237579 - Strict warnings for trailing comma in array initializers

Status: RESOLVED DUPLICATE of bug 219308

(Core :: JavaScript Engine, enhancement)

Description

[Paul Gregg]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; rv:1.7a) Gecko/20040219
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; rv:1.7a) Gecko/20040219

Similar feature request to Bug 103602, Bug 164400 & Bug 179628.

Basically asking for the JS engine to produce a warning when in strict mode for
a trailing comma in array initializers.

After seeing the Bug 235260, it prompted me to raise a RFE, after I was bitten
by a similar problem (array initializers not object initializers, which seem to
fire a strict warning).

The reason this bites for web developers is that IE4, 5, 5.5 and 6 increase the
reported length of the array by one (maybe more if there's more commas), which
can stuff up loop calculations. Moz1.6, NN4 and Opera7.11 all report the proper
array length while ignoring the trailing comma.

Therefore, it would be nice if Moz reported a warning in strict mode when the
trailing comma appears.

Reproducible: Always
Steps to Reproduce:
1. Load some JS with an array that has a trailing comma after the initializers.
Actual Results:  
No warning in strict mode.

Expected Results:  
Warning generated and displayed in JS console when an array initializer has a
trailing comma.

Will attach a reduced testcase soon.

Comment 1

[Paul Gregg]

Attachment: html testcase with JS array created with trailing comma

In Moz1.6, NN4 and Opera7.11:-
  groups.length = 4
In IE4, 5, 5.5 and 6:-
  groups.length = 5

This should raise a JS strict warning in the JS console.

Comment 2

[Paul Gregg]

Adding "testcase" keyword (would be nice if you could from the Create a New
Attachment page...), since I uploaded a simple testcase that should fine a
strict warning.

Comment 3

[Christian :Biesinger (don't email me, ping me on IRC)]

this sounds like a duplicate of bug 219308... so it looks the other person
wanted in bug 219308 comment 10 has shown up :)

*** This bug has been marked as a duplicate of 219308 ***

Comment 4

[Brendan Eich [:brendan]]

Re: comment 3, see bug 219308 comment 6, in particular:

  One problematic, subtle issue here: IE is simply botching the array length for
  [1,2,3,], which misleads code to expect (3 < a.length) and (3 in a) to be true.
  We will never emulate this bug, so just a warning may not be enough to cause
  the few who were misled by IE's bug to fix their code -- but it may bug the heck
  out of people who (based on righteous Perl or trad-C experience) want to put a
  trailing , in array initializers.

I went on to say that since a warning is not enough to gain compatibility, and
as we are *not* going to emulate IE's broken-ness in computing length, I'm not
in favor of adding a warning.

Maybe we could have a different kind of option than strict: "iebugs".  It would
warn about any construct buggily implemented by IE's JScript engine.  But I have
zero time or interest in doing that.

/be