Bug 112564 – Refactor Use initiatorType instead of initiatorName in CachedResourceRequest

• Component: WebCore Misc.
• Status: RESOLVED CONFIGURATION CHANGED

[pdeng6]

use initiatorType() to return "css" "img" "link" "xmlhttprequest" is better for meaning and keep consistent with its client Resource Timing.

[pdeng6]

Created attachment 193548 [details]
Patch

[WebKit Review Bot]

Please wait for approval from [e-mail], [e-mail], [e-mail], [e-mail] or [e-mail] before submitting, as this patch contains changes to the Chromium public API. See also https://trac.webkit.org/wiki/ChromiumWebKitAPI.

[WebKit Review Bot]

Comment on attachment 193548 [details]
Patch

Attachment 193548 [details] did not pass cr-linux-debug-ews (chromium-xvfb):
Output: http://webkit-commit-queue.appspot.com/results/17242008

[Adam Barth]

Comment on attachment 193548 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=193548&action=review

I'm not sure the benefits of this change are worth the code and API churn.

> Source/WebKit/chromium/public/WebCachedURLRequest.h:56
> -    WEBKIT_EXPORT WebString initiatorName() const;
> +    WEBKIT_EXPORT WebString initiatorType() const;

This API change is likely to break code in Chromium that consumes this API.

[James Simonsen]

(In reply to comment #4)
> (From update of attachment 193548 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=193548&action=review
> 
> I'm not sure the benefits of this change are worth the code and API churn.

Agreed. Also, this field is not just for Resource Timing, so we shouldn't restrict ourselves to matching that. I think Jochen and Marja are using it for something.

[Marja Hölttä]

Yeah, it's potentially used by the declarative web request api (battre@ & vabr@ can tell if it's used for real).

[pdeng6]

thanks for all comments.
cr-linux-debug-ews failed because I missed touch some MACRO.
I've searched the code, and found WebCachedURLRequest::initiatorName() only used in WebTestProxyBase::willRequestResource besides Resource Timing, and I guess the intent there is also return type rather than name, the real usage? :)

Thanks
Pan

[Dominic Battre]

(In reply to comment #6)
> Yeah, it's potentially used by the declarative web request api (battre@ & vabr@ can tell if it's used for real).

Please excuse my late reply. This is not used in the WebRequest API. According to code search only WebKit/Tools/DumpRenderTree/chromium/TestRunner/src/WebTestProxy.cpp calls this function.

[Joseph Pecoraro]

This seems stale and is only a name change. Lets close.