This is another example showing how eyes may get easily confused by
Javascript formal syntax:
var a = 10;
function test() {
return a;
var a = null;
}
console.log( test() )
The returned value obtained form invoking the "test()" function is
neither 10 nor null as it looks like at first sight in the above code.
Declarations are moved atop the functions by the lexical scanner so,
at some point, the internal representation may look like this:
var a = 10; // global declaration
function test() {
var a; // undefined value declaration (moved here by LS)
return a;
a = null; // non-reachable statement
}
console.log( test() )
Put all your variables atop the functions and you avoid related
problems (and probably shorten code as bonus).
--
Diego
On 17 Dic, 20:11, Garrett Smith <[email protected]> wrote:
> On 12/16/10, Darth Vincent Binamira <[email protected]> wrote:
>
> >> Next, in evaluation of statements, the IfStatement is evaluated. So
> >> what is `"a" in window"`? result? In a web browser, the global object
> >> its he global variable object is the window. Thus `"a" in window` is
> >> true.
>
> >> So `"a" in window` is true, and the Logical NOT operator (!) converts
> >> that to false. And:
> >> | if( !("a" in window) )
>
> >> Must be false!
>
> > I just tested this out in firebug's console and I'm a little bit confused.
> > `"a" in window` returned false, and `!("a" in window)` returned true.
>
> > Can you give a bit of clarification?
>
> I assume Firebug uses eval and IIRC it uses `with` for augmented
> scope. Expect different results in Firebug.
>
> > Sorry for asking this, but its really confusing. Thanks.
>
> I've added a bit to my explanation.
>
> [...]
>
> >> Example:
> >> | if (!("a" in window)) {
> >> | var a = 1;
> >> | }
> >> | alert(a);
>
> >> There are two passes of the code here. First, there is Variable
> >> Instantiation. This is where identifiers with variable statement,
> >> FunctionDeclaration, and parameters (for function code) are added to
> >> the variable object (VO), and given attribute {DontDelete}.
>
> So in Variable Instantiation, the code is first scanned for variable
> declarations (among other things).http://bclary.com/2004/11/07/#a-10.1.3
>
> The assigment `a = 1` is not part of the declaration, so it is skipped.
>
> When the interpreter sees `var a`, if the VO does not have an `a`
> property, then `a` is added as a property of the VO and given the
> value `undefined`.
>
> The global VO is also the global object, so the global object now has
> an `a` property. The global object is aliased by `window`. Thus, after
> variable instantiation, `"a" in window` is true.
>
> [...]
>
>
>
>
>
> >> The second pass is evaluation of Statements, such as the IfStatement
> >> or the assigmentExpression `a = 1`.
>
> >> In Variable Instantiation, `a` is added to the VO, if it does not
> >> already exist, and given value `undefined`. If the VO already has an
> >> `a` property, its value is unchanged.
>
> >> Next, in evaluation of statements, the IfStatement is evaluated. So
> >> what is `"a" in window"`? result? In a web browser, the global object
> >> its he global variable object is the window. Thus `"a" in window` is
> >> true.
>
> >> So `"a" in window` is true, and the Logical NOT operator (!) converts
> >> that to false. And:
> >> | if( !("a" in window) )
>
> >> Must be false!
>
> >> The result of the program is the creation of a global identifier `a`
> >> with value `undefined` and attributes {DontDelete}.
>
> [...]
>
> I should note that in ECMA-262 Ed. 5, the "variable object" was
> renamed to "lexical environment
> record".http://ecma262-5.com/ELS5_Section_10.htm#Section_10.5
>
> As stated, in variable instantiation, the variable declarations are
> added to the VO. I did not explain that FunctionDeclarations and
> formal parameters are added to the VO. They are, as explained in the
> spec.
> See
> also:https://groups.google.com/group/comp.lang.javascript/browse_thread/th...
>
> - which links to the closures article which explains in more detail.
> Of particular relevance to this
> discussion:http://jibbering.com/faq/notes/closures/#clIRExSc
> --
> Garrett
--
To view archived discussions from the original JSMentors Mailman list:
http://www.mail-archive.com/[email protected]/
To search via a non-Google archive, visit here:
http://www.mail-archive.com/[email protected]/
To unsubscribe from this group, send email to
[email protected]
