[
Lists Home |
Date Index |
Thread Index
]
- From: Tim Bray <tbray@textuality.com>
- To: Elliotte Rusty Harold <elharo@metalab.unc.edu>, xml-dev <xml-dev@xml.org>
- Date: Sat, 11 Mar 2000 10:34:29 -0800
At 01:07 PM 3/11/00 -0800, Elliotte Rusty Harold wrote:
>However a trick I picked up from the MathML 2.0 DTD is to use double
>indirection like this:
><!ENTITY % rd-prefix "rd">
><!ENTITY % rd-colon ":">
><!ENTITY % rd-test "%rd-colon;%rd-prefix;test">
><!ELEMENT %rd-test; (#PCDATA)>
>This works but for the life of me I can't see why. How come the rd-test
>entity is defined as "rd:test" instead of " rd : test"? Can anyone
>justify this behavior in terms of the spec?
Yeah, if you look at the big table in 4.4, this falls under "Reference
in Entity Value", with the behavior described under 4.4.6, "Included
in Literal". The first example has one of the dumbest errata in the spec,
the &YN; should be %YN; - what's there is wellformed but fails to illustrate
the point it's trying to make. I have some more on this in the annotated
spec.
Hadn't seen that trick before. Not sure whether it's disgusting or
brilliant. -Tim
***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************
|
