Herbert Wender wrote:
> BTW, the most poor minded people like me would prefer the simple
> old-fashioned way to escape metalanguage: Programs like TeX are aware that BackSlash
> indicates 'Take the following character code litteraly!'
hmm. I would say that is far from the truth.
a \ indicates that the following "thing" is a TeX
command sequence. And that is only if \ retains
its default catcode. Any character can have its
meaning changed in TeX, which is what makes
it such a fearsome language to parse without
reimplementing all of TeX's mouth over again.
> When I understand it right, the TEI processing application program is
> checking the stream of code numbers for splitting the metalanguage expressions from
> the text marked-up, then it checks if these expressions are allowed by the
> actual DTD or schema (= step 1 of 'validation'?) and then validates the structure
> of the given markup.
yes, more or less. although that work is seldom
done by ones own application, but rather mediated
through a standard XML parser.
> When this structure is recognized as vaild TEI encoding,
> I thought that the TEI processor never is looking at the content of the
> elements, not only in the case of the _formula_ element? Or is it now possible to
> define a language attribute in a way that the content of some elements is
> checked against the conventions of f.e. 'Neue Deutsche Rechtschreibung'?
not sure what N D R is, but yes, the new schema
languages can validated text nodes in the same way
as DTDs validate attributes. So you can ask
a schema-based validator to check that
is a valid date or not, or that
matches the regexp
If that is what you mean. and if you can
express Neue Deutsche Rechtschreibung
in terms of schema datatypes.
Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431