Entity replacement

James K. Tauber (jtauber@jtauber.com)
Sun, 1 Jun 1997 13:09:05 +0800

Messages sorted by: [ date ][ thread ][ subject ][ author ]

Am I correct in thinking that one benefit of requiring entity and element
structure to be synchronized is that a well-formed document is also
well-formed before general entity replacement; ie you can parse the
document before entity replacement, parse the entities and then just insert
the parse tree of the later into the former?

If this is true, then is there some similar constraint that could be
applied to use of parameter entities? This might already have been done in
the choice of where to use % in the productions but I can't quite work out
the pattern. It would be nice if parameter entity replacement could be
described without recourse to the % notation in the spec's productions.
Would it be helpful, for example, to increase the number of non-terminal
symbols in the grammar and then specify which non-terminal symbols can be
replaced by parameter entity reference?

PS Can people check out http://www.jtauber.com/xml/ and let me know
(off-list) what they think and what could be added?

Thanks

James

-- James K. Tauber / jtauber@jtauber.com Perth, Western Australia