[Webkit-unassigned] [Bug 87826] We should use NaN or separate boolean flag for null Date value

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=87826





--- Comment #3 from Kinuko Yasuda <kinuko at chromium.org>  2012-05-30 01:15:50 PST ---
(In reply to comment #2)
> (In reply to comment #1)
> > (In reply to comment #0)
> > > Currently we use the value 0 (in double) to indicate uninitialized or invalid Date value at several places, but this could be a valid date value as 0 epoch time (i.e. Jan 1 1970) and the binding code just returns Date(0) unless we use custom binding, while null should be returned instead.
> > > 
> > > We should probably use NaN instead 0 (which is automatically converted to Date null) to indicate uninitialized Date value, or should have some binding machinery support to return null.
> > 
> > Our binding code already supports NaN -> null Date mapping.
> 
> Ah, you mean 0->null mapping?
> I think we can use ImplementedAs to avoid custom binding code.
> 
> readonly attribute [ImplementedAs=lastModifiedDateForBinding] Date lastModifiedDate;
> 
> double lastModifiedDateForBinding()
> {
>     double value = lastModifiedDate();
>     return !value ? std::numeric_limits:quiet_nan() : value;
> }

Phew looks cool.  Let me try!

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.