Print

Print


Quoting Lou Burnard <[log in to unmask]>:

> Another way of addressing this issue might  be simply to make these
> elements a member of the att.spanning class
>
> That is, for each such element xxx, rather than invent a sister element
> xxxSpan, we modify its definition such that
> a) it  can take a @spanTo attribute
> b) we add a schematron rule which says that if @spanTo is supplied the
> element must be empty

This could also be expressed in the RNG grammar.

 Steve