Background:
I came across a very strange phenomenon while working with a node list. I wanted to use getElementsByClassName or something similar and then sort it. I decided one way would be to iterate through the nodelist and push each item to an array and sort the array. (This did work by the way but not as expected). I tried using the for (var i in nodeList)
to开发者_如何学Go iterate through, but it kept throwing an exception on the last few items, which were undefined. the weird part is I could instead use for (var i = 0; i < nodeList.length; i++)
to iterate through. I just tested it again and on a stackoverflow page I ran in my console the following code:
for (var i in document.getElementsByTagName("span"))
console.count("items");
console.log(document.getElementsByTagName("span").length);
It counted out to items: 382
but length gave 380
. As expected, when I entered document.getElementsByTagName("span")[380]
and document.getElementsByTagName("span")[381]
they came back undefined. This strange behavior does not occur on arrays (granted, nodeLists and arrays are different, but this does prove that it's not the different for loops causing the issue).
question:
Why does for(var i in nodeList)
constructs behave differently on nodeLists returning a couple of undefined items at the end?
The two additional properties that the for in
iteration statement catches are:
- length
- item
Let me give you a simple example. Let's say that there are 3 SPAN elements on the page.
var spans = document.getElementsByTagName( 'span' );
Now, spans
is a NodeList object which contains 5 properties:
- 0
- 1
- 2
- length
- item
The first 3 properties are indexes and they contain references to those SPAN elements. The other two properties - length and item - are two additional properties. All NodeList objects have these two properties.
The for in
statement iterates over all 5 properties of the NodeList object, which is probably not what you want. Therefore, use a regular for
statement.
var i, span;
for ( i = 0; i < spans.length; i++ ) {
span = spans[i];
// do stuff with span
}
for-in iterates through the all enumerable properties of the object such as length and item (this is your situation). This is where two more results come from. It will also enumerate everything added to object prototype.
for loops through the numeric indeces and doesn't take into consideration enumerable properties. That is why it much more reliable to use former way.
精彩评论