Mihai Sucan wrote:
Yet getElementById is defined as [2]:
<blockquote>
Returns the Element that has an ID attribute with the given value.
If no such element exists, this returns null.
If more than one element has an ID attribute with that value, what
is returned is undefined.
</blockquote>
Therefore... the appropriate behaviour for
getElementById("duplicate-ID") is to return null.
No, the spec says the behaviour is undefined and so UAs can return
whatever they like and still be considered conformant. I agree that it
should be defined, but it should be defined in the way most compatible
with existing implementations and since no existing implementation (at
least none that I know of) returns null, that's likely to break many pages.
Simply picking the last matching node is actually hiding a bug and
letting it go unnoticed. (Why the last one? Why not the first, for
example?)
All the tests I made up showed that the first was returned in all
browsers I tested, including FF, IE, Opera, Safari and iCab. None of
them returned the last.
This is one of my test cases:
<!DOCTYPE html>
<script>
window.onload = function() {
document.getElementById("foo").appendChild(document.createTextNode("This
One!"));
}
</script>
<p id="foo">1: </p>
<p id="foo">2: </p>
The result:
* P id="foo"
o #text: 1:
o #text: This One!
* #text:
* P id="foo"
o #text: 2:
--
Lachlan Hunt
http://lachy.id.au/
