These specifications seem to be addressing what I'm attempting to do on my own, but from a quick glance over them, it seems to be mostly referencing parts of a document as per the XML (eXtensible Markup Language) structure of the document (as in, give me the contents of the third <P> tag) which isn't all that bad but the URL (Universal Resource Locator)s get rather ugly, rather fast.
That's not to say that URLs aren't getting ugly now, but in trying to print these out, or even type them in is getting difficult to say the least.
[1] http://www.w3.org/TR/xpath