When I retyped (and shorten) it I might have missed a parentheses, but yes
you can only have (A or P.a) and the S* is to allow nesting. And the
problem is that the P.a never shows up in the insert dialog box at the S
level. I hope that helps, I still need to try the newer version.
Ellen
"EXT-Thompson, Steve" <steve.thompson2@boeing.com> wrote on 01/22/2009
01:09:22 PM:
> All,
> My apologies if my reply sounded like a real newbie question, but it
> wasn't rhetorical. If that construct is somehow flawed, that could
> explain why one parser does one thing, and another does something
> else. It would depend on the 'strictness' of the parser.
>
> I ask again, what's the ", S*" intended to do?
>
> But, more to the point, why would you expect to see a P.a after an
> A? According to the structure, you have EITHER a single A or a
> single P.a, followed by any number of para, fig, etc. OR at least
> one and possibly more of para, fig, etc. To have both an A and a P.
> a, the first portion of the structure should be "(A | P.a)+" or "(A |
P.a)*".
>
> Excerpt of your element structure:
> ( ( A | P.a) , ( para | fig | table | caution | note | warning )*)
>
> As it is, the "," says that what is in those parentheses is followed
> by what is in the next parentheses. Since there's only one instance
> of the first parentheses (no "+" or "*"), you shouldn't be able to
> have one of either A or P.a followed by one of the other. If one of
> the parsers is allowing the P.a after an A, that parser is not
> following this part of the structure. The part of the structure not
> shown here, the "or" ("|") part doesn't include "A" or "P.a", so it
> doesn't affect your problem.
>
> Or did I still miss something?
> Steve Thompson
> +1(316)977-0515