Bug 84960 – [GTK] Test WebKit2.RestoreSessionStateContainingFormData fails

• Status: RESOLVED
• Resolution: FIXED
• Component: Tools / Tests

[Carlos Garcia Campos]

Session State is not implemented in GTK+ port

[Carlos Garcia Campos]

The test crashes because it uses WKSessionStateCopyData() that requires an implementation of WebKit::encodeLegacySessionState(). The default implementation returns NULL.

[Sam P.]

How do you actually run this test?  I have worked on an implementation of this ap and tested it with minibrowser, however, I can not find documentation on how to run these type of tests.

[Michael Catanzaro]

Hi, thanks for working on this. Use the script Tools/Scripts/run-gtk-tests to run the tests. You will need to remove this test from the list of skipped tests up towards the top of that file.

That will run all the tests, though; I can't remember if I ever found a way to run a single cross-platform test or not.

[Sam P.]

Created attachment 261985 [details]
Patch

[Michael Catanzaro]

View in context: https://bugs.webkit.org/attachment.cgi?id=261985&action=review

> Source/WebKit2/ChangeLog:8
> +        This commit adds an implimentation of the LegacySessionStateCoding.  It is based off of

implementation

> Source/WebKit2/ChangeLog:9
> +        the mac implimentation, however instead of using core foundation dictionaries it uses

Ditto

> Source/WebKit2/PlatformGTK.cmake:367
> +    UIProcess/LegacySessionStateCodingNone.cpp

Instead of doing this, let's remove it from CMakeLists.txt and add it manually to all the other Platform*.cmake files.

[Carlos Garcia Campos]

*** Bug 88455 has been marked as a duplicate of this bug. ***

[Carlos Garcia Campos]

Comment on attachment 261985 [details]
Patch

I'm sorry if I didn't explain it correctly, I didn't mean we needed a new implementation of the legacy decoders. Those are there only for compatibility with Safari, but still used by the unit tests. We already have a cross-platform implementation of the SessionState, so we don't need to copy the mac legacy encoders/decoders, but simply use the new implementation.

[Carlos Garcia Campos]

Created attachment 262753 [details]
Patch

This is what I meant

[Sam P.]

That looks great!

[Zan Dobersek]

Comment on attachment 262753 [details]
Patch

r=me

[Carlos Garcia Campos]

Committed r191103: <http://trac.webkit.org/changeset/191103>