I am trying to write a GreaseMonkey script in which I want to find all of the links that are relative links. It seemed to me that the way to do that would be to match the contents of href against /^https?:///
.
But I find that when I access the anchor's href attribute, it's always normalized or cooked into a form that contains "http". That is, if the HTML contains:
<a id="rel" href="/relative/link">inner</a>
accessing
document.getElementById("rel").href
returns
http://example.com/relative/link
How can I access the raw data in the href attribute?
Alternately, is there a better way to find relative links?