Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upAdd creator URL, creator base URL and creator origin to browsing context #26499
Conversation
highfive
commented
May 12, 2020
|
Heads up! This PR modifies the following files:
|
|
@bors-servo try=wpt |
|
|
|
|
It seems we have a couple of unexpected test fixes and failures. After sorting them out:
|
|
Also, after going through the follow hyperlink implementation in servo/components/script/dom/htmlanchorelement.rs Lines 640 to 643 in 0fc2155 According to the Parse a URL algorithm, we should be parsing the href attribute value against the document's base URL. I am not sure though if that'll cause any unexpected test results as our implementation for document's |
|
Looking good! This is a nice correctness improvement. |
1ccb849
to
08c1d4b
|
@bors-servo try=wpt |
|
|
|
Some more unexpected test failures. Looking at the changes, I think adding a check for srcdoc in the document's fallback url implementation might be causing these additional test failures afaict. Will have to see what each failure entails. Let me know what you think @jdm. |
|
Yes, I was expecting that outcome :) |
|
08c1d4b
to
9b0b03a
|
A bunch of the new failures that I looked at like https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/semantics/embedded-content/the-img-element/move-element-and-scroll.html make sense since they rely on features we don't support yet. Looks good! |
|
@bors-servo r+ |
|
|
|
@jdm, I looked at the test failures but I am not sure what actually caused them in:
Also, I noticed there are many open issues that partly concern the test failures, but I am also not sure if we need to create any new issues to track any of these? |
|
sandboxed-iframe.html and dangling-markup-mitigation.html started resolving URLs correctly inside the srcdoc iframe for the first time, so the test failures represent tests that didn't run at all previously. I believe the failure in sizes-dynamic-001.html is caused again by the URLs now resolving correctly, and I think we don't implement intrinsic sizes for responsive images so the test now loads images that don't match the reference file. |
|
I see. Let me know if we require any other changes. |
|
|
utsavoza commentedMay 12, 2020
./mach build -ddoes not report any errors./mach test-tidydoes not report any errors