Bug 81196 – IETC HTML5: verify HTMLDataListElement - instanceof HTMLDataListElement fails.

• Component: DOM
• Status: RESOLVED FIXED

[Deepak Sherveghar]

IETC HTM5: verify HTMLDatalistElement fails.

http://samples.msdn.microsoft.com/ietestcenter/html5/show_forms_test.htm?datalistelement1

Add HTMLDatalistElement to DOMWindow.idl for consistency.
HTMLDatalistElement Constructor should be available on DOMWindow.

[Deepak Sherveghar]

Created attachment 131999 [details]
patch

patch

[WebKit Review Bot]

Comment on attachment 131999 [details]
patch

Attachment 131999 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/11957499

New failing tests:
fast/forms/datalist_prototype_constructor.html

[Deepak Sherveghar]

Created attachment 132004 [details]
Updated Patch

HTML5 datalist element is not enabled in chromium as its implementation is incomplete.
Hence skipping the test case in chromium.

[Kent Tamura]

Comment on attachment 132004 [details]
Updated Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=132004&action=review

> LayoutTests/ChangeLog:13
> +        * fast/forms/datalist_prototype_constructor-expected.txt: Added.
> +        * fast/forms/datalist_prototype_constructor.html: Added.

We usually use '-' to concatenate words, not '_'.

The test looks a copy of IETC code.  Is the license of the IETC test compatible with WebKit?

You need to update other tests such as LayoutTests/fast/dom/window/window-properties.html.

[Kent Tamura]

Comment on attachment 132004 [details]
Updated Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=132004&action=review

> Source/WebCore/ChangeLog:3
> +        IETC HTM5: verify HTMLDatalistElement. instanceof HTMLDataListElement fails.

HTM5 -> HTML5
HTMLData*l*istElement -> HTMLDataListElement

[Adam Barth]

> The test looks a copy of IETC code.  Is the license of the IETC test compatible with WebKit?

I believe so:

http://samples.msdn.microsoft.com/ietestcenter/support/copyright.htm

[Deepak Sherveghar]

Created attachment 132576 [details]
Updated Patch with review comments.

> HTM5 -> HTML5
> HTMLData*l*istElement -> HTMLDataListElement 
Done.

> You need to update other tests such as LayoutTests/fast/dom/window/window-properties.html.
I added 'HTMLDataListElement' to LayoutTests/fast/dom/Window/resources/window-properties.js.

Following tests needs to be rebaselined:
1. fast/dom/Window/window-properties.html 
2. fast/dom/Window/window-property-descriptors.html.

I have rebaselined it for GTK platform, for rest of the platforms I have added them to test_Expectations.txt/Skipped.

Since the above mentioned tests cover this change, I have removed the Layout test that I had copied earlier from IETC test suite.

[Deepak Sherveghar]

Created attachment 132578 [details]
Updated Patch

[Kent Tamura]

Comment on attachment 132578 [details]
Updated Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=132578&action=review

> LayoutTests/platform/mac/test_expectations.txt:417
> +
> +// Need rebaselining. https://bugs.webkit.org/show_bug.cgi?id=81196
> +BUGWK81196 : fast/dom/Window/window-properties.html = TEXT
> +BUGWK81196 : fast/dom/Window/window-property-descriptors.html = TEXT

You know how the results will change.  I recommend updating *-expected.txt manually rather than skipping the tests.

[Kent Tamura]

Comment on attachment 132578 [details]
Updated Patch

r- because of my comment.  Also, ENABLE_DATALIST was disabled for many ports, and we need no changes of Skipped/test_expectations.txt for such ports.

[Eric Seidel (no email)]

Comment on attachment 132578 [details]
Updated Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=132578&action=review

> Source/WebCore/page/DOMWindow.idl:412
> +        attribute HTMLDataListElementConstructor HTMLDataListElement;

Don't we have a [Conditional=DATALIST] idl option that does this these days?

[Ryosuke Niwa]

Comment on attachment 132578 [details]
Updated Patch

cq- given r-.

[alexander farkas]

This issue is more important than just for consistency. Due to the fact, that Safari still reports true on ('list' in document.createElement('input')) and gives a false positive for feature detection. Testing for the global HTMLDataListElement has become the famoust feature detection to distinguish between input[list]/datalist supporting browsers and none supporting browsers. This is also part of Modernizr.

So it is cruical to implement this property as soon as a basic datalist including UI is implemented.

@tkent
Currently Chrome "20.0.1113.0 canary" returns a false negative for datalist. Should I fill an issue in the Chrome bugtracker?

[Kent Tamura]

(In reply to comment #13)
> Currently Chrome "20.0.1113.0 canary" returns a false negative for datalist. Should I fill an issue in the Chrome bugtracker?

No, you don't need to file another issue.

[Kent Tamura]

If Deepak won't update the patch in a few days, someone in my team will take this over.

[Deepak Sherveghar]

Hi Kent,

I am tied up with other things, wont be able to pick this up. Feel free to assign this to your team.

[Keishi Hattori]

Created attachment 139144 [details]
Patch

[Kent Tamura]

Comment on attachment 139144 [details]
Patch

ok

[WebKit Review Bot]

Comment on attachment 139144 [details]
Patch

Clearing flags on attachment: 139144

Committed r115446: <http://trac.webkit.org/changeset/115446>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.