Provide a custom data slot in WebHistoryItem to store application data in the
history item. A client of WebBackForwardList can insert their custom data into
the item when created. In onIndexChanged, the custom data can be read by the
client.
The Find dialog draws on top of the WebView, including any
layers attached to the bottom of the screen. Adjust the view
dimenions returned to the layer code by the height of the Find
dialog if visible.
http://b/2453769
Merge commit '3ca122df9e9a25f7ed1c39ef18c5c4ba2e316fe8'
* commit '3ca122df9e9a25f7ed1c39ef18c5c4ba2e316fe8':
If we get TOUCH_UP before getting the WebCore's
which should not block either UI or WebKit. It handles
local file access, cache access and trim cache.
Move createCache, saveCache and most of getCache out
of WebCore thread so that slow IO and database will
not affect loading performance. getCache can be still
called from WebCore thread in the uncommon cases
like redirect and POST validation.
Move cache ticker from WebCore thread to WebViewWorkerThread.
Move setCookie from WebCore thread to WebViewWorkerThread.
Remove the unreferenced files in the cache directory
while trim cache.
Confirmed with our SQL expert, Vasu, there is no need
to wrap clearCache with end/startTransaction any more.
http://b/issue?id=2414792http://b/issue?id=2475242
If nativeMoveCursorToNextTextInput does not succeed, do not set the
default selection or perform the other actions which assumed that the
move worked.
Part of fix for http://b/issue?id=2478052
Requires a change to external/webkit.
This can cass a NPE since our Timer would have been destroyed by then.
Add call to VideoView.stopPlayback() when the view is dismissed by the user.
Fix b: 2476002, 2475786, 2475786
The history picture and the layers drawn may be out of sync,
so for now, don't draw the layers when the history picture is
drawn.
Also, remove native class check from drawExtras() since it
is already checked by the caller.
fixes http://b/2457350
fixes http://b/2453780
This is the framework part, moving classes around so the framework
no longer needs to link to android-common. Makes some APIs public,
others that didn't need to be public are private in the framework,
some small things are copied.
Drawing elements that appear atop or below layers need to be
drawn both in the proper order and with the correct canvas to
respect clipping and the matrix.
The logic of what to draw is still in WebView.java, but the
actual drawing calls are now triggered inside the layer code.
This still draws layers incorrectly after drawing the history;
will fix this in a future CL.
Move drawing to WebView.cpp.
Use inverseScale to simplify viewPort metrics.
Remove root layer; let WebView.cpp handle it exclusively.
Requires companion fix in external/webkit.
http://b/2457316http://b/2454127http://b/2454149
If mPreventDrag is cancelled due to time out, remove
the pending TOUCH_EVENTS and then send an ACTION_CANCEL
and skip the rest of touch events.
This should address the drag problem Flash plugin has.
Fix http://b/issue?id=245053
the pending TOUCH_EVENTS and then send an ACTION_CANCEL
and skip the rest of touch events.
This should address the drag problem Flash plugin has.
Fix http://b/issue?id=245053
This is to fix a particular case where the navigation cache
does not recognize that a textfield is now in focus. A message
was sent from the WebCore thread telling us to open the IME on
the textfield. Open it with the WebView as the served view.
Fix for http://b/issue?id=2457459
The message handler is instantiated on the WebCore thread and its presence is
checked on the browser thread. This requires synchronization.
Change-Id: I4b71c7a2470b60fa273dc2bcb46f645ed135ee11
Blocking network images is not as secure as completely blocking network loads.
Leave setBlockNetworkImage so clients can reduce bandwidth by disabling remote
images. This will allow file:// and content:// URIs to load.
Previously, we checked to see if the IME was active on the WebTextView,
in an attempt to fix http://b/issue?id=2266066 . If so, we move the
textfield on screen. However, that is resulting in
http://b/issue?id=2147168 , where a textfield far offscreen is being
scrolled on screen at the end of a zoom, and with this change, the
latter bug is fixed. Further, I am no longer able to reproduce the
earlier bug with this change.
Add a parameter for clearTextEntry to disable the focus controller,
and call clearTextEntry when releasing the trackball.
Requires a change to external/webkit
Fix for http://b/issue?id=2340871
