Ken
> A DOM data model contains entity references ... could you use
> that?
Not declaratively. I'd prefer to do it using a technique that can be mimiced
in XSLT, though I found out the hard way today that trying to use CDATA
marked sections to get a {name()} instruction through to a second
transformation is not possible. What I did find, however, was that if I put
the replacement string in a variable and then invoked that I got what I
wanted. Maybe I can do the same by putting &xyz; into a variable. Will
have to suck it and see if nobody can suggest anything easier.
> Does it cover the bases for ensuring you have the required
> declaration for the renamed entity so that the result is
> well-formed? Remember it is not an error for two declarations of the
> same entity (first one wins) so what happens if you rename "a" to "b"
> but "b" already exists? To be well-formed there must already be an
> entity named "b". If you add "a"'s declaration as a declaration for
> newly renamed "b", which definition with the old "b"'s get? Will "a"
> accidentally get the old b's declaration?
Don't think we need to cover all bases. We can simply say it is an error to
define an entity that already exists.
> What is the use case for entity renaming?
Two main ones:
1) Entity reference retention ( -> rather than  )
2) Localization/easier comprehension for people with limited English (&et;
or ∧ rather than &)
Martin
-- DSDL members discussion list To unsubscribe, please send a message with the command "unsubscribe" to dsdl-discuss-request@dsdl.org (mailto:dsdl-discuss-request@dsdl.org?Subject=unsubscribe)Received on Mon Dec 12 20:27:52 2005
This archive was generated by hypermail 2.1.8 : Tue Dec 13 2005 - 00:13:02 UTC