Re: The one element schema language

[Joe English <jenglish@flightlab.com>]

A few corrections:

Earlier I wrote:

> a relation '<<' on S is a strict weak order
> if there is a homomorphism from (S, <<) to (Z, <) where (Z, <)
> is a totally ordered set.

This is incorrect.  For h : (S, <<) -> (Z, <) to be a
homomorphism only requires that for all x,y:

	x << y ==> h(x) < h(y)

whereas for << to be a strict weak order requires the stronger condition:

	x << y <==> h(x) < h(y)

See <URL: http://www.sgi.com/Technology/STL/StrictWeakOrdering.html >
for an alternate definition.

More importantly, I got the use of 'imax' and 'imin' backwards
in the validation algorithm; it should read instead:

>     variable iprev : integer := -1;
>     while not end-of-document():
> 	case next-event() of
> 	    START-TAG(n):
> 		    if (imax(H,n)) < iprev then error "Invalid"
> 		    iprev := imin(H,n)
> 	    END-TAG(n):
> 		    iprev := imin(H,n)


--Joe English

  jenglish@flightlab.com