I'm not sure what you said about typenames and attribute
names, but I wonder if it is this problem. XML ELEMENT names are
confused about whether they are acting as programming language "type
names", or as programming language "field names".
In a "data oriented schema" (aka, Type Definition Language)
we might write
TYPE POINT
{
x : INT
y : INT
}
TYPE RECTANGLE
{
lower-left : POINT
upper-right : POINT
}
and then we might have instances:
<POINT> -- Point is the Type Name
<x>3</x> -- X is a field name
<y>4</y>
</POINT>
<RECTANGLE> -- type
<lower-left> -- field
<POINT> -- type
<x>3</x> -- field
<y>4</y>
<POINT>
</lower-left>
<upper-right>
<POINT>
<x>30</x>
<y>40</y>
<POINT>
</upper-right>
</RECTANGLE>
Programming languages make it clear when an identifier defines a
type (POINT), and when it uses a type (lower-left).
-- Mark