Bug 186105 – Fix the ENABLE(DATALIST_ELEMENT) build

• Status: RESOLVED
• Resolution: FIXED
• Component: Forms

[Aditya Keerthi]

If the DATALIST_ELEMENT flag is active, WebCore, WebKit and WebKitLegacy all fail to build.

[Aditya Keerthi]

Created attachment 341589 [details]
Patch

[Wenson Hsieh]

Comment on attachment 341589 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=341589&action=review

> Source/WebKit/WebProcess/Automation/WebAutomationSessionProxy.cpp:52
> +#include <WebCore/HTMLDataListElement.h>

Super minor nit — we generally prefer to put conditional header includes like this after the last unconditional header include (in this case, after `#include <wtf/UUID.h>`).

[Aditya Keerthi]

Created attachment 341593 [details]
Patch

[WebKit Commit Bot]

Comment on attachment 341593 [details]
Patch

Clearing flags on attachment: 341593

Committed r232306: <https://trac.webkit.org/changeset/232306>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Radar WebKit Bug Importer]

<rdar://problem/40661409>

[Darin Adler]

Comment on attachment 341593 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=341593&action=review

> Source/WebCore/rendering/RenderTheme.cpp:-1003
> -    auto* dataList = downcast<HTMLDataListElement>(input.list());

I don’t understand the rationale for the changes in this file. None of them seem to be needed.

[Aditya Keerthi]

(In reply to Darin Adler from comment #7)
> Comment on attachment 341593 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=341593&action=review
> 
> > Source/WebCore/rendering/RenderTheme.cpp:-1003
> > -    auto* dataList = downcast<HTMLDataListElement>(input.list());
> 
> I don’t understand the rationale for the changes in this file. None of them
> seem to be needed.

When attempting to build before the changes, I was met with a "No matching function for call to 'downcast'".

I now realize that I could have changed that specific line to use input.list().get() and kept the rest the same.

Is there a way that you prefer?

[Darin Adler]

(In reply to Aditya Keerthi from comment #8)
> When attempting to build before the changes, I was met with a "No matching
> function for call to 'downcast'".

OK, no strong preference either way.