DOCBOOK: Concrete proposal for #480954: Extend textobject to insertexternal files

[Norman Walsh <ndw@nwalsh.com>]

The content model of inlinemediaobject is currently:

inlinemediaobject ::=
(objectinfo?,
 (videoobject|audioobject|imageobject),
 (videoobject|audioobject|imageobject|textobject)*)

And the stylesheets support a hack whereby you can say

  <imageobject format="linespecific" fileref="somefile.txt"/>

and literally insert the content of somefile.txt into the result.

As this RFE suggests, it might be more reasonable to allow textobject
to directly insert a file by adding entityref and fileref attributes
to textobject.

If we do that, I suggest that we change the content model of
inlinemediaobject to:

inlinemediaobject ::=
(objectinfo?,
 textobject+|
 ((videoobject|audioobject|imageobject),
 (videoobject|audioobject|imageobject|textobject)*))

In other words, you would be able to have an inlinemediaobject with
one or more textobjects or one with a "proper" mediaobject and
alternatives, but you wouldn't be allowed to have a textobject and an
imageobject but put the textobject first. I think that preserves the
existing semantics.

This leaves open entirely the question of where textobject should be
allowed.  With this new semantic, it might make sense to allow
textobject directly in the verbatim environments and a few other
places. But we can leave that as a separate RFE, if we accept this
proposal, I think.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com>      | Clearness is so eminently one of
http://www.oasis-open.org/docbook/ | the characteristics of truth that
Chair, DocBook Technical Committee | often it even passes for truth
                                   | itself.--Joubert