RE: [docbook-apps] Database for docbook components (i.e. section, para, etc) questions

["Nilsson, David F" <dfnilsson@switch.com>]

> -----Original Message-----
> From: Peter Desjardins [mailto:pdesjardins@supplyscape.com] 
> Sent: Friday, March 16, 2007 3:18 PM
> To: Nilsson, David F; Nelson, Dean; docbook-apps@lists.oasis-open.org
> Subject: RE: [docbook-apps] Database for docbook components (i.e. 
> section, para, etc) questions

> > Developing the stylesheet was not trivial, but it produces
> > solid Word
> > documents that require no fix-up. We are, however, just getting
> > started with this type of conversion.

> I wrote about using the DocBook > HTML > Word technique. One of
> the limitations of that route is that the paragraphs in Word are
> not tagged with paragraph styles. The content is formatted
> perfectly because Word imports the CSS that is applied to the
> HTML, however it's all done by Word with formatting overrides of
> "normal" or some other stock paragraph style.

> Does your solution solve this problem? Do the resulting Word
> documents use paragraph styles consistently?

Yes. The XSLT stylesheet uses its rules to apply styles.

For example, one of the templates in the stylesheet looks for <para>
elements within <entry> elements. If the enclosing <table> element has a
"role" attribute set to "small" and the <entry> element is within
thead/row, the stylesheet writes WordML markup to set the style on the
paragraph to "TableHeadingSmall." That WordML would look something like
this:

  <w:tc><w:tcPr><w:tcW w:w="3420" w:type="dxa"/>
  </w:tcPr>
  <w:p><w:pPr><w:pStyle w:val="TableHeadingSmall"/>
  </w:pPr>
  <w:r><w:t>Blah, blah, blah.</w:t>
  </w:r>
  </w:p>
  </w:tc>

Here's what some of these elements are:

  w:tc = table cell
  w:tcPr = table cell properties
  w:p = paragraph
  w:pPr = paragraph properties
  w:pStyle = paragraph style
  w:r = run
  w:t = text

WordML is rather unusual XML because it just represents the Word 
formatting, and you set the formatting with properties, which 
may be attached to paragraphs, runs, tables, and other objects.

> For example, using my technique the contents of a DocBook <entry>
> element in a table heading row are mapped to an HTML <th>
> element. Word picks this up and applies any CSS I specify but
> tags the content of the heading row with "Normal + Arial, Bold,
> Centered." If I were setting up a Word template, I'd create a
> paragraph style "TableHeading" and use it here. I wish I could
> convince Word to tag content with paragraph styles based on
> something useful. The class names that reflect the original
> DocBook elements would be nice.

> Peter Desjardins

It's nice to have clear style names, but if your documents are the final
output, I would think it might not matter. The style names are most
valuable when you're going to be manipulating the formatting in some
way.

Sometimes, though, we have had customers who have specified Word
templates with named styles. But usually style names aren't an issue.

The DocBook XSL stylesheet, as I recall, makes an effort to use the
DocBook element names in styles.

Dave Nilsson