Bug 171748 – [Cocoa] Converting from WebCore::Cookie to NSHTTPCookie always marks cookies as session cookies

• Component: WebKit2
• Status: RESOLVED FIXED

[Blaze Burg]

I'm going to temporarily make the assertion wrong so this can be investigated separately from the "secure" flag bug (https://bugs.webkit.org/show_bug.cgi?id=171700). The test fix for that exposed this unexpected failing test assertion.

[Blaze Burg]

Committed r216272: <http://trac.webkit.org/changeset/216272>

[Blaze Burg]

Un-closing, going to track the fix here too.

[Blaze Burg]

Looks like another round tripping issue:

(lldb) po cookie
<NSHTTPCookie version:1 name:"OtherCookieName" value:"OtherCookieValue" expiresDate:(null) created:2017-05-05 21:18:10 +0000 sessionOnly:TRUE domain:".www.w3c.org" partition:"none" path:"/path" isSecure:TRUE>

(lldb) po cookie2.get()
<NSHTTPCookie version:0 name:"OtherCookieName" value:"OtherCookieValue" expiresDate:2017-05-06 00:04:50 +0000 created:2017-05-05 21:18:10 +0000 sessionOnly:FALSE domain:".www.w3c.org" partition:"none" path:"/path" isSecure:FALSE>

Also note that the expires date is getting dropped, afaict. Is that something we are supposed to preserve?

[Michael Catanzaro]

I think it's sessionOnly if it doesn't have an expiresDate... then the expiration date is when you close the browser.

[Blaze Burg]

The NS equivalent here is NSHTTPCookieDiscard:

https://developer.apple.com/reference/foundation/nshttpcookiediscard

"String value must be either "TRUE" or "FALSE". This cookie attribute is optional. The default is "FALSE", unless this cookie is version 1 or greater and a value for NSHTTPCookieMaximumAge is not specified, in which case it is assumed to be "TRUE"."

[Brady Eidson]

(In reply to Brian Burg from comment #5)
> The NS equivalent here is NSHTTPCookieDiscard:
> 
> https://developer.apple.com/reference/foundation/nshttpcookiediscard
> 
> "String value must be either "TRUE" or "FALSE". This cookie attribute is
> optional. The default is "FALSE", unless this cookie is version 1 or greater
> and a value for NSHTTPCookieMaximumAge is not specified, in which case it is
> assumed to be "TRUE"."

Ignore the discard thing, I think.

[Brady Eidson]

(In reply to Michael Catanzaro from comment #4)
> I think it's sessionOnly if it doesn't have an expiresDate... then the
> expiration date is when you close the browser.

Michael is right.

If it has an expires date, it's a persistent cookies.
If it doesn't, it's session.

[Brady Eidson]

Yah, the test does the discard thing, but in reality having an expires date is what defines session or not.

[Brady Eidson]

Except the docs say:
"A boolean value that indicates whether the receiver should be discarded at the end of the session (regardless of expiration date)."
lol wut.

[Radar WebKit Bug Importer]

<rdar://problem/32027327>

[Blaze Burg]

Created attachment 309232 [details]
Patch

[WebKit Commit Bot]

Comment on attachment 309232 [details]
Patch

Clearing flags on attachment: 309232

Committed r216292: <http://trac.webkit.org/changeset/216292>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.