* Revert "Exposed tooltip longPress action when available (#117338)"
This reverts commit 38e3930f39d8efca0a27f9663129d27a9093cca8.
* Revert "fixes android_semantics_integration_test to expect long press for tootip"
* Prevent possibility of null exceptions on widget.bottomSheet
* New approach that fixes bug in updating the bottomSheet
* Real-world test for bottomSheet error
* Allow bottomSheet to animate out after being killed, even if it was rebuilt
* Go back to the simple solution of SizedBox.shrink
* cacheWidth cacheHeight support for web canvaskit
* comments
* clarifying comment for loadTestImageProvider class
Co-authored-by: alanwutang11 <alpwu@google.com>
* Replace PanGestureRecognizer in TextSelection with TapAndDragGestureRecognizer
* add tracking of _DragState to new tap_and_drag recognizer and remove some legacy double tap code from text_selection.dart and add logs"
* add dragTapCount, a tap count that is persistent for an entire drag and is set to null on drag end vs the regular tap count which is reset on a timer
* basic double tap to drag functionality and add a local dragTapCount in text_selection.dart to use with the timer callback
* Add offsetFromOrigin and localOffsetFromOrigin to DragUpdateDetails similar to LongPressMoveUpdateDetails, eliminates the need to hold the state of lastDragStartDetails
* make a generic baselongpressgesturerecognizer
* Revert "make a generic baselongpressgesturerecognizer"
This reverts commit aad8f7433bd01e4cd016d527af832c3b1f15fac5.
* rename tap_and_drag to selection_recognizers
* add mixin for consecutivetap
* tap and long press gesture recognizer
* Revert "Revert "make a generic baselongpressgesturerecognizer""
This reverts commit 181350c36718f644eada3e45c1b7b5939f90a340.
* Revert "Revert "Revert "make a generic baselongpressgesturerecognizer"""
This reverts commit 4d69775967858dfd66dd9429e1713da598908a85.
* Add support for secondary button clicks on drag gesture recognizer and separate drag end and tap up callback
* get test running
* rename tapCount to consecutiveTapCount
* dispose timer properly
* add some comments to tests
* Add comments
* Make ConsecutiveTapMixin private and move logic to increment tap count into mixin
* stop tracking pointer when gesture is rejected and detect drags on touch devices
* onCancel for TapAndDrag
* have the TapAndDragGestureRecognizer handle tap downs and tap ups on touch and mouse devices
* add drag to move cursor for android and iOS, and pointer device kind to DragUpdateDetails
* get tests running
* refactor TapAndDragGestureRecognizer moving some logic into _check methods
* Handle cancel properly on TapAndDragGestureRecognizer, having both onTapCancel and onDragCancel, also fix tests
* Fix test mouse drag selects and cannot drag cursor, save _initialPosition based on dragStartBehavior (either on tapDown or dragStart)
* determine if drag has a sufficient global distance to accept and fix some cancel behavior, making _checkCancel clearer
* give up pointer on drag end
* properly stop tracking pointer, fixes test for right click on Apple and non-apple platforms
* clean up some comments from last commit
* remove drag on touch for now
* fix Can select text by dragging with a mouse due to dragStart only being fired on the first PointerMoveEvent, the previous pan gesture recognizer would fire both dragStart and dragUpdate
* Revert "fix Can select text by dragging with a mouse due to dragStart only being fired on the first PointerMoveEvent, the previous pan gesture recognizer would fire both dragStart and dragUpdate"
This reverts commit 124dc79bc3389672c76d7c014ce04edab297abc6.
* correctly use _initialPosition for checkStart and call _checkUpdate after _checkStart if localDelta is not zero
* updates
* fix double tap chains
* Add docs
* Address analyzer
* more analyzer, only issues left are with print statements
* add deadlineTimer to fix conflict with ForcePressGestureRecognizer
* Revert "add deadlineTimer to fix conflict with ForcePressGestureRecognizer"
This reverts commit 3b29ddfff4cde4845edd481ecefb789fea2a0781.
* remove unecessary changes to tests
* secondaryButton should not drag
* Revert "Revert "add deadlineTimer to fix conflict with ForcePressGestureRecognizer""
This reverts commit 0a008f029f5796acd48c17c1897c0b700d5ef3a7.
* updates
* Revert "updates"
This reverts commit 4803b8443a2b67f0b8d29e9a01f712dfcb0f588c.
* Revert "Revert "Revert "add deadlineTimer to fix conflict with ForcePressGestureRecognizer"""
This reverts commit 79251a7af88d5dbb1460a960afc77e65dea18bff.
* fix shift + tap + drag tests, this was happening because a double tap + drag was being registered and not a single tap, added a duration to pumpAndSettle to fix this
* remove TapAndLongPressGestureRecognizer
* fix cupertino text field tests related to shift + tap + drag
* deadline timer try 2
* more logs
* Should reset taps when tap cancel is called, and should wait until gesture is accepted to initiate a drag
* should clear _down and _up when gesture is rejected
* remove erroneous log
* fix selectable text double tap chains test
* dont restart timer until tap up
* reset consecutiveTapCount on drag end
* fix selectableText test
* fix material text field tests
* reject TapAndDragGestureRecognizer when it is neither a tap nor a drag
* remove prints
* clean up
* shift aware
* clean up
* fix cupertino test
* fix text field focus tests
* Add 100ms delay to cupertino test, to prevent a double tap
* clean up test comments
* add comment to test
* uncomment test
* remove longpress changes
* Fix drag on mobile
* remove debug
* Fix drag to move cursor on iOS
* left over from drag fix
* add tests for drag on touch devices
* add test for double tap + drag mouse devices
* add tests
* Fix bug where initialPosition was used before it was set
* Address some review comments and fix issue where if double tap was held too long then long press gesture recognizer would take over
* remove _isDoubleTap flag since it is no longer needed due to previous commit
* Add docs for onTapCancel and onDragCancel
* analyzer fixes
* Do not test selection handles on macOS, since macOS does not support touch
* Add assert for dragStartBehavior
* add double tap + drag tests to cupertino
* use kDoubleTapTimeout instead of const Duration(milliseconds: 300) for readability
* analyzer issues
* update docs
* update more docs
* address comments
* more doc updates
* fix docs
* unused import
* fix docs
* Add more tests
* Add more tests and reject a tap up if we have exceeded the tap tolerance
* updates
* Address comments
* fix test naming
* update documentation
* move selection_recognizers to selection_gestures
* fix analyzer
* fix analyzer
* keysPressedOnDown instead of isShiftPressed
* update docs
* update docs
* Add drag update throttle to TapAndDragGestureRecognizer
* update comments
* missed from merge
* Replace _ConsecutiveTapMixin with _TapStatusTrackerMixin
* updates
* correctly cancel tap when when past tap tolerance with new implementation
* Should call tap and drag cancel if we are giving up a pointer without succesfully tracking a PointerUpEvent
* comments
* move pastTapTolerance to tap tracker
* move pastTapTolerance to tap tracker
* clean up check for nulls and remove use of consecutiveTapCountWhileDragging
* move call to super.acceptGesture to top
* remove print
* clean up
* Fix tests where both PanGestureRecognizer and TapAndDragGestureRecognizer lost
* clean up
* _GestureState -> _DragState
* more docs clean up
* more clean up
* Add onSecondaryTapCancel
* Add docs
* more docs
* Fix broken isPointerAllowed when attempting a right click drag - the _initialButtons is never reset
* revert debug flag
* make primaryPointer private
* Add support for upper count limit in TapAndDragGestureRecognizer, the tap counter should not be allowed to grow infinitely unless that is desired
* fix analyzer
* Use new TapDrag details objects and callbacks
* clean up docs
* clean up and add test for upperLimit
* Add docs for TapAndDragGestureRecognizer and remove some ambiguity of onStart onUpdate and onEnd parameters
* Address review comments
* analyzer fixes
* Call cancel before rejecting the gesture so we can still access _initialButtons
* Recognizer should reject any pointer differing from the original
* Revert "Recognizer should reject any pointer differing from the original"
This reverts commit afd9807480bd11e119bdd2b7d520631511973bab.
* Address reviewer comments
* Correct cancel behavior
* Fix consecutive tap + drag because _dragStart state was not being set when consecutive tap is greater than one
* Add more tests
* Add documentation on behavior with TapGestureRecognizer and DragGestureRecognizer
* more docs
* more docs
* remove comments
* updates
* fix multiple pointer behavior
* only handle the primary pointer
* Clean up dangerous assumptions in gesture details objects
* forgot from rebase
* update docs
* updates
* Clean up some redundant code
* remove whitespace
* fix tests as a result of #115849
* update test docs
* Fix same test from last commit for material variants
* More clean up of redundant code and update docs
* Clean up didStopTrackingLastPointer and untie TapAndDragGestureRecognizer cancel behavior from TapStatusTrackerMixin.currentUp state
* untie pastTapTolerance
* updates
* Add slopTolerance
* update docs
* Have secondary tap handled by TapGestureRecognizer
* update docs
* fix analyzer and address comments
* Add more docs
* Update cancel behavior tol not call on tap cancel when a drag has been accepted
* Change cancel behavior to only cancel if the tap down callback has been sent and merge tapcancel and dragcancel
* update docs;
* Rename selection_gestures to tap_and_drag_gestures
* Address some reviewer comments
* make deadline and slopTolerance private
* updates
* updates
* Address review comments
* remove _initialButtons
* fix docs
* trackTrap -> trackTap
* fix analyzer
* Add test to verify that tap up is called when recognizer accepts before handleEvent is called
* implement Diagnosticable for Details objects;
* sentTapDown == wonArenaForPrimaryPointer, so the implementation now only uses sentTapDown
* Count user tap up immediately and do not wait to win the arena
* Do not need to call super from TapAndDragGestureRecognizer.acceptGesture anymore because mixin implementation is gone
* Do not start selection drag on Android, iOS, and Fuchshsia touch devices if renderEditable does not have focus, this fixes many scubas
* Address reviewer comments
* fix test
* TapAndDragGestureRecognizer should wait for other recognizer to lose before winning the arena
* Address review comments
* Dont check for drag if the start was already found
* Only check for a drag if it has not already been found"
* fix from rebase
Co-authored-by: Renzo Olivares <roliv@google.com>
* Implemented Scrim Focus for BottomSheet so that assistive technology users can focus and tap on the scrim to close the BottomSheet, which they could not do before the change . The Scrim Focus's size changes to avoid overlapping the BottomSheet.
* isCanvasKit implement and test
* isCanvasKit implement and test
* ++
* forgot license
* make isCanvasKit a getter
* addressed comments
* forgot to change names of integration test files
* typo
* simplified tests
* comments
* Reland "Inject current `FlutterView` into tree and make available via `View.of(context)` (#116924)"
This reverts commit 9102f2fe0bd26db6074ac4a17785296cd341ecb9.
* remove window placeholder
* Reland "Adds API in semanticsconfiguration to decide how to merge child semanticsConfigurations (#110730)"
This reverts commit 7549925c8c44dea92c7bc75be676c17b7613f87f.
* makes markNeedsSemanticsUpdate more robust
* address comment
* wip
* remove temp text file
* fix tests
* add test
* default to off
* restore gitignore
* update
* apply annotation to cupertino icons as well
* update reference to library in icon_tree_shaker.dart
* update tests
* fix tests
* remove hack to skip non-const check on web
* add hint about how much reduction and test
This improves the documentation of the `compute()` function as follows:
* Instead of making `compute` a top-level constant, we make it a
function. This allows the generated API docs to show a function
signature with parameters and their names, making it *much* clearer
to users what function is being documented.
* We mention that on web-backends this is running `compute()` on the
normal eventloop whereas on on native platforms it runs in a
separate isolate.
* We mention that callback, message and result have to be sendable
across isolates. We also mention that they may be copied.
* We link to both `Isolate.run` & `SendPort.send` for more information.
* We use `M` for the type of `message` instead the rather confusing
`Q`.
