| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently, when converting coercing a JavaScript string to a method parameter
of type Java char, we always use a value of 0.
However, when coercing a JavaScript object to method parameter of type Java
char array, we behave differently. In this case, if the element is a JavaScript
string, we use the first character of the string for the char value.
This change modifies the behaviour in the array case to match that in the
non-array case.
Bug: 5619677
Change-Id: Ifeee8940af0bb01df58eeff6c9dd7e1b40688790
|
|\
| |
| |
| |
| | |
* commit '7c2e38b61270efe55f065d765228ace66b54f169':
Enable SPDY protocol.
|
| |\ |
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Bug: 5524112
Reenable SPDY protocol.
Change-Id: I0144791020b64e20a12b7cd45ac2427100bef8a7
Signed-off-by: Selim Gurun <sgurun@google.com>
|
|\ \ \
| |/ /
| | |
| | |
| | | |
* commit 'e5c39206de45588678bf718a8274432068d702f6':
Remove inset for WebTextView
|
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Bug: 5383911
By removing the inset we cover webkit's text field completely
Change-Id: Ib3edf21beb4e26b38b2195a9e336523336f26eb7
|
|\ \ \
| |/ /
| | |
| | |
| | | |
* commit '0de683c1b300c8e788c178c63b7dc71189c0a3d8':
Fix combobox metrics and asset scaling
|
| |\ \ |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug: 5521553
Change-Id: I630eda2ffd90d0c986f00c5a2a4b523e93537192
|
|\ \ \ \ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Bug 5445861 : complex joining characters positioned in wrong place.
1. Set font size and scale for harfbuzz object based on Chrome's CL:http://trac.webkit.org/changeset/77504.
2. Truncate harfbuzz value to pixels for y position.
3. Ran through the page_cycler/intl2 tests with 5 iterations.
Change-Id: I38dcb07b2d23042e40781ea914c8c6e73f33c220
|
|\ \ \ \ \
| | |/ / /
| |/| | |
| | | | |
| | | | |
| | | | |
| | | | | |
into ics-mr1
* commit 'f51315051fbfa03d02d8c098ae70e9eb227d4200':
Fallback mode when running out of textures for layers
|
| |\ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
- Implements a degraded rendering mode when running out of textures
- Speed up rendering by reusing the same SkBitmap in RasterRenderer
- Some refactoring
bug:5279231
Change-Id: I52943a87bed56c49b01b47fea4fa1a5c49e09e93
|
|\ \ \ \ \ \
| |/ / / / /
| | | | | |
| | | | | |
| | | | | | |
* commit 'bbffb087457241fb32b7bf6327323f5c2695f8f7':
Use anti-aliasing in prefetch tiles
|
| |\ \ \ \ \ |
|
| | | |_|_|/
| | |/| | |
| | | | | |
| | | | | |
| | | | | | |
bug:5534278
Change-Id: I5929752a9edcf3cd1c51360e0dc4d8df6ae6bb1e
|
|\ \ \ \ \ \
| |/ / / / /
| | | | / /
| |_|_|/ /
|/| | | | |
* commit '1067d8e648ceae4db56ce5cc2b3dac1786d557d1':
Fix CTS regressions from merge to r74534.
|
| |\ \ \ \ |
|
| | |/ / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
WebKit change http://trac.webkit.org/changeset/73385 removed
overflow dimensions from the calculation of layer dimensions.
This means that the optimisation inside the nav cache to skip
building a frame cache when one of the layer dimensions is 0
is no longer valid (as the layer may contain only overflow
content).
Bug: 5572431
Change-Id: I3c973ad1ed57b07217d19f5d4fba7bfe522a2165
|
|\ \ \ \ \
| |/ / / /
| | | | |
| | | | |
| | | | | |
* commit '9074815a647bc5401963370e16b71e758fc3f550':
Support pausing drawing
|
| |\ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Bug: 5594608
Change-Id: I09590f8c04904957b079f0458022b960599eb47a
|
| |\ \ \ \ \
| | |_|/ / /
| |/| | | |
| | | | | | |
ics-mr1
|
| | | |/ /
| | |/| |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Bug 5268793
The scrolled area for WebView is 0,0 - maxX,maxY while
ScrollView can have any origin. On RTL pages, the origin
is negative when it can scroll. This scrolls the page
to the right when it detects an RTL page.
Cherry-picked from master CL: I6a91b97f
Framework CL: I2537b84a
Change-Id: I6a91b97f90f5ba4c323c56dda9d125108d9502c4
|
|\ \ \ \ \ |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Bug 5598631
The assertion message was misleading. This corrects it.
Change-Id: Ic609804ebac09e6d92fa11dc078aad62126b8690
|
|\ \ \ \ \ \
| |/ / / / /
|/| / / / /
| |/ / / / |
Change-Id: Idc88ff5f613a7c8f771b067f4a01967a32ede276
|
| |\ \ \ \
| | |_|_|/
| |/| | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Complex languages (Arabic, Thai, etc.) currently need to override the
base font (e.g. Roboto) with an explicitly loaded copy of the right
fallback font (e.g. DroidNaskh). To keep from running out of file
descriptors, a caching mechanism was put in place, but the cached
data contained the font size, so the first size seen on the page was
applied to all subsequent renderings of that font, regardless of the
size requested. Size is now also considered in the cache.
Bug: 5509443
Change-Id: I616954980122e89da14ce86efe99d5e4343a40c2
|
|\ \ \ \ \
| |/ / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
ics-mr1
* commit '7c7fadaed03fc262d86394fc08ae3948ebc904e1':
Add support for Bengali and Tamil gpos/gsub tables
|
| |\ \ \ \
| | |_|_|/
| |/| | | |
|
| | |/ /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
HB_Script_Bengali and HB_Script_Tamil are now recognized in
FontAndroid.cpp and map to Lohit-Bengali.ttf and Lohit-Tamil.ttf,
allowing Harfbuzz to properly load the gpos/gsub tables for those
languages.
Bug: 5578901
Change-Id: I20ed542dd5a616535222dcb5c808f3716a062e80
|
| |\ \ \
| | |_|/
| |/| | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
bug:5367327
Cherry pick of: dc193f393a845e88f55ecf0055386fc4e4204537
Change-Id: Iea6e4a564fc003d8e2ad8e1d6d9d9b64d016e43d
|
|\ \ \ \
| |/ / /
| | | |
| | | |
| | | |
| | | |
| | | | |
youtube video" into ics-mr1
* commit 'c97291b2e7a966abf8dbad5cf2e0e8b033c3cd90':
Add necessary null pointer check to avoid crash on youtube video
|
| |\ \ \
| | |/ /
| |/| |
| | | | |
into ics-mr1
|
| | | |
| | | |
| | | |
| | | |
| | | | |
bug:5562347
Change-Id: I9b53a58137d70955bde5334c7c8b581343a8c2ce
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Bug 5268793
The scrolled area for WebView is 0,0 - maxX,maxY while
ScrollView can have any origin. On RTL pages, the origin
is negative when it can scroll. This scrolls the page
to the right when it detects an RTL page.
Framework CL: I2537b84a
Change-Id: I6a91b97f90f5ba4c323c56dda9d125108d9502c4
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Getting the app package name from JNI. Cache the string after the
first call.
Add a header to every request with:
X-Requested-With: com.package.name
Change-Id: If9d8c131f6c4203036678516ccae6a7ec9131df9
|
|\ \ \ \
| |/ / /
| | | |
| | | |
| | | | |
* commit 'b15d3e9d4201362f2617955b89fbd88ecbeb06c4':
Fix extras clipping
|
| |\ \ \ |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Bug: 5556494
Change-Id: I82e8282271dfc623afa9b1ef0d0061ca9c3df5f4
|
|\ \ \ \ \
| |/ / / /
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
memory consumption" into ics-mr1
* commit '00d524fd5dc3e995e2b517da48e479694260ce2f':
Change the default webView behavior to minimize the memory consumption
|
| |\ \ \ \
| | |/ / /
| |/| / /
| | |/ / |
consumption" into ics-mr1
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
For now, we just limit the expansion on x and y direction.
In the future, we need to dynamically allocate and deallocate the memory.
browser change is : c/148242
bug:5522175
Change-Id: I8562c39c55913ab89848afe8a36e8a2784c12727
|
|\ \ \ \
| |/ / /
| | | |
| | | |
| | | | |
* commit '82ccbd8ce95211b4b5afad6a653462fd3beb771b':
Remove the global button lock
|
| | |/
| |/|
| | |
| | |
| | |
| | |
| | | |
Bug: 5558699
Bug: 5572238
Change-Id: Ic199ffda5bc4aa09d39ee54221c6fcce6a91ca05
|
|\ \ \
| |/ /
| | |
| | |
| | | |
* commit '54464c408c893079c1486aec03329f1de8ec6a29':
Doesn't lock the entire drawing...
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
This caused us to potentially waiting on the completion of a tile
painting. On sites where skia struggle with, this had a large impact on
scrolling performances.
This is only part one of the solution -- we need to also get rid of the
global button lock.
We also lower the priority of the texture gneeration thread.
bug:5558699
Change-Id: I251354d6cabb2fd7ebfd665c30eff7fe90c3d316
|