USERS_FULL by using
ActivityManagerInternal.handleIncomingUser.
Caution:
The application with INTERNAL_SYSTEM_WINDOW usually use user 0
to show the window. But, the current user is user 10, WindowManager
know the focus windows is belong to user 0 rather user 10. That's
why user 10 can't copy the the text from systemui directly reply to
the other applications.
Readability:
ClipboardService use callingUid everywhere but actaully it is not
appropriated to fix this kind of bug. This patch refactor the naming
to produce two name. i.e. intendingUid and intentdingUserId that are
validated by ActivityManagerInternal.handleIncomingUser.
Test: manual test
Test: atest android.widget.cts.TextViewTest
Test: atest CtsTextTestCases
Test: atest CtsContentTestCases
Bug: 123232892
Bug: 117768051
Change-Id: Ie3daecd1e8fc2f7fdf37baeb5979da9f2e0b3937
Otherwise it's hard to keep track of things happening or not, since
the request would be postponed, leading to bugs elsewhere.
Test: Wait for time to pass on lock screen
Test: Enter/leave AOD
Fixes: 132955195
Change-Id: I54f806f9dd2d1cf4acac420bab08d4579f867a5f
This reverts commit 1600c5a615.
Reason for revert: Rolling forward for Q-Finalization
Bug: 129975435
Change-Id: I1ffb8162cb5e6f386fd3c417fabfd4298ef86ffd
A previous change restricted rules that are added to the rule dependency
graph to resource ids rather than just integers greater that zero.
Restore the previous behavior and also allow for negative resource ids
to work as expected.
Bug: 132447676
Test: loaded maps and view looks as expected
Change-Id: I88d86c77c696d1e494a743fdd2e398283383c64e
This reverts commit f6ed8afa40.
Reason for revert: QT SDK Finalization. Will be merged again on/after May 13th
Bug: 129975435
Change-Id: If94098b7cc9cf75cf9782d2b70e01881f9c40430
variants
Updated various framework Views to have equivalent BlendMode APIs
to replace the deprecated PorterDuff equivalents.
Updated InspectableProperty annotations to refer to the same
xml attributes as the original tintmode APIs
Bug: 126726419
Test: Added CTS tests to verify new BlendMode APIs
Change-Id: Id9ab36d3d4d29f351250723e9d13d49bc6062c83
Merged-In: Id9ab36d3d4d29f351250723e9d13d49bc6062c83
As per the suggestion from API council, we now have a subclass for event
of each category.
Bug: 129344540
Test: atest frameworks/base/core/tests/coretests/src/android/view/textclassifier/
Test: atest cts/tests/tests/view/src/android/view/textclassifier/cts/
Test: atest frameworks/base/packages/ExtServices/tests/src/android/ext/services/notification/SmartActionsHelperTest.java
Change-Id: Ic43b33c2176447c40e64bd0e410e906d5fb9c4cc
For splits with package id 0x80 and higher, the resource ids are
negative. RelativeLayout builds a dependency graph to indicate in which
order the layout height and width need to be processed. Since the ids
are less than 0, RelativeLayout is incorrectly assuming the layouts are
not order dependent.
Bug: 72869300
Test: manual
Change-Id: I98f58f11733c2976fc5c1b4152949cf80660f657
It's unintuitive that hasSelection() can return false, while
getSelectionStart() and getSelectionEnd() return meaningful non -1
values.
Test: TreeHugger
Change-Id: Iea2157f4b3b2c13c74547ba31eb4131496f939cc
TEST_MAPPING allows sharing test configs, use that
to avoid duplication.
Test: atest --test-mapping frameworks/base/core/java/android/widget
Change-Id: I441f36fabcc26bc8198ab49eebc433932b344ae7
The CL updates Magnifier#getPosition() and Magnifier#getSourcePosition()
to return coordinates relative to the main application window, rather
than relative to the surface backing it. Working with window relative
coordinates is the better approach for public APIs. The CL also makes
View#getLocationInSurface public. Magnifier CTS tests still need it,
when testing MAX_IN_SURFACE behavior.
Bug: 124012182
Test: atest CtsWidgetTestCases:android.widget.cts.MagnifierTest
Change-Id: I684f27aa466e0390c7801b616331f4682784c423
After method rename procedure b/126354063 we discovered that
androidx-platform-dev already depends on that methods added during fixit
week, so revert deletion of these methods and deprecate them.
Bug: b/126354063
Test: PopupWindowTests
Change-Id: I17b985542ffd661358059740bdf620e1bfb5b550
This CL renames 4 methods for PopupWindow based on API council
suggestions. Also remove unused import in order to `repo upload`
Bug: b/126354063
Test: All ListPopupWindow, MenuPopup and PopupWindow passed
Change-Id: I5db93b1c18a7a3dc9033cb9c5bbe89107ac22024
If they were null, then the Parcelable would fail to work.
Bug: 126726802
Test: manual
Change-Id: I7929ffa2f20e5de1c8e68e8263cca99496e9d014
Exempt-From-Owner-Approval: Trivial API annotations
Add range annotations to ToggleButton#getDisableAlpha() and
ViewFlipper#getFlipInterval().
Rename AbsListView#getDrawSelectorOnTop() to #isDrawSelectorOnTop()
Test: m framework
Change-Id: Ieda5b45d10677a9be19a4fc61e8e25d491633428
Fix: 126391791
In API 26 we deprecated ZoomButton and ZoomButtonsController, but
we missed ZoomControls that is used together with the former classes.
Test: None
Bug: 126610647
Change-Id: I0207551b2aaf511ad37ad759475916616fe9240b
The CL fixes the light source position when the magnifier is first
drawn. Before this CL, the source position was set only after the
magnifier was drawn for the first time, causing an incorrect shadow for
the first draw. This was noticeable if the magnifier was trigerred and
not moved (the shadow was updated and started looking fine when the
magnifier was moved and a new draw happened).
Bug: none
Test: manual testing
Change-Id: Iabdd34430a5def3608197294cf694c5045a7d8d3
This CL fixes a bug in AbsListView when it move offset,but its ChildCount was zero and the firstChild
view was null object reference
Bug: N/A
Change-Id: I1c9ccd9807201fd107fe14e051f3f8e7b0cc530d
Signed-off-by: cnx421 <cnx421@gmail.com>
