Simon Harvey wrote:
Could you or anyone else elaborate on that a bit more. I thought that xsl
was the technology needed to do conversions. Why can it do html but not pdf.
How does FO relate to xsl?
First of all let's make a difference between XSL and XSLT. As per W3C
recommendations, XSL = XSL Transformations (XSLT) + XSL-FO vocabulary.
XSLT is XML transformation language, with predefined XML, HTML and text output
formats. HTML case is simple - due to ubiquitous web browsers it's just simple
XML to HTML transformation, XML -> transformation -> HTML.
But because pdf is binary non-markup-based data format it's unfeasible to
generate pdf with XSLT (may be only simplest pdf documents). That's by design,
recall XSLT is *XML* transformation language.
So, they have created XSL-FO (another descendant of DSSSL, which was (and it
is) actually XPath+XSLT+XSL-FO all in one). The idea of XSL is to allow to
produce high-quality document presentation (usually printed one) by expressing
presentation semantics in XSL-FO and then formatting XSL-FO to actual
device-specific presentation, see [1]. So there is one more player in this
scenario - XSL-FO formatter, which formats (and sometimes renders) XSL-FO.
Usual scenario is XML -> transformation -> XSL-FO -> formatter -> pdf. But
that's not only scenario. Having GUI-based formatter like AntennaHouse XSL
formatter it can be XML -> transformation -> XSL-FO -> screen. Otherwise it's
possible to XML -> transformation -> XSL-FO -> formatter -> printer. Another
interesting new approach could be
XML -> transformation -> XSL-FO -> transformation -> WordML document (for
Microsoft Word 2003).
[1]
http://www.w3.org/TR/xsl/slice1.html...n-and-Overview
--
Oleg Tkachenko
http://www.tkachenko.com/blog
Multiconn Technologies, Israel