Attribute Filter predicates

[Rod Widdowson]

> Since it is a compatibility thing, and since we guess that most of the
> weirder combinations probably aren't used much, presumably it could be
> cleaned up later.

Leave this with me.  I am wondering whether there may be a middle way that
leaves the weirdness in one place (probably the filter engine), allows
context insensitive parsing, and allows the predicates to concentrate on
what they should do in the normal place.

I'll need to "interview this in the code" and see what comes out at the end.
The screw cases are likely to be the logical rules, which function in both
cases.  But they can be a bit weird in the <AttributeRule> case anyway.

Rod