[Webkit-unassigned] [Bug 177304] WebDriver: Add support to import and run W3C tests

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=177304

--- Comment #4 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to Brian Burg from comment #3)
> Comment on attachment 321434 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=321434&action=review
> 
> Overall, importing WebDriver tests to a separate directory on a different
> timescale seems like the right thing to do. We don't want WebDriver test
> updates to be blocked on rebaselining updated W3C layout tests.

Exactly!

> However, I really don't want two copies of the same WPT tools in the tree.
> Especially because they import pytest, html5lib, and other dependencies into
> their tree instead of using versioned modules, so there is a lot of 3rd
> party code sitting around. This is an invitation for cruft and problems.

Agree.

> Carlos and I talked in IRC about ways to avoid doing this. Imported layout
> tests probably only use wptserve, so we can try to import just that for
> layout tests and import the full tools to WebDriverTests. Or, we can have a
> shared copy of tools and update the two test suites independently.

I've checked it and layout tests don't need wptrunner and pytest at all, so we can update the importer to import only wptserve and its dependencies from tools directory. This way we won't have two copies of wptrunner, and we'll also reduce a bit the size since there are more things in tools that are unused.

> > Tools/Scripts/run-webdriver-tests:46
> > +option_parser.add_option('--display-server', choices=['xvfb', 'xorg', 'weston', 'wayland'], default='xvfb',
> 
> I think this option won't make any sense on other platforms than GTK?

Right, same way it happens in run-webkit-tests, I don't think we can add options that depends on the value of other values. What we could do is adding groups to include the port specific options when we add support for more ports.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20170922/af139f13/attachment-0001.html>