You need to
before you can comment on or make changes to this bug.
Ideally, these security checks would be in some common part of the custom bindings that all descendants of V8Element had access to. Unfortunately, there doesn't seem to be such a place, and V8AttrCustom.cpp needs to have access to these checks as well. Therefore, these should probably go in V8BindingSecurity (which will need to be expanded to something more than a typedef).
Created an attachment (id=45842) [details]
Attachment 45842 [details] did not pass style-queue:
Failed to run "WebKitTools/Scripts/check-webkit-style" exit_code: 1
WebCore/ChangeLog:7: Line contains tab character. [whitespace/tab] 
WebCore/bindings/v8/V8BindingSecurity.h:68: Could not find a newline character at the end of the file. [whitespace/ending_newline] 
Total errors found: 2
Created an attachment (id=45843) [details]
patch minus style errors
style-queue ran check-webkit-style on attachment 45843 [details] without any errors.
(From update of attachment 45843 [details])
Sorry, I think i steered you slightly wrong here. This code can go in the generic BindingSecurity object since it's not dependent on V8:
That way the code will be shared with JSC eventually.
(In reply to comment #5)
> (From update of attachment 45843 [details] [details])
> Sorry, I think i steered you slightly wrong here. This code can go in the
> generic BindingSecurity object since it's not dependent on V8:
> That way the code will be shared with JSC eventually.
That'll require a State<Binding> parameter then, right? Is that ok? I don't have a great idea of how the State is used.
Created an attachment (id=45902) [details]
Functions are now in BindingSecurity.h.
I'm not sure whether these functions belong in BindingSecurity.h or whether we should add BindingSecurity.cpp, so I defaulted to not adding a new file.
style-queue ran check-webkit-style on attachment 45902 [details] without any errors.
(From update of attachment 45902 [details])
This looks great. Unfortunately, these functions need to be in the header because of the template parameter.
I'm about to land a change that moves this file, so we're going to conflict. I can land this if you'd rather not deal with the merge conflict.
Committed r52812: <http://trac.webkit.org/changeset/52812>