1225009 - String#match no longer respects Symbol.match

Status: RESOLVED DUPLICATE of bug 887016

(Core :: JavaScript: Standard Library, defect)

Description

[ljharb]

User Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_5) AppleWebKit/601.2.7 (KHTML, like Gecko) Version/9.0.1 Safari/601.2.7

Steps to reproduce:

In Firefox 40, the following code gives `true`:
`var o = {}; o[Symbol.match] = function () { return 42; }; ''.match(o) === 42`

In Firefox 41 and later (including latest nightly, it gives `false` - because it's returning `null` instead.


Actual results:

I got `null` instead of the value the Symbol.match matcher returned.


Expected results:

It should have returned `42`.

Comment 1

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

I can't reproduce the behavior described in comment 0 for Firefox 40.0.3. It seems to be similar for 40 and 42. Doesn't Symbol.match control use as regular expression and not the return value (and for that reason returns null and false)?

Comment 2

[Tooru Fujisawa [:arai]]

Symbol.match is implemented from Firefox 40 (bug 1054755), but ES6 compliant String.prototype.match/RegExp.prototype[@@match] are not yet implemented. see bug 887016.

In current implementation, o.toString() == "[object Object]" is parsed as RegExp pattern (so, single character class), and it's used in match, and the match fails.  it won't return 42 anyways.

So, the question is that why/how it returns 42 and true on your Firefox 40.
Can you still reproduce it?

Comment 3

[ljharb]

I can still reproduce it. I'm using Firefox 40.0 on Windows 8 and also 40.0 on any Mac OS on Browserstack.

Why was `Symbol.match` added in the first place if it wasn't meant to be fully compliant?

Comment 4

[ljharb]

Ah wait - scratch that. It is indeed returning null for me in Firefox 40.

In that case, this may be a duplicate of bug 887016.

I'm still confused why a well-known symbol would be added that isn't compliant with the spec, but that's neither here nor there - I'll just have to patch it in the es6-shim until it's fixed :-)