Hi Jens,
You wrote:
> I have been looking forward to try-out you DocBook/WML, but just
> discovered you do not support Chapter/Appendix although you do
> support the
> Book element. I will have to see if I can write these templates.
>
> N.B. According to http://www.explain.com.au/oss/docbook/
> supported.html,
> both chapter and appendix are supported
I'll check that out. These should both be supported as both toplevel
and component elements.
BTW, I'd like to use the SourceForge bug/RFE tracking system for
these issues. I'll look into what is needed to set that up.
> By default you param.xsl set the wordml.template to
>
> <xsl:param name="wordml.template" select=""/>
>
> If you want to define the empty string, you should properly use
>
> <xsl:param name="wordml.template" select="''"/>
>
> instead, otherwise e.g. Saxon dies.
I can't see that problem. It is defined that way in params/
wordml.template.xml,
and my copy of param.xsl has:
<xsl:param name='wordml.template'/>
<xsl:param name='pages.template'/>
That should be OK.
> It was not immediately obvious, what value to use here, but it must
> be the
> following value, I should use
>
> <xsl:param name="wordml.template" select="'template.xml'"/>
I've tried to make that clear in the documentation on the Explain
website.
> Some of my documents contains XML processing instruction, which I
> would
> like to keep even if I transform them from DocBook to Word and back
> again.
> Is the a Word ML (cover it all element), just like the \special
> macro in TeX
No, there are no special WordML elements. You must use some feature
of Word to encode the information, and then translate it back again.
I'd suggest defining a paragraph style, say "xmlpi", and character
styles, say "xmlpi-target" and "xmlpi-data", and using those to
encode the values.
This brings up the issue of customisation, which I haven't really
addressed yet. You want to be able to interpose your own custom
handling of elements and styles on top of the processing provided by
the round-tripping system. I'll have a think about this over the
next couple of weeks.
> Most documents contains graphics in one form or another. Word will
> not for
> a forseable future (if ever) support my SVG figures, but I will
> properly
> have to enclose a kind of BASE64 raster versions of my images. Will
> that
> be possible with an implementation of a imagedata template, and
> will I be
> able to "reverse" my BASE64 spec/or reference to a raster image
> back to a
> reference to my svg file ?
Graphics are currently supported by linking to the external image
file. See imageobject-imagedata.
> Should the files specification.xml and supported.xml located in
> wordml/
> not be moved to docsrc/ and included in the document generation
> process ?
Possibly, I'll raise the issue with the other DocBook developers.
HTHs,
Steve Ball
