Proposal 12046 (data element in figgroup)

[Robert D Anderson <robander@us.ibm.com>]

This is partly a question for Deborah Pickett, partly for anybody else,
related to the proposal to extend the content model of figgroup:
http://lists.oasis-open.org/archives/dita/200703/msg00087.html

I put my name on this one just so somebody could speak to it on the call,
though Deborah has already submitted the proposal. I was reading through it
today to prepare for tomorrow's call, and noticed that it suggests
expanding this model:
<!ELEMENT figgroup      ((%title;)?,
                         (%figgroup; | %xref; | %fn; | %ph; |
                          %keyword;)* )                      >
to this:
<!ELEMENT figgroup (title?, (figgroup | %basic.ph; | %fig.cnt;)* )>

The problem I see here is that both basic.ph and fig.cnt contain the xref
and fn elements. I tried this new model with different parsers -- two were
happy, but one (usually reliable) parser crashed with an "ambiguous content
model" error.

Deborah - do you have a suggestion on which of these three courses to
pursue?
1. If you can show me that the parser is in error, I will open a bug report
against it - but it seems from the XML spec that this is technically
forbidden. [1]
2. We revise the proposal to explicitly list all of basic.ph except <xref>
and <fn>
3. We revise the proposal to forget about adding more phrase-like elements,
and just add fig.cnt -- making the model
(figgroup | ph | keyword | fig.cnt)

[1]: http://www.w3.org/TR/REC-xml/#sec-element-content
"For compatibility, it is an error if the content model allows an element
to match more than one occurrence of an element type in the content model."

Thanks,

Robert D Anderson
IBM Authoring Tools Development
Chief Architect, DITA Open Toolkit
(507) 253-8787, T/L 553-8787 (Good Monday & Thursday)