Bug 143799 – Number.parseInt is not === global parseInt in nightly r182673

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Jordan Harband]

The spec requires that `Number.parseInt === parseInt`, but that's not the case in WebKit nightly.

[Jordan Harband]

Linking to other Number.parseInt bug.

[Jordan Harband]

(fwiw, this was also a bug in Firefox 37, but it's now fixed in their nightly v40)

[Jordan Harband]

Created attachment 250902 [details]
Patch

[Jordan Harband]

Created attachment 250947 [details]
Patch

[Darin Adler]

Comment on attachment 250947 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=250947&action=review

> Source/JavaScriptCore/runtime/NumberConstructor.cpp:76
> +    putDirectWithoutTransition(vm, Identifier::fromString(&vm, "parseInt"), numberPrototype->globalObject()->parseIntFunction(), DontEnum | Function);

Why isn’t this using vm.propertyNames->parseInt?

[Jordan Harband]

(In reply to comment #5)

> Comment on attachment 250947 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=250947&action=review
> 
> > Source/JavaScriptCore/runtime/NumberConstructor.cpp:76
> > +    putDirectWithoutTransition(vm, Identifier::fromString(&vm, "parseInt"), numberPrototype->globalObject()->parseIntFunction(), DontEnum | Function);
> 
> Why isn’t this using vm.propertyNames->parseInt?

I didn't want to couple the `Number` function name with the global name - they're the same string, and the same function, but they're conceptually not linked in the same way. Would you prefer I change it to do so?

[WebKit Commit Bot]

Comment on attachment 250947 [details]
Patch

Clearing flags on attachment: 250947

Committed r182938: <http://trac.webkit.org/changeset/182938>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.