After finding a window in the window list we turn around and look in
the AppWindowToken.windows list for it. If it is a child of a window
in that list we should use the parent windows index as the search
result. Instead we gave up and ended up inserting the window at the
beginning of the windows list.
Bug 7357465 fixed.
Change-Id: If77f343b8597bfbb0b7fa41dedf7972d78d03020
Fix for bug 6630331.
This is a workaround for a shortcoming in the Thai shaper in Harfbuzz.
De facto (but not documented) shaping rules dictate that a tone mark
followed by a SARA AM should be rendered with the NIKHAHIT below the
tone mark, rather than above it as would be indicated by a
straightforward interpretation of the GSUB and GPOS rules. This patch
adds an additional substitution of the form MAI THO + SARA AM ->
NIKHAHIT + MAI THO + SARA AA. The three tone marks affected are MAI EK,
MAI THO, and MAI CHATTAWA, based on an usage analysis of the Thai
dictioary included in ICU.
When we upgrade to harfbuzz-ng, we can revert this workaround.
Change-Id: I209ed094e7c89ea7f6fd6e983d37e396c6da1f64
- remove unnecessary calls to resetRtlProperties().
- now reset of RTL properties will only be done when adding a View
(and no more when removing it)
Change-Id: I0d42128c9f7df6085fb92bb5af5c9bd4d1ba88a3
getChar(String) and getChar(String, char) had wrong description in their
javadoc.
Also most of the get methods with default value were missing the corresponding
@param tag for the default value parameter.
Change-Id: I0f38b3caacf9cabd70e0c1ada36af3662f5566d0
The color of timer and backgroud in MediaController
are too closer to distinguish.
Change-Id: Id60ecbc26233857c7ef291ef891c9d4720309dfa
Signed-off-by: Roger Chen <cxr514033970@gmail.com>
Cherrypicked from external contribution.
b/7648349
VelocityTracker is a powerful tool---too powerful, in this
case, because many devices (I'm looking at you, Nexus 7)
have touchscreens that report total garbage on takeoff and
landing.
PanelView now uses its own incredibly crude velocity tracker
called FlingTracker, which implements a short sliding window
of the last 8 motion events (7 intervals) over which touch
velocity is averaged. There's also a little bias toward more
recent touch events so that the overall velocity of small
circular gestures will tend to favor the exit tangent.
The end result is a primitive low-pass filter on touch
velocity that should help us avoid situations where one (or
even two!) stray MotionEvents at the end of a gesture won't
invalidate the overall thrust.
Bug: 7422342
Change-Id: Idae38d1957727e400493324af4eee357ba5baa27
* commit '49be1ef39cfa2c718cb56dee088338ea97019c11':
Doc change: cleanup filtering sidebars and api level table. Changes links pointing to billing and other play docs that have moved.
