Re: Using WPT as unit tests for polyfill libraries?

Ah, thanks for the heads up. I guessed somewhere in the spec, it requires undefined to be treated the same as missing for optional arguments.

Looks like blink and webkit are doing it wrong right now.

I will give web idl a through read when I got the time. Hope I won’t send a lot of incorrect tests to embarrass myself. :)

> On Feb 19, 2015, at 3:53 PM, Ms2ger <> wrote:
> Hash: SHA1
> On 02/19/2015 05:25 AM, Glen Huang wrote:
>> @Ms2ger
>> Glad to hear they can be merged.
>> As for classList.force, since we are explicitly passing an
>> ECMAScript value to it, according to WebIDL, it will first be
>> converted to boolean by using ToBoolean(undefined), which returns
>> false. So toggle("name", undefined) should actually remove the
>> class and not toggling it. But I haven't go through WebIDL very
>> throughly, so I'm not absolutely sure that's correct, will go
>> through it more carefully when I'm writing the tests.
> Off-list, since not really relevant: the overload resolution algorithm
> <>
> treats undefined as triggering the default.
> Ms2ger
> lWjNCS7D1XN0eUnucir380n+8QrI5s2vv4tcr0KxiLWhNaJENFHL2K4FaNM2QeMI
> gr3BPdW82jwnLw534J3EfdxSIDcerwnu2A1GJjSst5KijLrGpXK/DAdKrmjuOGoW
> P2W2TvqtcB1nlx8Th1b+nj4MVWXaQuwQHqh5MdOq/aADBg0O4ae9D1ZW0rZwD+0p
> 0tdA52iYVk6nCIJ5KmqBYuMWwHj2LmDLc66IDyIjgRuu5Vb5dDlpO0J6PqQlYSYw
> TabUnD9rAzPTd0v0dsGFXZIMGeFHQ+OsO5uvHWlA1XyJErzcmIc6ChJ6sB0PCL0=
> =jD9c

Received on Friday, 20 February 2015 10:46:57 UTC