I did not yet see this variation, which I personally like the best:

Given an array:

var someArray = ["some", "example", "array"];

You can loop over it without ever accessing the length property:

for (var i=0, item; item=someArray[i]; i++) {
  // item is "some", then "example", then "array"
  // i is the index of item in the array
  alert("someArray[" + i + "]: " + item);
}

See this JsFiddle demonstrating that: http://jsfiddle.net/prvzk/

This only works for arrays that are not sparse. Meaning that there actually is a value at each index in the array. However, I found that in practice I hardly ever use sparse arrays in Javascript... In such cases it's usually a lot easier to use an object as a map/hashtable. If you do have a sparse array, and want to loop over 0 .. length-1, you need the for (var i=0; i<someArray.length; ++i) construct, but you still need an if inside the loop to check whether the element at the current index is actually defined.

Also, as CMS mentions in a comment below, you can only use this on arrays that don't contain any falsish values. The array of strings from the example works, but if you have empty strings, or numbers that are 0 or NaN, etc. the loop will break off prematurely. Again in practice this is hardly ever a problem for me, but it is something to keep in mind, which makes this a loop to think about before you use it... That may disqualify it for some people :)

What I like about this loop is:

• It's short to write
• No need to access (let alone cache) the length property
• The item to access is automatically defined within the loop body under the name you pick.
• Combines very naturally with array.push and array.splice to use arrays like lists/stacks

The reason this works is that the array specification mandates that when you read an item from an index >= the array's length, it will return undefined. When you write to such a location it will actually update the length.

For me, this construct most closely emulates the Java 5 syntax that I love:

for (String item : someArray) {
}

... with the added benefit of also knowing about the current index inside the loop

I did not yet see this variation, which I personally like the best:

Given an array:

var someArray = ["some", "example", "array"];

You can loop over it without ever accessing the length property:

for (var i=0, item; item=someArray[i]; i++) {
  // item is "some", then "example", then "array"
  // i is the index of item in the array
  alert("someArray[" + i + "]: " + item);
}

See this JsFiddle demonstrating that: http://jsfiddle.net/prvzk/

This only works for arrays that are not sparse. Meaning that there actually is a value at each index in the array. However, I found that in practice I hardly ever use sparse arrays in JavaScript... In such cases it's usually a lot easier to use an object as a map/hashtable. If you do have a sparse array, and want to loop over 0 .. length-1, you need the for (var i=0; i<someArray.length; ++i) construct, but you still need an if inside the loop to check whether the element at the current index is actually defined.

Also, as CMS mentions in a comment below, you can only use this on arrays that don't contain any falsish values. The array of strings from the example works, but if you have empty strings, or numbers that are 0 or NaN, etc. the loop will break off prematurely. Again in practice this is hardly ever a problem for me, but it is something to keep in mind, which makes this a loop to think about before you use it... That may disqualify it for some people :)

What I like about this loop is:

• It's short to write
• No need to access (let alone cache) the length property
• The item to access is automatically defined within the loop body under the name you pick.
• Combines very naturally with array.push and array.splice to use arrays like lists/stacks

The reason this works is that the array specification mandates that when you read an item from an index >= the array's length, it will return undefined. When you write to such a location it will actually update the length.

For me, this construct most closely emulates the Java 5 syntax that I love:

for (String item : someArray) {
}

... with the added benefit of also knowing about the current index inside the loop

I did not yet see this variation, which I personally like the best:

Given an array:

var someArray = ["some", "example", "array"];

You can loop over it without ever accessing the length property:

for (var i=0, item; item=someArray[i]; i++) {
  // item is "some", then "example", then "array"
  // i is the index of item in the array
  alert("someArray[" + i + "]: " + item);
}

See this JsFiddle demonstrating that: http://jsfiddle.net/prvzk/

This only works for arrays that are not sparse. Meaning that there actually is a value at each index in the array. However, I found that in practice I hardly ever use sparse arrays in Javascript... In such cases it's usually a lot easier to use an object as a map/hashtable. If you do have a sparse array, and want to loop over 0 .. length-1, you need the for (var i=0; i<someArray.length; ++i) construct, but you still need an if inside the loop to check whether the element at the current index is actually defined.

What I like about this loop is:

• It's short to write
• No need to access (let alone cache) the length property
• The item to access is automatically defined within the loop body under the name you pick.
• Combines very naturally with array.push and array.splice to use arrays like lists/stacks

The reason this works is that the array specification mandates that when you read an item from an index >= the array's length, it will return undefined. When you write to such a location it will actually update the length.

For me, this construct most closely emulates the Java 5 syntax that I love:

for (String item : someArray) {
}

... with the added benefit of also knowing about the current index inside the loop

I did not yet see this variation, which I personally like the best:

Given an array:

var someArray = ["some", "example", "array"];

You can loop over it without ever accessing the length property:

for (var i=0, item; item=someArray[i]; i++) {
  // item is "some", then "example", then "array"
  // i is the index of item in the array
  alert("someArray[" + i + "]: " + item);
}

See this JsFiddle demonstrating that: http://jsfiddle.net/prvzk/

This only works for arrays that are not sparse. Meaning that there actually is a value at each index in the array. However, I found that in practice I hardly ever use sparse arrays in Javascript... In such cases it's usually a lot easier to use an object as a map/hashtable. If you do have a sparse array, and want to loop over 0 .. length-1, you need the for (var i=0; i<someArray.length; ++i) construct, but you still need an if inside the loop to check whether the element at the current index is actually defined.

Also, as CMS mentions in a comment below, you can only use this on arrays that don't contain any falsish values. The array of strings from the example works, but if you have empty strings, or numbers that are 0 or NaN, etc. the loop will break off prematurely. Again in practice this is hardly ever a problem for me, but it is something to keep in mind, which makes this a loop to think about before you use it... That may disqualify it for some people :)

What I like about this loop is:

• It's short to write
• No need to access (let alone cache) the length property
• The item to access is automatically defined within the loop body under the name you pick.
• Combines very naturally with array.push and array.splice to use arrays like lists/stacks

The reason this works is that the array specification mandates that when you read an item from an index >= the array's length, it will return undefined. When you write to such a location it will actually update the length.

For me, this construct most closely emulates the Java 5 syntax that I love:

for (String item : someArray) {
}

... with the added benefit of also knowing about the current index inside the loop