commit | 2ab9dfaee5c791e0d7b2c0893a55ced8677e2e2a | [log] [tgz] |
---|---|---|
author | wenson_hsieh@apple.com <wenson_hsieh@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc> | Sun Jul 14 03:36:19 2019 +0000 |
committer | wenson_hsieh@apple.com <wenson_hsieh@apple.com@268f45cc-cd09-0410-ab3c-d52691b4dbfc> | Sun Jul 14 03:36:19 2019 +0000 |
tree | 6936c86a80ffd1dc85c58199703671c1faf7dbeb | |
parent | 615a3e0bc7e49bb529533f6adc6ffb4ce43c6e9b [diff] |
[Text autosizing] [iPadOS] Further adjust our heuristics to determine text autosizing candidates https://bugs.webkit.org/show_bug.cgi?id=199780 <rdar://problem/52289088> Reviewed by Simon Fraser. Source/WebCore: Our current idempotent text autosizing candidate heuristic makes the right judgment call most of the time, but there is still a large batch of text autosizing bugs left unfixed by the first iteration of the heuristic added in r246781. This patch attempts to address most of these bugs by adjusting the decision-tree-based heuristic once again, mostly with improvements to the model generation pipeline. During the first iteration, I placed emphasis on tuning the max tree depth and min leaf size hyperparameters when coming up with my decision tree, and didn't consider the inclusion or exclusion of each feature as a hyperparameters. As such, the trees generated using the pipeline tended to use too many features, and as a result, tended to have cross-validation overall accuracy scores hovering around 73%. In this revised model generation pipeline, I now consider the inclusion of each feature (along with max depth and min leaf size, as before) as a hyperparameter. Since this increases the number of hyperparameters by many orders of magnitude, a naive grid search (as described in the prior ChangeLog entry) is no longer a tractible procedure for tuning hyperparameters to the training algorithm. Instead, I now use a stochastic greedy algorithm to search for good sets of hyperparameters; this process begins with seeding some number (usually 20-24) of "searchers" with completely randomized sets of hyperparameters (i.e. random max depth, random leaf size, and random subsets of features). I then evaluate the average performance of each set of hyperparameters by using them to generate 2000 decision trees over 90% of the training data, and then cross-validating these trees against the remaining 10%. These cross-validation scores are aggregated into a single confusion matrix, which is then passed into a loss function that computes a single value indicating how well training with the set of hyperparameters generalized to cross-validation data. After experimenting with various loss functions, I settled on the following: `k(false positive rate)^2 + (false negative rate)^2` ...where a constant k is chosen to penalize false positives (i.e. broken layout) more harshly than false negatives (small text). Additionally, squaring the false negative and false positive rates seems to help avoid converging on solutions that heavily favor reducing only false positives or false negatives, or vice versa. The stochastic algorithm starts by computing a loss value for the randomly generated configuration. Then, for an indefinite number of iterations, it randomly mutates the configuration (e.g. by adding or removing features, or changing min leaf size or max tree depth) and computes a new loss value for the mutated configuration. If the mutated configuration performs better (i.e. achieves lower loss) than the current configuration, I set the current configuration to be the mutated configuration. Otherwise, I keep the current (non-mutated) configuration as-is. The stochastic algorithm then proceeds, ad-infinitum, with this current configuration. Of course, since each mutation is small, this strategy so far is prone to leaving each searcher stuck in local optima. To mitigate this, for each searcher, I keep track of a side-table of configurations that have already been tested; when random mutations would normally lead to testing a configuration that has already been tested, each searcher instead increases the chance of applying additional mutations. This has the effect of searchers initially exhausting similar configurations, and expanding to test more and more dissimilar configurations as the local alternatives all turn out to be worse. This allows searchers to effectively jump out of local optima after being stuck for a long time. So, using these strategies, I simultaneously ran a handful of searchers until they all appeared to converge (a process that takes 8-12 hours on my current dataset). Many of the searchers achieved configurations with cross-validation scores of 81% and above, up from the 73% of the previous attempt. These additionally have the added bonus of reducing the number of features, often making the final trees themselves shallower and simpler to understand than before. This patch introduces one such decision tree generated using a set of hyperparameters acquired via this stochasic search algorithm; it appears to simultaneously use fewer features, and achieve better cross-validation performance. Test: fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-candidates.html * css/StyleResolver.cpp: (WebCore::StyleResolver::adjustRenderStyleForTextAutosizing): Adjust the early return to bail if either (1) the element is a candidate and the computed size is already equal to the boosted size, or (2) the element is not a candidate and the computed size is already equal to the specified size. Since the autosizing candidate heuristic depends on styles specified on the element itself (as opposed to styles on any element in the ancestor chain), a parent may be an autosizing candidate, but a child of it may not. * rendering/style/RenderStyle.cpp: (WebCore::RenderStyle::isIdempotentTextAutosizingCandidate const): Revamp the idempotent text autosizing candidate heuristic. See the explanation above for more details. * rendering/style/RenderStyle.h: Remove some bits from RenderStyle's autosizeStatus, now that we care about fewer bits of information from the inherited flags. * rendering/style/TextSizeAdjustment.cpp: (WebCore::AutosizeStatus::updateStatus): * rendering/style/TextSizeAdjustment.h: LayoutTests: Rebaseline an existing idempotent text autosizing test, and add an additional test case. * fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-candidates-expected.txt: * fast/text-autosizing/ios/idempotentmode/idempotent-autosizing-candidates.html: git-svn-id: http://svn.webkit.org/repository/webkit/trunk@247421 268f45cc-cd09-0410-ab3c-d52691b4dbfc
WebKit is a cross-platform web browser engine. On iOS and macOS, it powers Safari, Mail, iBooks, and many other applications.
Visit WebKit Feature Status page to see which Web API has been implemented, in development, or under consideration.
On macOS, download Safari Technology Preview to test the latest version of WebKit. On Linux, download Epiphany Technology Preview. On Windows, you'll have to build it yourself.
Once your bug is filed, you will receive email when it is updated at each stage in the bug life cycle. After the bug is considered fixed, you may be asked to download the latest nightly and confirm that the fix works for you.
On Windows, follow the instructions on our website.
Run the following command to clone WebKit's Git SVN repository:
git clone git://git.webkit.org/WebKit.git WebKit
or
git clone https://git.webkit.org/git/WebKit.git WebKit
If you want to be able to commit changes to the repository, or just want to check out branches that aren’t contained in WebKit.git, you will need track WebKit's Subversion repository. You can run the following command to configure this and other options of the new Git clone for WebKit development.
Tools/Scripts/webkit-patch setup-git-clone
For information about this, and other aspects of using Git with WebKit, read the wiki page.
If you don‘t want to use Git, run the following command to check out WebKit’s Subversion repository:
svn checkout https://svn.webkit.org/repository/webkit/trunk WebKit
Install Xcode and its command line tools if you haven't done so already:
xcode-select --install
Run the following command to build a debug build with debugging symbols and assertions:
Tools/Scripts/build-webkit --debug
For performance testing, and other purposes, use --release
instead.
You can open WebKit.xcworkspace
to build and debug WebKit within Xcode.
If you don't use a custom build location in Xcode preferences, you have to update the workspace settings to use WebKitBuild
directory. In menu bar, choose File > Workspace Settings, then click the Advanced button, select “Custom”, “Relative to Workspace”, and enter WebKitBuild
for both Products and Intermediates.
The first time after you install a new Xcode, you will need to run the following command to enable Xcode to build command line tools for iOS Simulator:
sudo Tools/Scripts/configure-xcode-for-ios-development
Without this step, you will see the error message: “target specifies product type ‘com.apple.product-type.tool’, but there’s no such product type for the ‘iphonesimulator’ platform.
” when building target JSCLLIntOffsetsExtractor
of project JavaScriptCore
.
Run the following command to build a debug build with debugging symbols and assertions for iOS:
Tools/Scripts/build-webkit --debug --ios-simulator
For production builds:
cmake -DPORT=GTK -DCMAKE_BUILD_TYPE=RelWithDebInfo -GNinja ninja sudo ninja install
For development builds:
Tools/gtk/install-dependencies Tools/Scripts/update-webkitgtk-libs Tools/Scripts/build-webkit --gtk --debug
For more information on building WebKitGTK+, see the wiki page.
For production builds:
cmake -DPORT=WPE -DCMAKE_BUILD_TYPE=RelWithDebInfo -GNinja ninja sudo ninja install
For development builds:
Tools/wpe/install-dependencies Tools/Scripts/update-webkitwpe-libs Tools/Scripts/build-webkit --wpe --debug
For building WebKit on Windows, see the wiki page.
Run the following command to launch Safari with your local build of WebKit:
Tools/Scripts/run-safari --debug
The run-safari
script sets the DYLD_FRAMEWORK_PATH
environment variable to point to your build products, and then launches /Applications/Safari.app
. DYLD_FRAMEWORK_PATH
tells the system loader to prefer your build products over the frameworks installed in /System/Library/Frameworks
.
To run other applications with your local build of WebKit, run the following command:
Tools/Scripts/run-webkit-app <application-path>
Run the following command to launch iOS simulator with your local build of WebKit:
run-safari --debug --ios-simulator
In both cases, if you have built release builds instead, use --release
instead of --debug
.
If you have a development build, you can use the run-minibrowser script, e.g.:
run-minibrowser --debug --wpe
Pass one of --gtk
, --jsc-only
, or --wpe
to indicate the port to use.
Congratulations! You’re up and running. Now you can begin coding in WebKit and contribute your fixes and new features to the project. For details on submitting your code to the project, read Contributing Code.