202168 - "arguments" passed as an argument not treated as an array

Status: VERIFIED INVALID

(Rhino Graveyard :: Core, defect)

Description

[Larry Blair]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.3) Gecko/20030312
Build Identifier: Rhino 1.5R4

Date:	 Tue, 15 Apr 2003 21:01:15 +0200
From:	"Igor Bukanov" <[e-mail]>
Subject:	Re: Is "arguments" an array or not in Rhino?

Larry Blair wrote:
> The following function works in spidermonkey, but not in Rhino:
> 
> function f(){
>   var c = [].concat(arguments);
>   return c;}
> var x = f("A", "B");
> die (x.length+"|"+x[0]+"|"+x[1]);
> 
> Spidermonkey results = 2|A|B
> Rhino results = 1|[object Arguments]|undefined
> 
> Btw, the following variation works in Rhino.  Unfortunately, our 1900 
> app files don't use it:
> 
> function f(){
>   var c = [].concat([arguments[0], arguments[1]]);
>   return c;}
> var x = f("A", "B");
> die (x.length+"|"+x[0]+"|"+x[1]);
> 

This is a bug in Rhino. In fact, this is a regression since 1.5R3. 
Could you file a bugzilla.mozilla.org report for it?

Regards, Igor



Reproducible: Always

Steps to Reproduce:
1.
2.
3.

Comment 1

[Phil Schwartau]

Larry, what version of SpiderMonkey are you using? The current versions
of SpiderMonkey and Rhino give essentially the same output for your test:

---
function f()
{
  var c = [].concat(arguments);
  return c;
}

var x = f("A", "B");
print(x.length + "|" + x[0] + "|" +x[1]);
---


OUTPUT IN SPIDERMONKEY  1|[object Object]|undefined
OUTPUT IN RHINO         1|[object Arguments]|undefined


This is correct. The |arguments| object inside a function
is just an object that happens to have a |length| property,
and properties |0|, |1|, etc. storing the function's parameters.

But it is just an ordinary object, not an instance of an Array.
To see this, we use the |instanceof| operator :

function test()
{
  print(arguments instanceof Array);
}

The output of this in both SpiderMonkey and Rhino is "false".

Comment 2

[Phil Schwartau]

Here is the relevant portion of the ECMA-262 Edition 3 spec
(See http://www.mozilla.org/js/language/E262-3.pdf)

---
10.1.8 Arguments Object
When control enters an execution context for function code,
an arguments object is created and initialised as follows:

• The value of the internal [[Prototype]] property of the arguments object
  is the original Object prototype object, the one that is the initial value
  of Object.prototype (section 15.2.3.1).

• A property is created with name callee and property attributes {DontEnum}.  
  The initial value of this property is the Function object being executed.
  This allows anonymous functions to be recursive.

• A property is created with name length and property attributes {DontEnum}.  
  The initial value of this property is the number of actual parameter values   
  supplied by the caller.

• For each non-negative integer, arg, less than the value of the length 
  property, a property is created with name ToString(arg) and property   
  attributes {DontEnum}. The initial value of this property is the value
  of the corresponding actual parameter supplied by the caller. The first
  actual parameter value corresponds to arg = 0, the second to arg = 1,
  and so on. In the case when arg is less than the number of formal parameters   
  for the Function object, this property shares its value with the corresponding 
  property of the activation object. This means that changing this property   
  changes the corresponding property of the activation object and vice versa.
---


Note in particular, the [[Prototype]] property of the arguments object
is the original Object prototype object, not the Array prototype object.

Have to mark this one Invalid, sorry -

Comment 3

[Phil Schwartau]

Marking verified. 

Thank you for this report, it illuminates an interesting question.

For details on Array.prototype.concat(), and what it does when it is
supplied a non-array parameter, see bug 169795, "Array.concat(function)
doesn't add function (or anything with .length but no elements) to the array"

Comment 4

[Igor Bukanov]

Targeting as resolved against 1.5R5