commit | 39563cf141605683bb8d76abb74eaa0c764d6c65 | [log] [tgz] |
---|---|---|
author | Morten Stenshorne <mstensho@chromium.org> | Wed May 02 05:42:23 2018 |
committer | Blink WPT Bot <blink-w3c-test-autoroller@chromium.org> | Wed May 02 06:00:54 2018 |
tree | 93500004de69b0e8d1b5fd8d8b56d8b528378b8f | |
parent | c83c53e0cc5ec8a7427c8039e772a3bbf1bb9f50 [diff] |
[LayoutNG] Improve handling of adjoining floats. A float always needs to be positioned by its block parent, so passing them around to children, parents or siblings really shouldn't be necessary (and it *was* somewhat confusing, since nobody but the direct float parent is allowed to touch them anyway, apart from placing them into temporary exclusion spaces). The main reason for passing them around like that, was for other blocks to determine, based on the list of floats being empty or not, the need for relayout once the BFC offset was resolved. Instead, confine the list of unpositioned floats to the block parent of those floats, and introduce the concept of adjoining float types (none, left, right, both). Adjoining floats occur when the BFC offset is unknown, meaning that their position may be affected by the current layout algorithm. Adjoining float types will now be the thing that's both input to and output from the layout algorithms. Having something other than "none" means that a block's BFC offset is unknown, but that doesn't automatically mean that we have to abort and re-layout if the BFC offset gets resolved. If the "floats BFC offset" is known, for instance, those adjoining floats may be positioned right away. Still we need to know about them (positioned or not), to get clearance correct. We're going to need to treat adjoining floats specially when applying clearance. Will deal with that in a later CL. For now, we just keep track of the adjoining float types, so that the clearance machinery can tell that there are floats there that may not yet be positioned. That used to be taken care of the list of unpositioned floats, but, as previously stated, adjoining floats are special, and we need to know about them, whether they are positioned or not. This is a preparatory CL for that. Each time we add an unpositioned float, we need to update the types of adjoining floats, so that these can be returned from the algorithm if necessary. Whether they end up being positioned right away or not isn't relevant. Adjoining is adjoining. Note that we don't have to #include the header file for unpositioned floats as much as before now, but I'll clean that up in a follow-up CL, because it turned out that there were quite a few translation units that got stuff for free via that header file, instead of explicitly including what they need. Had to rewrite how we deal with floats in HandleNewFormattingContext() and LayoutNewFormattingContext(), since those depended on a list of all preceding unpositioned floats to place them into a temporary exclusion space, to figure out whether to let the child's margin be adjoining with the current margin strut or not. Instead of using a temporary exclusion space, we now position floats via the regular mechanisms, and initially assume that the child's margin is going to be adjoining. This means that we have to abort and roll back if there are preceding unpositioned floats. This is no different from how we lay out regular blocks, though. What *is* different is that if it turns out that the child's margin has to be separated from the strut, we'll have to abort and roll back once *again* (but only once). The algorithms now need to keep track of whether they need to abort if the BFC offset *changes*, rather than if it is *resolved*. We only allow the offset to resolve and optionally change *once* afterwards, though. Cq-Include-Trybots: master.tryserver.chromium.linux:linux_layout_tests_layout_ng Change-Id: Ie527d659213049f180ebedc764e1d7f4926a5876 Reviewed-on: https://chromium-review.googlesource.com/1030191 Commit-Queue: Morten Stenshorne <mstensho@chromium.org> Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org> Reviewed-by: Koji Ishii <kojii@chromium.org> Cr-Commit-Position: refs/heads/master@{#555314}
The web-platform-tests Project is a W3C-coordinated attempt to build a cross-browser testsuite for the Web-platform stack. Writing tests in a way that allows them to be run in all browsers gives browser projects confidence that they are shipping software that is compatible with other implementations, and that later implementations will be compatible with their implementations. This in turn gives Web authors/developers confidence that they can actually rely on the Web platform to deliver on the promise of working across browsers and devices without needing extra layers of abstraction to paper over the gaps left by specification editors and implementors.
Clone or otherwise get https://github.com/w3c/web-platform-tests.
Note: because of the frequent creation and deletion of branches in this repo, it is recommended to “prune” stale branches when fetching updates, i.e. use git pull --prune
(or git fetch -p && git merge
).
The tests are designed to be run from your local computer. The test environment requires Python 2.7+ (but not Python 3.x).
On Windows, be sure to add the Python directory (c:\python2x
, by default) to your %Path%
Environment Variable, and read the Windows Notes section below.
To get the tests running, you need to set up the test domains in your hosts
file.
The necessary content can be generated with ./wpt make-hosts-file
; on Windows, you will need to preceed the prior command with python
or the path to the Python binary (python wpt make-hosts-file
).
For example, on most UNIX-like systems, you can setup the hosts file with:
./wpt make-hosts-file | sudo tee -a /etc/hosts
And on Windows (note this requires an Administrator privileged shell):
python wpt make-hosts-file >> %SystemRoot%\System32\drivers\etc\hosts
If you are behind a proxy, you also need to make sure the domains above are excluded from your proxy lookups.
The test server can be started using
./wpt serve
On Windows: You will need to preceed the prior command with python
or the path to the python binary.
python wpt serve
This will start HTTP servers on two ports and a websockets server on one port. By default the web servers start on ports 8000 and 8443 and the other ports are randomly-chosen free ports. Tests must be loaded from the first HTTP server in the output. To change the ports, copy the config.default.json
file to config.json
and edit the new file, replacing the part that reads:
"http": [8000, "auto"], "https":[8443]
to some ports of your choice e.g.
"http": [1234, "auto"], "https":[5678]
After your hosts
file is configured, the servers will be locally accessible at:
http://web-platform.test:8000/
https://web-platform.test:8443/ *
*See Trusting Root CA
Tests can be run automatically in a browser using the run
command of the wpt
script in the root of the checkout. This requires the hosts file setup documented above, but you must not have the test server already running when calling wpt run
. The basic command line syntax is:
./wpt run product [tests]
On Windows: You will need to preceed the prior command with python
or the path to the python binary.
python wpt run product [tests]
where product
is currently firefox
or chrome
and [tests]
is a list of paths to tests. This will attempt to automatically locate a browser instance and install required dependencies. The command is very configurable; for example to specify a particular binary use wpt run --binary=path product
. The full range of options can be see with wpt run --help
and wpt run --wptrunner-help
.
Not all dependencies can be automatically installed; in particular the certutil
tool required to run https tests with Firefox must be installed using a system package manager or similar.
On Debian/Ubuntu certutil may be installed using:
sudo apt install libnss3-tools
And on macOS with homebrew using:
brew install nss
On other platforms, download the firefox archive and common.tests.zip archive for your platform from Mozilla CI.
Then extract certutil[.exe]
from the tests.zip package and libnss3[.so|.dll|.dynlib]
and put the former on your path and the latter on your library path.
The wpt
command provides a frontend to a variety of tools for working with and running web-platform-tests. Some of the most useful commands are:
wpt serve
- For starting the wpt http serverwpt run
- For running tests in a browserwpt lint
- For running the lint against all testswpt manifest
- For updating or generating a MANIFEST.json
test manifestwpt install
- For installing the latest release of a browser or webdriver server on the local machine.Some optional components of web-platform-tests (test components from third party software and pieces of the CSS build system) are included as submodules. To obtain these components run the following in the root of your checkout:
git submodule update --init --recursive
Prior to commit 39d07eb01fab607ab1ffd092051cded1bdd64d78
submodules were requried for basic functionality. If you are working with an older checkout, the above command is required in all cases.
When moving between a commit prior to 39d07eb
and one after it git may complain
$ git checkout master error: The following untracked working tree files would be overwritten by checkout: […]
followed by a long list of files. To avoid this error remove the resources
and tools
directories before switching branches:
$ rm -r resources/ tools/ $ git checkout master Switched to branch 'master' Your branch is up-to-date with 'origin/master'
When moving in the opposite direction, i.e. to a commit that does have submodules, you will need to git submodule update
, as above. If git throws an error like:
fatal: No url found for submodule path 'resources/webidl2/test/widlproc' in .gitmodules Failed to recurse into submodule path 'resources/webidl2' fatal: No url found for submodule path 'tools/html5lib' in .gitmodules Failed to recurse into submodule path 'resources' Failed to recurse into submodule path 'tools'
then remove the tools
and resources
directories, as above.
On Windows wpt
commands must be prefixed with python
or the path to the python binary (if python
is not in your %PATH%
).
python wpt [command]
Alternatively, you may also use Bash on Ubuntu on Windows in the Windows 10 Anniversary Update build, then access your windows partition from there to launch wpt
commands.
Please make sure git and your text editor do not automatically convert line endings, as it will cause lint errors. For git, please set git config core.autocrlf false
in your working tree.
By default pregenerated certificates for the web-platform.test domain are provided in tools/certs
. If you wish to generate new certificates for any reason it's possible to use OpenSSL when starting the server, or starting a test run, by providing the --ssl-type=openssl
argument to the wpt serve
or wpt run
commands.
If you installed OpenSSL in such a way that running openssl
at a command line doesn't work, you also need to adjust the path to the OpenSSL binary. This can be done by adding a section to config.json
like:
"ssl": {"openssl": {"binary": "/path/to/openssl"}}
On Windows using OpenSSL typically requires installing an OpenSSL distribution. Shining Light provide a convenient installer that is known to work, but requires a little extra setup, i.e.:
Run the installer for Win32_OpenSSL_v1.1.0b (30MB). During installation, change the default location for where to Copy OpenSSL Dlls from the System directory to the /bin directory.
After installation, ensure that the path to OpenSSL (typically, this will be C:\OpenSSL-Win32\bin
) is in your %Path%
Environment Variable. If you forget to do this part, you will most likely see a ‘File Not Found’ error when you start wptserve.
Finally, set the path value in the server configuration file to the default OpenSSL configuration file location. To do this, copy config.default.json
in the web-platform-tests root to config.json
. Then edit the JSON so that the key ssl/openssl/base_conf_path
has a value that is the path to the OpenSSL config file (typically this will be C:\\OpenSSL-Win32\\bin\\openssl.cfg
).
To prevent browser SSL warnings when running HTTPS tests locally, the web-platform-tests Root CA file cacert.pem
in tools/certs must be added as a trusted certificate in your OS/browser.
The master branch is automatically synced to http://w3c-test.org/.
Pull requests are automatically mirrored except those that modify sensitive resources (such as .py
). The latter require someone with merge access to comment with “LGTM” or “w3c-test:mirror” to indicate the pull request has been checked.
Each top-level directory matches the shortname used by a standard, with some exceptions. (Typically the shortname is from the standard's corresponding GitHub repository.)
For some of the specifications, the tree under the top-level directory represents the sections of the respective documents, using the section IDs for directory names, with a maximum of three levels deep.
So if you're looking for tests in HTML for “The History interface”, they will be under html/browsers/history/the-history-interface/
.
Various resources that tests depend on are in common
, images
, and fonts
.
In the vast majority of cases the only upstream branch that you should need to care about is master
. If you see other branches in the repository, you can generally safely ignore them.
Save the Web, Write Some Tests!
Absolutely everyone is welcome (and even encouraged) to contribute to test development, so long as you fulfill the contribution requirements detailed in the Contributing Guidelines. No test is too small or too simple, especially if it corresponds to something for which you've noted an interoperability bug in a browser.
The way to contribute is just as usual:
git checkout -b topic
.If you spot an issue with a test and are not comfortable providing a pull request per above to fix it, please file a new issue. Thank you!
We have a lint tool for catching common mistakes in test files. You can run it manually by starting the lint
executable from the root of your local web-platform-tests working directory like this:
./wpt lint
The lint tool is also run automatically for every submitted pull request, and reviewers will not merge branches with tests that have lint errors, so you must fix any errors the lint tool reports.
In the unusual case of error reports for things essential to a certain test or that for other exceptional reasons shouldn't prevent a merge of a test, update and commit the lint.whitelist
file in the web-platform-tests root directory to suppress the error reports.
For more details, see the lint-tool documentation.
Sometimes you may want to add a script to the repository that‘s meant to be used from the command line, not from a browser (e.g., a script for generating test files). If you want to ensure (e.g., for security reasons) that such scripts won’t be handled by the HTTP server, but will instead only be usable from the command line, then place them in either:
the tools
subdir at the root of the repository, or
the tools
subdir at the root of any top-level directory in the repository which contains the tests the script is meant to be used with
Any files in those tools
directories won't be handled by the HTTP server; instead the server will return a 404 if a user navigates to the URL for a file within them.
If you want to add a script for use with a particular set of tests but there isn't yet any tools
subdir at the root of a top-level directory in the repository containing those tests, you can create a tools
subdir at the root of that top-level directory and place your scripts there.
For example, if you wanted to add a script for use with tests in the notifications
directory, create the notifications/tools
subdir and put your script there.
We can sometimes take a little while to go through pull requests because we have to go through all the tests and ensure that they match the specification correctly. But we look at all of them, and take everything that we can.
OWNERS files are used only to indicate who should be notified of pull requests. If you are interested in receiving notifications of proposed changes to tests in a given directory, feel free to add yourself to the OWNERS file. Anyone with expertise in the specification under test can approve a pull request. In particular, if a test change has already been adequately reviewed “upstream” in another repository, it can be pushed here without any further review by supplying a link to the upstream review.
Search filters to find things to review:
infra
(excluding vendor exports)docs
(excluding vendor exports)If you wish to contribute actively, you're very welcome to join the public-test-infra@w3.org mailing list (low traffic) by signing up to our mailing list. The mailing list is archived.
Join us on irc #testing (irc.w3.org, port 6665). The channel is archived.