- From: William F Hammond <hammond@csc.albany.edu>
- Date: 07 Aug 2002 19:03:02 -0400
- To: www-html@w3.org
Jens Müller <jens.lists@unfaehig.de> writes:
> > <!ENTITY % charstuff "#PCDATA|ltc|gtc|amp|dollar|euro">,
> >
> > and, of course, use of the character entities would not be precluded.
> < etc. are characters (represented by named entity references) and
> treated as such in any transformation and in any parser.
"<" after a parse is merely "<". I do not want to place
numeric values in, say, an XSLT sheet. In fact, on this topic see
complaints about the necessity of using raw unicode values with MathML
in the log for www-math@w3.org since user agents do not always read
dtd's.
Moreover, I do not to want make transformation branching decisions on
CDATA points. Transformation branching is sometimes necessary with
special characters such as "<".
> What is the advantage of having an element present a character? Why
> not have <letter-a /> representing "a" etc.?
No, since there is no output format I imagine in which "a" is a
special character.
In this vein, however, something like <specchar name="ltc"/>
would be OK; with that the short form <ltc/> would be only for
convenience, and I would then limit the short forms to "ltc", "gtc",
and "amp".
-- Bill
Received on Wednesday, 7 August 2002 19:03:04 UTC