Open kosloot opened 5 years ago
Yes, it is valid indeed. I agree it would be nicer when it's homogenous but I don't think we can/should enforce that in FoLiA itself, if tools like ucto/frog pose such extra constraints then that's fair enough I'd say.
Hmmm. imposing such implicit semantics might hit us hard in the future. But making homogeneity a strict prerequisite is maybe hard to implement. (can a DTD express this at all?) Maybe it is good to document this as 'good behavior' for FoLiA users. Or have the validator point it out? I will consider adding warnings (or even fatal errors) to the tools in my control.
see also issue #42
Given this FoliA example:
it is accepted by the validator and also folialint. There is maybe an issue here with the
<div>
having both a Sentence AND a Paragraph. This is valid FoLiA, but maybe it is against the 'gut feeling' that in this case the Sentence should be embedded in a Paragraph. Should this feeling be formalized? (an how) And if not, this has ramifications for a lot of FoLiA based software, liike Ucto, Frog, TICCL etc., that assumes OR sentences OR paragraphs with sentences.