<html>
<head>
<base href="https://bugs.webkit.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - Pending API request URL no set when loading Alternate HTML or plain text"
href="https://bugs.webkit.org/show_bug.cgi?id=136916#c14">Comment # 14</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - Pending API request URL no set when loading Alternate HTML or plain text"
href="https://bugs.webkit.org/show_bug.cgi?id=136916">bug 136916</a>
from <span class="vcard"><a class="email" href="mailto:cgarcia@igalia.com" title="Carlos Garcia Campos <cgarcia@igalia.com>"> <span class="fn">Carlos Garcia Campos</span></a>
</span></b>
<pre>Comment on <span class=""><a href="attachment.cgi?id=243974&action=diff" name="attach_243974" title="Now also including the xcode file changes">attachment 243974</a> <a href="attachment.cgi?id=243974&action=edit" title="Now also including the xcode file changes">[details]</a></span>
Now also including the xcode file changes
View in context: <a href="https://bugs.webkit.org/attachment.cgi?id=243974&action=review">https://bugs.webkit.org/attachment.cgi?id=243974&action=review</a>
<span class="quote">>>> Source/WebKit2/UIProcess/WebPageProxy.cpp:860
>>> + m_pageLoadState.setPendingAPIRequestURL(transaction, ASCIILiteral("about:blank"));
>>
>> I think it’s a little peculiar to actually set the URL to about::blank here. Do we agree that explicitly giving it this particular URL makes sense? Does WebKit already treat this as if it came from that URL in other ways or is that a new concept we are introducing? If WebKit does already use this URL then why are we having to inject it specifically here?
>>
>> I think this would read better without the local variable even though the other function does have it in a local variable since it uses it twice. I think blankURL().string() is better than ASCIILiteral("about:blank").
>>
>> m_pageLoadState.setPendingAPIRequestURL(m_pageLoadState.transaction(), blankURL().string());
>
> I found comments in FrameLoader::reload that say, "If a window is created by javascript, its main frame can have an empty but non-nil URL."
>
> Maybe that’s how the plain text string should work. Have an empty URL, not "about:blank".</span >
This is what we are doing in other load methods, see WebPageProxy::loadHTMLString() and WebPageProxy::loadData(). You can see in the unit test I'm adding that it also checks existing cases like WKPageLoadData that indeed returns about:blank. So, I don't know if it's the best to do it, but it's consistent with what we do already.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>