[Webkit-unassigned] [Bug 200165] WebSockets: response Set-Cookie header not handled when using platform APIs

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=200165

--- Comment #11 from Carlos Garcia Campos <cgarcia at igalia.com> ---
Comment on attachment 375338
  --> https://bugs.webkit.org/attachment.cgi?id=375338
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=375338&action=review

>> Source/WebKit/NetworkProcess/NetworkSocketChannel.cpp:109
>> +        m_session->networkStorageSession()->setCookiesFromDOM(m_request.firstPartyForCookies(), SameSiteInfo::create(m_request),
> 
> You should probably check m_session before using it, since it's weak.

I'm doing it, see the condition above:

if (!serverSetCookie.isEmpty() && m_session) {

>> Source/WebKit/NetworkProcess/cocoa/WebSocketTaskCocoa.mm:87
>> +    // FIXME: support extensions and server cookie.
> 
> Actually, you should probably make an accessor for m_session on NetworkSocketChannel then move your cookie setting code from NetworkSocketChannel::didConnect to WebSocketTaskSoup.cpp because the cocoa implementation won't be setting cookies like this.

Ok, isn't it possible to use the same approach to set the cookies?

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190802/4c1e5040/attachment.html>