Re: DOCBOOK: docbook vs latex

[Ramon Casellas <casellas at infres dot enst dot fr>]

On Tue, 3 Sep 2002, Doug du Boulay wrote:

>
> My understanding was that DocBook output was a product of the
> toolchain and not DocBook itself.  We should expect the tools
> to improve in due course. On the other hand the DocBook mathematical
> construct deficiencies could probably be addressed rather quickly once
> they were identified (and perhaps Ramon has already done that part).
>


Not very much. In fact, I have no experience with DTDs, but more or less
this is what I use (and to me, it has been good enough to write my
dissert. in networking/computer science. Of course, I would like more
experienced/knowledgeable people to give their opinions. Maybe if this is
made a proposal, Oasis could get in touch with AMS). I apologize for not
being formal and not give a proper content model.


<mathelement>
</mathelement> is the wrapper. I don't think it is really needed.


<mathhypothesis>
<mathproposition>
<mathlemma>
<mathremark>
<mathexample>
<mathdefinition>
<maththeorem>

are the "big" blocks. Personally, I think they should be at the
"formalpara" level.
Each "big" block has all common attributes (id, role, etc) and an optional
<title> and one or more of the following
<para>, <mathcondition>, <mathassertion>,...

<mathcondition> and <mathassertion> have 1+ <para>... etc.


The last "bit" is <mathproof>. IMVHO, I think it should be part of the big
blocks (à la proof of the theorem) but people using LaTeX tend to put it
after, that is
\begin{theorem}
\end{theorem}
\begin{proof}
\end{proof}



For example:


<mathelement>
	<mathhypothesis id="chapter:hyp:1">
	<title>Boredom</title>
	<para>All Tasks are boring.<para>
	</mathhypothesis>
<mathelement>


<mathelement>
	<maththeorem id="chapter:theorem:name">
		<title>Lazy Man</title>
		<mathcondition><para>Given a Task T</para></mathcondition>
		<mathassertion>
			<para>Do it tomorrow</para>
			<equation>
			....
			</equation>
		</mathassertion>
		<mathproof>
		<para>By <xref linkend="chapter:hyp:1"/> we can start by
proving the lazy nature of mankind...</para>
		</mathproof>
	</maththeorem>
</mathelement>


There is another point regarding the boolean expression of conditions,
but I usually put just one single composed condition (if and b or c)

Of course, bear in mind that I used them "as is" for my own purposes.


in mathelem.mod.xsl (DB2LaTeX) most templates are of the form

 <!--
 ##########################################
 #
 #  \begin{thm}[title]
 #
 #  \end{thm}
 #
 ##########################################
 -->
 <xsl:template match="mathelement/maththeorem">
     <xsl:text>\begin{thm}[</xsl:text>
     <xsl:call-template name="normalize-scape">
         <xsl:with-param name="string" select="title"/>
     </xsl:call-template>
     <xsl:text>]&#10;</xsl:text>
     <xsl:variable name="id"> <xsl:call-template name="label.id"/>
</xsl:variable>
     <xsl:apply-templates/>
     <xsl:text>\end{thm}&#10;</xsl:text>
 </xsl:template>


time for "café"
Thanks for reading

Ramon