| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
| |
removed, there is a slight chance that getRealObject()
for WebView's JavaObject be null.
Fix http://b/issue?id=2159815
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Requires a corresponding change to frameworks/base. We were previously
letting WebTextView handle clicks, determine the change in selection,
and pass that down to webkit. This sometimes resulted in a different
placement of the caret if the WebTextView and the webkit-rendered
textfield did not line up exactly. Now, we pass the click directly
to webkit, which determines the new selection and passes that info
back to the WebTextView. This also has the benefit of letting
the WebTextView reflect changes in the selection that originated from
webkit. Also remove some unused parameters.
|
| |
|
|
|
|
|
|
|
|
|
| |
Fix for http://b/issue?id=2021424. When trying to draw the matches,
if we find that the page needs to be scrolled to put the match on
screen, we request a scroll, do a viewInvalidate and stop, so that
the matches will be drawn after the invalidate. Unfortunately, for
some values, we do not end up scrolling, because contentToView turns
the scroll into a no op. So we forever request scrolls, and never
draw the matches. Fix this by getting a return value from scrollBy,
and if it returned false, go ahead and draw the matches.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
While the page is loading, the nav cache is not rebuilt.
Double-click zooms out the web page by using the nav cache to
find the left edge of the column -- but fails to work during
page load. This change rebuilds the nav cache (if the page is
loading) each time a mouse click is sent to webkit.
This doesn't fix the bug where the first double click doesn't
align the column correctly, but helps with subsequent clicks.
Also, pass scale information to getBlockLeftEdge so it can
restrict its search to the area that will be zoomed to. Default
to the point clicked if no alignment info can be found.
|
| |
|
|
|
|
|
|
| |
In CachedFrame, add methods to find the next textfield and to
determine which ImeAction should be associated with a given textfield.
In WebView, uses these apis to determine the ImeAction and jump
to the next textfield and scroll it on screen. Requires a change to
frameworks/base.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add api to get the left edge of the block from the current (x,y). The code
was copied from Cary's change.
Todo:
1.need some tuning as we can see from nytimes.com that some times the left
edge is not correct.
2.currently nav cache is not up to date while loading. This means the left
edge may not be correct during loading.
3.if (x,y) is over an edit text box, or image, it should return the left
edge of it. Currently it is not working as expected.
Added the code to store the extra scale factor, so that back/forward history
works correctly.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
We already cleared it in the above case, when the
user clicks on empty space. If the user clicks
on a node, and it is not a text input, clear it.
This fixes a bug on the Google home page, where
the user has typed into the textfield, but then
clicks on "more" and is then unable to click on
Gmail, which is covered by the (invisible)
WebTextView.
|
| |
|
|
|
|
|
|
|
| |
The cursor position is preserved, so additional trackball
navigation will start from where the cursor ring was drawn
last. Hiding the cursor gets rid of confusion where the
cursor ring stays visible when clicking on a link that
does not go to a new page, or when the new page has a
link at the same physical location as the old page.
|
| |
|
|
|
|
|
|
| |
This fixes a bug where the DOM has moved the focus, but the WebTextView
is still sending keys to the old focus. Requires a change in
frameworks/base. This part of the change just gives WebView a way
to get a pointer to the currently focused node (regardless of the
cursor).
|
| |
|
|
| |
user events)
|
| |
|
|
|
|
| |
This helps superpudu recompute the cursor ring more
frequently when the floating bar moves. It also helps
update the nav cache when navigating during a page load.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Longpressing on a WebView while the WebTextView is up calls
WebTextView::performLongPress(). This used to be fine, but now
we may still leave the WebTextView up while the user is longpressing
on something else. Check to ensure that the focus is the same
as the cursor before calling WebTextView's version. Also,
send a click if the user longpresses on a different WebTextView,
so it will now behave appropriately. Requires a change in
frameworks/base.
Provide a setter on CachedHistory to set the mousebounds, and
set it in selectBestAt, so that getSimulatedMousePosition
uses the correct information. Also edit some comments to refer
to the cursor rather than the focus. In WebView.cpp, use
the delete TypingCommand, making deleting and cutting work
once again.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Consolidate key handling. Simplify interface between android
and webkit so multiple clients can use common code to pass keys.
Add helper to return if cached node is plugin.
Use WebView::m_clickedOnPlugin to remember active plugins.
Add WebView.MOVE_CURSOR to replay key events the plugin
doesn't want.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Create a new function that sets the cursor data to the WebViewCore,
and call it each time we change the cursor. Prior to this change,
by touching on a non-cursored, non-focused textfield, you could
set the cursor on the current CachedRoot without setting the data
on the WebViewCore. Then, when we get a new frame cache, we
check the data on the WebViewCore (which is out of date). Since
it does not think there was a cursor (or it thinks it was somewhere
else), we do not restore the cursor on the new cache. This change
fixes that by ensuring that the cursor information is up to date.
|
| |
|
|
|
|
| |
In drawCursorRing, we potentially use a different value than bounds
for the cursor ring. If we use that other value, also use it for
the invalidation to clear it later.
|
| |
|
|
|
|
|
|
| |
Add clicking, key debug statements in WebViewCore.cpp.
Fix dumping nav tree to use NamedNodeMap.
Set click point to be start of text field, end of text area, middle
of others.
Make CachedRoot::getSimulatedMousePosition, callers const
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
When the WebViewCore picture set is recomputed, check
to see if the trackball has moved since the last time.
If so, rebuild the nav cache even if dom version checking
is disabled.
If there's no focus node ready to receive a key event,
check to see if the cursor is hovering over a plugin, and
give it the event instead.
Update all cursor state in both places it could change.
|
| |
|
|
|
|
|
|
|
| |
Generate mouse move event after sending scroll event
Move cursor update in nav cache to UI thread
Remove updating nav cache on mouse move
Remove node parameter from mouse move
Remove plugin focus hack from mouse move
Add interface to get native cursor position
|
| |
|
|
| |
also, set hasCursorBounds only if cursor ring is visible
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Make the blinking caret active when the user actually starts editing
text. In WebView.cpp, determine whether to disable the blinking
caret and pass that info up to the Java call. Requires a matching
change in frameworks/base. Also add some comments for clarification,
and remove duplicate calling points for functions. Also change
glue for nativeSetFocusControllerActive to reflect its new name/
function.
This change also reflects a behavioral change. If the WebView or
its window lose focus, we do not restore the blinking caret when
focus returns.
|
| |
|
|
|
|
|
|
|
|
|
| |
Everytime the picture set is rebuilt, the nav cache
is also built if the dom changes. This doesn't catch
javascript that changes locations of objects without
updating the dom. To handle these cases, do a hit-test
on the cursor location and see if it finds the same
node used to build the cursor cache.
This helps track the gmail message floating tool bar.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
WebViewCore.cpp:
Add parameter to CachedRoot::findAt to suppress check for
hidden nodes. This helps find a best match for newly built
nav caches.
Require that the new cursor node closely match the bounds
of the prior cursor node. This may need tuning.
CacheBuilder.cpp:
Remove isInput flag on cached nodes (no longer used)
CachedFrame.cpp:
Add hideCursor to complement clearCursor. Hide prevents
the cursor from drawing but does not move it. Clear removes
it altogether so that the next movement starts from the viewPort
edge.
CachedHistory.cpp:
Don't special case text fields when doing history navigation.
This special casing in part allowed setting the focus to
the homepage input, but since focus is no longer set by nav,
it is not required.
CachedNode.cpp:
Add hideCursor; clean up debugging
CachedRoot.cpp:
Use navBounds from history instead of cursor bounds to
determine next move. Clean up some obsolete code.
WebView.cpp:
Add hideCursor; call it when appropriate.
|
| |
|
|
|
|
|
| |
Since we have a way to bring up the soft keyboard manually
(holding down the menu key), and trying to bring it up
using our existing call results in a crash, only bring it
up when the click is on a textfield/textarea.
|
| |
|
|
|
|
|
|
|
|
|
| |
In the old navigation model, we set focus to a node, and checked to see
if that node was still valid (could have changed since we built the
navigation cache). Then we sent a message back to the UI thread to
tell it to draw the ring differently. In the new model, we are essentially
moving a mouse, so we do not concern ourselves with whether the node is
valid. Here I have removed function and jni to call sendMarkNodeInvalid,
the notion of an invalid node in WebView.cpp, and the INVALID flavor
of CursorRing. Needs a change in frameworks/base to work properly.
|
| |
|
|
|
|
| |
Remove obsolete recomputeFocus, old moveCursor code
Make sure focus is !null before using it.
|
| |
|
|
|
|
|
|
|
|
| |
Previously, many methods, variables, etc referred to the focus.
Now, they have been changed to refer to the cursor when
appropriate. Some references were changed to the focusCandidate,
since they may apply to either the focus or the cursor, which
is a candidate for focus. Also removed unneeded method
unblockFocus. Will not work without a corresponding
change in frameworks/base.
|
| | |
|
| |
|
|
|
|
| |
Instead of returning the node that currently has focus,
return the node that the cursor is on if it can handle
key events.
|
| |
|
|
|
|
|
|
|
| |
Add convenience routines to consolidate cache builder
and current focus fetches into one place.
Remove unused functions.
Use wantsKeyEvents attribute where applicable.
|
| |
|
|
|
|
|
| |
This change corresponds to a change in frameworks/base which changes the
name of the TextDialog to WebTextView. It also changed the method updateTextEntry
to rebuildWebTextView. So, the native side needs to call the newly named
method.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Older code treated the trackball as a four way dpad with
equivalents to moving up, down, left and right by generating
arrow key events. This change makes the trackball solely
generate mousemove events.
The old arrow keys in turn were mapped to be as close as
possible to tab-key events that moved the focus. The new
model leaves focus-changes to the DOM.
Clicking the dpad is distinguished from pressing the enter
key to be more compatible with desktop-authored web pages.
|
| |
|
|
| |
http://b/issue?id=1742161
|
| |
|
|
|
|
|
| |
remove unused commonKitFocus function in WebView
leave recompute focus position for now, but don't
look at node index as a condition to abort fixOutOfDateFocus()
|
| |
|
|
|
|
| |
isClick is always set to true by the callers, so remove
it and simplify the code
(http://b/issue?id=1666780; see also change in framework/base)
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Check to see if the potentially larger hit-test bounds
can be used in place of the normal bounds, or if the
normal bounds can be used in place of the individual
text bounds.
Construct a region out of the individual focus ring
rectangles, then see if any text is drawn inside the
bounds but outside of the focus ring. If not, use one
rectangle instead of the rings.
|
| |
|
|
|
|
| |
Besides sharing those routines in SkiaUtils.cpp, also remove our (now redundant)
android_setrect, etc. calls, since the type-conversion extras we added earlier
take care of this by using constructors.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If WebViewCore::sendFinalFocus is called from the UI thread,
postpone calls to WebViewCore::notifyFocusSet until after
the focus is set.
Don't allow WebViewCore::setSelection or
WebViewCore::passToJs to change the webkit
focus.
In WebView::notifyFocusSet, read the cache after it has been
updated.
In WebView::focusIsTextArea and WebView::nativeFindAll,
don't allow the new frame cache to fix the focus.
If CachedFrame::setFocus has already set the focus, don't
recompute its position.
Add/revise debugging.
|
| |
|
|
|
|
|
|
| |
pages. The previous fix, 142939-p9, allowed larger pages to be searched fully, but not with a length larger than a 16 bit number. In change 143089-p9, bitmap's width and height were promoted to 32 bits, so our previous fix is no longer necessary, and prevents larger pages from working. This removes the prior fix.
BUG=1730904
Automated import of CL 143515
|
| |
|
|
|
|
|
|
|
|
|
| |
only searches the first part of large pages. Cap the width and height of the page so we do not overflow a 16 bit int and think the page is smaller than it is when performing the find.
Original author: scroggo
Merged from: //branches/cupcake/...
Original author: android-build
Merged from: //branches/donutburger/...
Automated import of CL 143340
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
