The name of the DOM function is actually getElementsByClassName, not getElementByClassName, simply because more than one element on the page can have the same class, hence: Elements.
The return value of this will be a NodeList instance, or a superset of the NodeList (FF, for instance returns an instance of HTMLCollection). At any rate: the return value is an array-like object:
var y = document.getElementsByClassName(‘foo’);
var aNode = y[0];
If, for some reason you need the return object as an array, you can do that easily, because of its magic length property:
var arrFromList = Array.prototype.slice.call(y);
//or as per AntonB’s comment:
var arrFromList = [].slice.call(y);
As yckart suggested querySelector(‘.foo’) and querySelectorAll(‘.foo’) would be preferable, though, as they are, indeed, better supported (93.99% vs 87.24%), according to caniuse.com:
querySelector(all)
getElementsByClassName
Don’t use w3schools to learn something
Refer to MDN for accurate information
Another option is to use querySelector(‘.foo’) or querySelectorAll(‘.foo’) which have broader browser support than getElementsByClassName.
http://caniuse.com/#feat=queryselector
http://caniuse.com/#feat=getelementsbyclassname