ES6: Object Literal Methods toString is missing method name

[Joseph Pecoraro]

* SUMMARY
Object Literal Methods toString is missing method name.

* TEST
o = { method(a,b){} };
o.method.toString();

* EXPECTED:
method(a,b){}

* ACTUAL
(a,b){}

[Joseph Pecoraro]

Function.prototype.string:
https://people.mozilla.org/~jorendorff/es6-draft.html#sec-function.prototype.tostring

Note that the output can, and should be, the syntax of a `MethodDefinition`, not `FunctionDeclaration FunctionExpression` in this case.

I'm not sure what would be best for computed method names:

    o = { ['method'](a,b){} };
    o.method.name; // ""
    o.method.toString(); // => ???

That is a MethodDefinition.

Firefox outputs "function (a,b){}" in that case, which I think makes some sense.

[Joseph Pecoraro]

Wow, I didn't realize this, but this was in my test and failing: =(
LayoutTests/js/object-literal-methods.html

[Joseph Pecoraro]

(In reply to comment #2)
> Wow, I didn't realize this, but this was in my test and failing: =(
> LayoutTests/js/object-literal-methods.html

Oh, never mind. This broke with r181810!
<https://webkit.org/b/142853> Function.prototype.toString should not decompile the AST

[Geoffrey Garen]

It looks like Firefox strigifies by concatenating "function " + function.name + [all text from opening paren to closing brace], with a separate code path for arrow functions. I think that's a reasonable solution.

[Joseph Pecoraro]

(In reply to comment #4)
> It looks like Firefox strigifies by concatenating "function " +
> function.name + [all text from opening paren to closing brace], with a
> separate code path for arrow functions. I think that's a reasonable solution.

This will add back the space for an anonymous function:

  (function(a,b){}).toString();
  // Currently   : "function(a,b){}"
  // Always Space: "function (a,b){}"

But this does match Firefox, always adding the space. So I'll go that route.

[Joseph Pecoraro]

Created attachment 249305 [details]
[PATCH] Proposed Fix

I want this to run through all tests on EWS because I only did a subset based on what the previous change had changed.

[Geoffrey Garen]

Comment on attachment 249305 [details]
[PATCH] Proposed Fix

r=me

[WebKit Commit Bot]

Comment on attachment 249305 [details]
[PATCH] Proposed Fix

Clearing flags on attachment: 249305

Committed r181901: <http://trac.webkit.org/changeset/181901>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.