This should be the last keyboard CL.
- Fix bug: couldn't focus All Apps button in some cases when the All
Apps column was skipped over. Also added test case for this.
- Stop explicitly passing countX and countY to handleKeyEvent, as
these had to match the matrix dimensions anyways.
- Rename createSparseMatrix() - there were 3 methods of the same name,
but all had different purposes. This is confusing both from a
readability standpoint and also when looking at stack traces.
Change-Id: I08ba8411674fcea43a608856c114dee8dbd22398
- Handle NextPageFirstItem as first focusable item in reading order
- Handle PreviousPageLastItem as last focusable item in reading order
- Check the hotseat after the workspace in both cases above
- Dpad horizontal navigation (left/right) uses these as a last
resort (Rule3) to guarantee an item takes focus if a page exists
Note that it is necessary to search for a focusable item because
widgets are not yet focusable.
Bug: 25591057
Change-Id: I953648bd76c657d660a38427fdd4108bf9963c23
The content provider is the first component started with the main process,
which makes it ideal for initiating LauncherAppState
Change-Id: I03db82a8949ad1376af53a0033592f421dbf0b15
- Pressing Tab wraps around to the first item from the folder name,
and vice versa when pressing Shift+Tab.
- When tapping off the folder while editing the text, the folder
requests focus. We handle the following cases from that state:
- Pressing an arrow key or Tab gives focus to the first item.
- Pressing Shift+Tab gives focus to the last item.
- Fix slight corner case where moving from folder name to an item
didn't update mIsEditingName to false. So when clicking off of
the folder, it gave focus to the folder (as mentioned above)
instead of closing the folder like it usually does when icons are
focused. Not a huge deal, but still worth fixing.
Bug: 25687579
Change-Id: I1bec844c8ccd09529a11b9e3a1d92b3bdf7b2eb3
Because going to overview mode scales down the workspace, it was
thinking the touch was moving even though your finger was still. If
the "movement" was large enough, it was treated as a scroll, causing
jank. This was especially prevalent on tablets due to their size.
Bug: 25779718
Change-Id: Idb7833e0087bd24ca840f6afc451bf221f6bc047
The extra folder cell space added in ag/790800 looks bad on phones in
portrait mode, so reverting it here. However, we keep the extra space
on tablets or phones in landscape mode, since it provides more room
for long app names (bug 22462641).
Bug: 25662215
Change-Id: I2a37b884458ee557c9b8cff0c3edef16bfc50efb
The first of these cases seems like they should already have been
present, and the last couple allow you to switch pages by hitting
arrow keys on the edges of the hotseat.
Bug: 25589939
Change-Id: I9378b209250f7dc376fa97efde979bcee2979537
The All Apps button creates a number of edge cases, mainly because it
causes the hotseat to sometimes have an extra column than the workspace.
Previously, we sort of swept these problems under the rug by simply
ignoring the All Apps button if other icons were present in the hotseat,
with the assumption that those other icons should get focus instead of
the All Apps button. (If possible, we want to stay in the same column
when moving from the workspace to the hotseat.) But this doesn't always
work, as in the attached bug where the hotseat doesn't get focus at all
when the All Apps button is an obvious candidate for it.
By adding a specialized column in the focus matrix for the All Apps
button, we ensure that moving down to the hotseat stays within the
original column when possible, while also allowing the focus to switch
to the All Apps button if appropriate. Furthermore, we take care to skip
over the All Apps column when necessary in order to maintain all
previous functionality.
Bug: 25590522
Change-Id: I5d6a8ee69de8834314c4689246fe7d54329b2eef
There's a case where hitting the left/right arrow in the first/last
column wouldn't give focus to an icon on the next page because it is
vertically too far from the first icon to be considered. This makes
for a bad user experience, because there is no way to switch pages
with the keyboard. So now we brute-force search for an appropriate
icon if none was found in this case.
Bug: 25434120
Change-Id: Ifdead0e3b458717ccb33e2f0ec7c15f1fcce4b95
> As of android M, the platform somtimes throws DeadObjectException
instead of TransactionTooLargeException
Bug: 25305031
Change-Id: I3a98db8a5041aef01640d16f9beb71b72bf598f1
When an icon grabs focus, it already snaps to its own page, so
the explicit snapToPage() is redundant. And in the case where no icon
will grab the focus, the explicit call puts us in a weird state
because the page snaps but the focus doesn't change (see bug).
(This happens, for instance, if you are moving from an icon on the
bottom right of page 1 and the only icon on page 2 is in the top
left. In this case, the second icon won't take focus because the
angle is too wide according to FocusLogic#handleDpadHorizontal().)
Bug: 25434120
Change-Id: I17b0ef7d86426a2599bc948bf2313dca7e112364
Clearing the focus was causing the first child in the new page to
immediately take focus, which caused FocusIndicatorView jank when
using a keyboard.
Bug: 25256728
Change-Id: I5ab31ebc3fe370d7ac9e9792b30dab3467023738
> Pressing 'enter' enters the focus inside the widget.
> If the widget has only one focusable, it directly performs click
> Pressing 'esc' brings the focus back to the parent host.
Bug: 25199581
Change-Id: Ia7fd5d929f5945d61173abd349660a5efbe74afc
- Seperate search bar height from its top and bottom paddings
- Report this height plus 8dps top and bottom for the widget options
- Use a separate bottom padding to adjust spacing between QSB and
the workspace
- Add tests to verify the height of the search bar in both normal and
tall modes (ran on Nexus 6, 7, and 10)
Change-Id: Ia7557785d1b38194eee869d0460456e8f33447d5
There is a race condition that occurs primarily (maybe entirely) with
Android Auto, it seems because they hijack the Home intent or something
similar. I'm not exactly sure how Android Auto works, but if I pair my
phone with the Desktop Head Unit (car dashboard emulator), I can repro
the NPE fairly easily by simply force closing Android Auto and then
disconnecting my phone from the DHU. If I don't force close Android
Auto, then pressing home launches Android Auto or other apps that I
assume handle some custom intent, such as Car Home Ultra, instead of
normal Home intents such as Launcher3/Google Now Launcher. So I think
what's happening is that, when the phone is disconnected from the car,
Android Auto restores and launches the real home intent (Launcher 3)
around the same time that it destroys the previous home intent
(Android Auto, Car Home Ultra, etc.). This could cause the NPE if both
intents are actually Launcher 3, as is the case when Android Auto is
already closed, because mWorkspace is set to null in
Launcher#onDestroy() (something like onNewIntent() --> post() called
--> onDestroy() --> post() runs). This is consistent with the fact
that I can guarantee a repro if I use postDelayed() instead of post().
Long-winded explanation aside, I think this fix is safe, especially
since we already have a null check for mWorkspace in onNewIntent(),
just not inside the post().
Bug: 24610231
Change-Id: I42f75b83946f375d947be1961a1f2a03a3707a84
Previously, any child of ShortcutAndWidgetContainer was added to the
matrix, causing widgets (which aren't focusable) to be considered as
potential targets to gain focus when an arrow key was pressed. But if
the algorithm chose them, they couldn't take the focus so nothing
happened (i.e. the focus stayed on the app/folder it was on before).
Bug: 25126768
Change-Id: Id55fc310f7f58fb8795cce51dcefe4fd1210f788
Sunny Goyal
Tony Wickham
Winson Chung