[rfc-i] draft-iab-xml2rfc-03, "2.56 <td>"

[julian.reschke at gmx.de (Julian Reschke)]

<https://greenbytes.de/tech/webdav/draft-iab-xml2rfc-03.html#rfc.section.2.56>

It seems to me that the content model for <td> and <th> is flawed; it 
allows to mix block content with phrasing content.

The grammar currently says:

>     (t+
>      | (text
>         | artwork
>         | bcp14
>         | br
>         | cref
>         | dl
>         | em
>         | eref
>         | figure
>         | iref
>         | ol
>         | relref
>         | sourcecode
>         | strong
>         | sub
>         | sup
>         | tt
>         | ul
>         | xref)+)

...so it special cases <t>, which is good. However, it allows mixing 
text content with non-<t> block level elements, such as lists.

We probably need something like:

     ( (t
         | artwork
         | dl
         | figure
         | ol
         | sourcecode
         | ul
       )+
      | (text
         | bcp14
         | br
         | cref
         | em
         | eref
         | iref
         | relref
         | strong
         | sub
         | sup
         | tt
         | xref)*)

instead.

(I note that the fact this wasn't reported earlier is that likely nobody 
has tried to implement it yet...)

Best regards, Julian