Re: DL content model

Murray Maloney (murray@sco.COM)
Mon, 14 Nov 94 10:41:56 EST

>
> > > * change to DL content model from (DT*, DD?)+ to (DT | DD)*.
> > > OK, I guess.
> >
> > This doesn't match what was discussed at the Working group meeting
> > in Chicago. I thought we had agreed on:
> >
> > <!ELEMENT DL - - (DT*, DD)+>
>
> Doesn't this allow the following?
>
> <DL>
> <DD>Definition without preceding term</DD>
> </DL>

Yes, but it also allows this:

<DL>
<DT> Term 1
<DD> Definition 1a
<DD> Definition 1b
<DT> Term 2
<DT> Term 3
<DD> Definition 2 and 3
</DL>

The content model that actually describes many actual
definition lists is:

<!ELEMENT DL - - ( ( (DT+, DD) | (DT, DD+) )* )>

That is, I can have one or more terms followed by a definition,
and I can have a term followed by one or more definitions.

I was talked out of this content model, but it does describe what's needed.
I would accept as an alternative anything that allowed a mixture
of DT and DD elements.

>
> And if it does, wouldn't
>
> <!ELEMENT DL - - (DT+, DD)+>
>
> be more sensible? Or am I missing something?

Yes, you are missing the need to have multiple definitions
for a single term. And while Tim BL pointed out that
I could use a nested list to have multiple defn's, I am
not happy with that.
>
>
> Wkr
>
> /G
>
> --
> G|ran \berg University of Lule}, SWEDEN Student MSc CS
> K}rhusv{gen 5:601 http://www.luth.se/~goggi/ Adm./CoAdm. of
> S-976 54 Lule} goggi@dc.luth.se {www,wais,ftp,gopher}.luth.se
> ______________________________________________________________________________
>