Fun impossible Firefox JS challenge

Andreas Rossberg rossberg at google.com
Thu Apr 12 08:06:07 PDT 2012


On 12 April 2012 16:27, Peter van der Zee <ecma at qfox.nl> wrote:
> On Thu, Apr 12, 2012 at 4:12 PM, Andreas Rossberg <rossberg at google.com> wrote:
>>> Haha nice try even with unicode escapes it still refers to "true" the
>>> boolean not the function.
>>
>> That's another FF deviation from the standard, though.
>
> Identifiers with unicode escapes have the meaning of their canonical
> value. So wouldn't that (tru\u0065 referring to the bool) be valid and
> according to the spec?

Yes, but keywords (and other verbatim tokens, like 'true') are
recognized _before_ canonicalization. At least that is the intention
of the spec, Waldemar told me. (I filed
https://bugs.ecmascript.org/show_bug.cgi?id=277 a few weeks ago as a
request for making the wording clearer.)

/Andreas


More information about the es-discuss mailing list