Bug 97139 – nrwt: convert chromium TestExpectations to the new syntax

• Status: RESOLVED
• Resolution: FIXED
• Component: New Bugs

[Dirk Pranke]

nrwt: convert chromium TestExpectations to the new syntax

[Dirk Pranke]

Created attachment 164790 [details]
Patch

[Ryosuke Niwa]

Comment on attachment 164790 [details]
Patch

rs=me.

[Dirk Pranke]

Committed r129057: <http://trac.webkit.org/changeset/129057>

[Ojan Vafai]

Comment on attachment 164790 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=164790&action=review

> Tools/Scripts/convert-test-expectations:7
> +

For thoroughness sake, maybe put a FIXME to delete this file once all the conversions are done?

> LayoutTests/platform/chromium/TestExpectations:154
> +webkit.org/b/68610 fast/dom/MicroData

Now that I see this in practice, I think it's pretty confusing to not have [ Skip ] indicating that the test is skipped. WDYT?

> LayoutTests/platform/chromium/TestExpectations:157
> +crbug.com/103939 http/tests/appcache/wrong-content-type.html [ Failure Skip ]

Sorry if my memory is foggy here. Didn't we decide to now allow an expectation type if the test is skipped?

> LayoutTests/platform/chromium/TestExpectations:229
> +dom/xhtml/level2/html/HTMLDocument17.xhtml [ Failure Skip WontFix ]

In the end we want this to just be [ WontFix ], right? Will we just need to do another pass over the file once we support WontFix without Skip or other listed expectations?

[Ojan Vafai]

Comment on attachment 164790 [details]
Patch

Whoops. Didn't mean to overwrite the review flag.

[Dirk Pranke]

(In reply to comment #3)
> (From update of attachment 164790 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=164790&action=review
> 
> > Tools/Scripts/convert-test-expectations:7
> > +
> 
> For thoroughness sake, maybe put a FIXME to delete this file once all the conversions are done?
> 

Well, I already landed this, but yeah, I'll add on shortly.

> > LayoutTests/platform/chromium/TestExpectations:154
> > +webkit.org/b/68610 fast/dom/MicroData
> 
> Now that I see this in practice, I think it's pretty confusing to not have [ Skip ] indicating that the test is skipped. WDYT?
>

I'm inclined to agree. 

> > LayoutTests/platform/chromium/TestExpectations:157
> > +crbug.com/103939 http/tests/appcache/wrong-content-type.html [ Failure Skip ]
> 
> Sorry if my memory is foggy here. Didn't we decide to now allow an expectation type if the test is skipped?
>

I also have a very foggy memory; I think you're right but I forgot to implement that. I'll file a bug.

 
> > LayoutTests/platform/chromium/TestExpectations:229
> > +dom/xhtml/level2/html/HTMLDocument17.xhtml [ Failure Skip WontFix ]
> 
> In the end we want this to just be [ WontFix ], right? Will we just need to do another pass over the file once we support WontFix without Skip or other listed expectations?

Right, on both counts. I'll roll that into the Skip bug.

[Dirk Pranke]

Turns out the new chromium TestExpectations file didn't actually get committed, due to a merge conflict. 

I landed it separately in http://trac.webkit.org/changeset/129063 .

[Tim Horton]

Removed the script in http://trac.webkit.org/changeset/189132