<?xml version="1.0"?>
<slideshow>
<title>AxKit</title>
<metadata>
<speaker>Matt Sergeant</speaker>
<email>matt@axkit.com</email>
<organisation>AxKit.com Ltd</organisation>
<link>http://axkit.com/</link>
<logo scale="0.4">ax_logo.png</logo>
<background>redbg.png</background>
</metadata>
<slide>
<title>Introduction</title>
<point level="1">Perl's XML Capabilities</point>
<point level="1">AxKit intro</point>
<point level="1">AxKit static sites</point>
<point level="1">AxKit dynamic sites (XSP)</point>
<point level="1">Advanced AxKit</point>
</slide>
<slideset>
<title>XML with Perl Introduction</title>
<slide>
<title>Basis : XML::Parser</title>
<point level="1">XS wrapper around expat</point>
<point level="1">expat - James Clark</point>
<point level="1">XML::Parser - originally by Larry Wall</point>
<point level="2">Now maintained by Clark Cooper</point>
<point level="1">Allows different "styles" of parsing</point>
<point level="1">Default style is callback/event based stream parsing</point>
<point level="1">Also implements a "push" parser</point>
</slide>
<slide>
<title>XML::Parser usage</title>
<point level="1">SAX-like API</point>
<point level="1">register callback handler methods</point>
<point level="2">start tag</point>
<point level="2">end tag</point>
<point level="2">characters</point>
<point level="2">comments</point>
<point level="2">processing instructions</point>
<point level="2">... and more</point>
<point level="1">Non validating XML parser</point>
<point level="1">dies (throws an exception) on bad XML</point>
</slide>
<slide>
<title>XML::Parser code</title>
<source_code><![CDATA[
my $p = XML::Parser->new(
Handlers => {
Start => \&start_tag,
End => \&end_tag,
# add more handlers here
});
$p->parsefile("foo.xml");
exit(0);
sub start_tag {
my ($expat, $tag, %attribs) = @_;
print "Start tag: $tag\n";
}
sub end_tag {
my ($expat, $tag) = @_;
print "End tag: $tag\n";
}
]]></source_code>
</slide>
<slide>
<title>XML::Parser - maintaining state</title>
<source_code><![CDATA[
my $ctxt = MyParser->new();
my $p = XML::Parser->new(
Handlers => {
Start => sub { $ctxt->parse_start(@_) },
End => sub { $ctxt->parse_end(@_) }
});
$p->parse("<some><xml>here</xml></some>");
package MyParser;
sub new {
return bless {}, shift;
}
sub parse_start {
my ($self, $expat, $tag, %attribs) = @_;
$self->{some_state} = "Open $tag";
}
sub parse_end {
my ($self, $expat, $tag) = @_;
print "Last state: $self->{some_state}\n";
}
]]></source_code>
</slide>
<slide>
<title>PYX</title>
<point level="1">Turns XML into a stream of text events</point>
<point level="1">Based on SGML ESIS streams</point>
<point level="2">)start</point>
<point level="2">(end</point>
<point level="2">Aattribute=value</point>
<point level="2">-text</point>
<point level="2">?processing instruction</point>
<point level="1">Comments are dropped</point>
</slide>
<slide>
<title>PYX example</title>
<source_code>
$ pyx presentation.xml | more
(slideshow
-\n
-
(title
-Developing XML Applications with Perl and AxKit
)title
-\n
-
(metadata
-\n
-
(speaker
-Matt Sergeant
)speaker
-\n
-
(email
-matt@axkit.com
)email
...
</source_code>
</slide>
<slide>
<title>PYX usage</title>
<point level="1">pyx xmlfile | something | pyxw</point>
<point level="1">pyx generates PYX streams</point>
<point level="1">pyxw generates XML from PYX</point>
<point level="1">pyxhtml parses HTML to PYX</point>
</slide>
<slide>
<title>HTML Tidy, or html2xml</title>
<point level="1">Java utility by Dave Ragget of the W3C</point>
<point level="1">Tidy converts HTML to XHTML</point>
<point level="1">PYX can do the same:</point>
<point level="2">pyxhtml file.html | pyxw > file.xhtml</point>
<point level="1">pyxhtml uses HTML::TreeBuilder</point>
</slide>
<slide>
<title>XPath intro</title>
<point level="1">W3C Standard for locating nodes within an XML document</point>
<point level="1">A subset of XPath is used for "matching" nodes in XSLT</point>
<point level="1">Looks like directory paths: /path/to/node</point>
<point level="1">But that's an abbreviated syntax...</point>
<point level="2">/child::path/child::to/child::node</point>
<point level="1">Full grammar containing expressions, calculations, functions, etc</point>
</slide>
<slide>
<title>XPath Examples</title>
<point level="1">Find title of XHTML document</point>
<source_code>/html/head/title</source_code>
<point level="1">Find all hrefs in <a> tags</point>
<source_code>/html/body/descendant::a[@href]/@href</source_code>
</slide>
<slide>
<title>XML::XPath</title>
<point level="1">Full implementation of W3C XPath on a DOM-like API</point>
<point level="1">Easy to use:</point>
<source_code><![CDATA[
my $xp = XML::XPath->new(filename => "foo.xml");
print $xp->findvalue("/html/head/title");
]]></source_code>
<point level="1">Can also process HTML:</point>
<source_code><![CDATA[
my $xp = XML::XPath->new(
filename => "pyxhtml foo.html | pyxw |");
]]></source_code>
</slide>
<slide>
<title>XML::XPath continued</title>
<point level="1">Every node has methods findvalue(), findnodes() and find()</point>
<source_code><![CDATA[
my $xp = XML::XPath->new(filename => "foo.xml");
foreach my $link ($xp->findnodes('/html/body/a[@href]')) {
print "Link: ", $link->findvalue('@href'), "\n";
# or :
# print "Link: ", $link->getAttribute('href'), "\n";
}
]]></source_code>
<point level="1">Not 100% DOM compatible, but close</point>
</slide>
<slide>
<title>XML::XPath Implementation</title>
<point level="1">XML::Parser and SAX parsers build an in-memory tree</point>
<point level="1">Hand-built parser for XPath syntax (rather than YACC based parser)</point>
<point level="1">Garbage Collection yet still has circular references (and works on Perl 5.005)</point>
<image>pointers.png</image>
</slide>
<slide>
<title>XML to XML Conversions</title>
<point level="1">Use XSLT (or XPathScript, see later)</point>
<point level="1">Either use the command line (e.g. xsltproc from libxslt)</point>
<point level="1">... or call from Perl:</point>
<source_code><![CDATA[
use XML::LibXSLT;
use XML::LibXML;
my $parser = XML::LibXML->new();
my $xslt = XML::LibXSLT->new();
my $source_dom = $parser->parsefile('foo.xml');
my $style_dom = $parser->parsefile('bar.xsl');
my $stylesheet = $xslt->parse_stylesheet($style_dom);
my $results = $stylesheet->transform($source_dom);
print $stylesheet->output_string($results);
]]></source_code>
</slide>
<slide>
<title>SAX Filters</title>
<point level="1">SAX events passed between SAX handlers</point>
<point level="1">e.g. tags to lower case:</point>
<source_code><![CDATA[
my $ya = XML::Handler::YAWriter->new( AsFile => '-' );
my $lc = XML::Filter::ToLower->new( Handler => $ya );
my $parser = XML::Parser::PerlSAX->new( Handler => $lc );
package XML::Filter::ToLower;
@ISA = ('XML::Filter::Base');
sub start_element {
my ($self, $element) = @_;
$element->{Name} = lc($element->{Name});
$self->{Handler}->start_element($element);
}
sub end_element {
my ($self, $element) = @_;
$element->{Name} = lc($element->{Name});
$self->{Handler}->end_element($element);
}
]]></source_code>
</slide>
<slide>
<title>SAX Filters (cont)</title>
<point>Almost any level of complexity possible</point>
<point>Chain as many filters together as you like</point>
<point>Fast way of processing XML</point>
<point level="2">We don't have to build a tree in memory</point>
<point>Possible to output to HTML using XML::Handler::HTMLWriter</point>
</slide>
</slideset>
<slideset>
<title>AxKit Introduction</title>
<slide>
<title>AxKit Introduction</title>
<point level="1">XML Publishing Framework</point>
<point level="1">XML Application Server</point>
<point level="1">Designed for Content Diversification</point>
<point level="2">Same content, delivered differently</point>
<point level="2">HTML, WAP, PDF, SOAP, etc</point>
<point level="1">Built in C, Perl, and mod_perl/Apache</point>
<point level="1">Plugin Architecture, allows operation to be changed easily</point>
</slide>
<slide>
<title>How does it work?</title>
<point level="1">Apache handle</point>
<point level="1">Embedded perl interpreter via mod_perl</point>
<point level="1">Apache Config directives</point>
<point level="1">Transformation Pipeline</point>
<point level="1">Language Modules implement stylesheets and other engines</point>
<point level="1">XSP for Dynamic Content</point>
<point level="1">Caching</point>
<point level="1">Apache::Filter Support</point>
</slide>
<slide>
<title>mod_perl</title>
<point level="1">mod_perl used to implement most of AxKit</point>
<point level="1">... though more code being ported to C all the time</point>
<point level="1">Gives us bytecode compiled Perl</point>
<point level="1">Makes development faster</point>
</slide>
<slide>
<title>Configuration Directives</title>
<point level="1">Config directives written in the C code</point>
<point level="1">Implemented on top of Apache</point>
<point level="1">Fits in with httpd.conf and .htaccess files</point>
<point level="1">28 Config directives in total</point>
<point level="1">Setup the pipeline, debugging options, and various other flags</point>
</slide>
<slide>
<title>Config Directives examples</title>
<source_code><![CDATA[
AxDebugLevel 5
AxCacheDir /tmp/axkit_cache
<Location /apps/calendar>
AxAddProcessor application/x-xsp .
<AxMediaType handheld>
AxAddProcessor text/xsl /styles/cal2wml.xsl
</AxMediaType>
<AxMediaType screen>
AxAddProcessor text/xsl /styles/cal2html.xsl
</AxMediaType>
</Location>
]]></source_code>
</slide>
<slide>
<title>Transformation Pipeline</title>
<point level="1">AxKit is a pipeline engine</point>
<point level="1">Output of one stage goes to input of the next</point>
<point level="1">Allows us to build up our application in stages</point>
<point level="1">Pipeline passed as one of:</point>
<point level="2">String</point>
<point level="2">DOM Tree</point>
<point level="2">SAX events</point>
</slide>
<slide>
<title>Language Modules</title>
<point level="1">Every stage in the pipeline implemented by a language module</point>
<point level="1">Language modules can either use a stylesheet for transformation or not</point>
<point level="2">Examples of modules not using a stylesheet are "XSL:FO to PDF", XSP, and the module which produces this slideshow</point>
<point level="1">Different implementations of same language possible, e.g. XSLT</point>
<point level="2">XML::LibXSLT</point>
<point level="2">XML::Sablotron</point>
<point level="2">XML::XSLT</point>
<point level="2">XML::Xalan</point>
<point level="2">XML::Transformiix</point>
</slide>
<slide>
<title>Languages modules available</title>
<point level="1">XSLT</point>
<point level="1">XPathScript</point>
<point level="1">XSP</point>
<point level="1">AxPoint (this slideshow)</point>
<point level="1">XMLNews::HTMLTemplate</point>
<point level="1">Template Toolkit w/XPath plugin</point>
</slide>
<slide>
<title>Caching</title>
<point level="1">XML Transformation can be slow, caching vital to making a fast XML Publishing Engine</point>
<point level="1">Cache uses the filesystem, not memory or DBM files</point>
<point level="1">Cache results of transformation pipeline, where appropriate</point>
<point level="1">Cache results of dynamic processing by implementing has_changed() method</point>
<point level="1">Using filesystem allows Apache to deliver the cache directly:</point>
<source_code>Apache->request->filename = <cache-file>;
return DECLINED;
</source_code>
</slide>
<slide>
<title>Cache (cont.)</title>
<point level="1">Fastest possible implementation of a cache</point>
<point level="1">Default storage place is .xmlstyle_cache directory in same directory as XML file</point>
<point level="1">Overridable using AxCacheDir directive</point>
<point level="1">Files named as MD5 hash of XML filename, media type, style name and optionally run-time parameters that may affect the cache</point>
<point level="1">Does use large amounts of space, so can be turned off</point>
</slide>
<slide>
<title>Non Cache-able content</title>
<point level="1">POST requests usually not cacheable</point>
<point level="1">XSP pages usually not cacheable</point>
<point level="1">Can turn off the cache in Perl code:</point>
<source_code>$r->no_cache(1);</source_code>
<point level="1">Cache can be modified using plugins, so cache can be based on phase of the moon</point>
</slide>
<slide>
<title>Apache::Filter Support</title>
<point level="1">Alternate plugin "Provider" module</point>
<point level="1">AxKit can get XML from previous mod_perl handler</point>
<point level="1">Works with Apache::ASP, HTML::Mason, etc</point>
</slide>
<slide>
<title>Apache::Filter example</title>
<point level="1">ASP Script</point>
<source_code><![CDATA[
<?xml version="1.0"?>
<?xml-stylesheet href="/foo.xsl" type="text/xsl"?>
<page>
<head><title>Hello World</title></head>
<body>
<section>
<title>Test Page</title>
<para>
Hello World
</para>
<para>
The time is now: <%= localtime %>
</para>
</body>
</page>
]]></source_code>
</slide>
<slide>
<title>Apache::Filter example (stylesheet)</title>
<source_code><![CDATA[
<xsl:stylesheet version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:template match="/">
<html>
<head><title><xsl:value-of select="/page/head/title"/> </title></head>
<body>
<xsl:apply-templates/>
</body>
</html>
</xsl:template>
<xsl:template match="section">
<h1><xsl:value-of select="title"/></h1>
<xsl:apply-templates select="para"/>
</xsl:template>
<xsl:template match="para">
<p><xsl:apply-templates/></p>
</xsl:template>
</xsl:stylesheet>
]]></source_code>
</slide>
<slide>
<title>Apache::Filter example (results)</title>
<source_code><![CDATA[
<html>
<head><title>Hello World</title></head>
<body>
<h1>Test Page</h1>
<p>
Hello World
</p>
<p>
The time is now: Tue May 8 07:41:33 2001
</p>
</body>
</html>
]]></source_code>
</slide>
<slide>
<title>Gzipped Output</title>
<point level="1">Many HTTP/1.1 browsers will accept gzip encoded results (modulo bugs)</point>
<point level="1">AxKit can check Accept-Encoding header for this capability</point>
<point level="1">Will then gzip content before delivery</point>
<point level="1">Cache stores regular and gzipped copy</point>
<point level="1">Reduces used bandwidth by up to 85%</point>
</slide>
<slide>
<title>Developer's Perspective</title>
<point level="1">Two ways to tell AxKit how to process XML</point>
<point level="2">Processing Instructions in the XML</point>
<source_code><![CDATA[
<?xml-stylesheet href="/styles/foo.xsl" type="text/xsl"?>
]]></source_code>
<point level="2">Configuration Directives in httpd.conf</point>
<source_code><![CDATA[
AxAddProcessor text/xsl /styles/foo.xsl
]]></source_code>
<point level="1">PIs act as override mechanism</point>
<point level="1">Also PIs are sometimes appropriate, e.g. for XSP files</point>
</slide>
<slide>
<title>Processing Instructions</title>
<point level="1">Follows the W3C xml-stylesheet REC</point>
<point level="1">Works almost identically to HTML's LINK element</point>
<point level="1">Processing instruction is just <?target data?>, so attributes are called "psuedo attributes"</point>
<point level="1">Allowed attributes:</point>
<point level="2">href - the location of the stylesheet</point>
<point level="2">type - the mime type of the stylesheet</point>
<point level="2">alternate - yes or no, specifies alternate styling</point>
<point level="2">media - media type this stylesheet applies to</point>
<point level="2">title - group this stylesheet belongs to</point>
<point level="1">Only href and type are required</point>
</slide>
<slide>
<title>Config Directives</title>
<point level="1">More complex and more powerful than PIs</point>
<source_code><![CDATA[
AxAddProcessor text/xsl /styles/foo.xsl
AxAddRootProcessor text/xsl /styles/article.xsl article
AxAddDocTypeProcessor text/xsl /styles/docbook.xsl \
"-//OASIS//DTD DocBook XML V4.1.2//EN"
AxAddDTDProcessor text/xsl /styles/bar.xsl /dtds/bar.dtd
AxAddURIProcessor text/xsl /styles/book.xsl \
"book.*\.xml$"
]]></source_code>
<point level="1">Fully documented in perldoc AxKit</point>
</slide>
<slide>
<title>AxKit vs Cocoon</title>
<point level="1">Cocoon is a Servlet implementing some of the same functionality</point>
<point level="1">AxKit is tied to Apache</point>
<point level="1">Cocoon 2 [alpha] adds features like AxKit config directives</point>
<point level="1">Both implement XSP, and collaborate on the spec</point>
<point level="1">Both open source</point>
<point level="1">Perl and C vs Java</point>
<point level="1">Use what fits your environment</point>
</slide>
</slideset>
<slideset>
<title>AxKit for Static Content</title>
<slide>
<title>XML Publishing</title>
<point level="1">Using XML for the source of all our web site's content</point>
<point level="1">Semantically rich original data</point>
<point level="1">Delivery to different media types</point>
<point level="2">Example - slides</point>
<point level="3">Deliver to HTML</point>
<point level="3">Or PDF</point>
<point level="3">Or plain text</point>
<point level="2">AxKit can detect via plugins, which version the user wants</point>
</slide>
<slide>
<title>XML Based Content</title>
<point level="1">DocBook</point>
<point level="1">XHTML</point>
<point level="1">TEI XML</point>
<point level="1">RSS</point>
<point level="1">MyML - invent your own!</point>
<point level="1">Or use Provider interface...</point>
<point level="2">OpenOffice</point>
<point level="2">MS Word (via libwv)</point>
<point level="2">Content Management System</point>
<point level="1">Or use other mod_perl modules (Mason, ASP, etc) to generate XML</point>
</slide>
<slide>
<title>The Value of Semantics</title>
<point level="1">HTML has low semantic value</point>
<point level="1">Example point - farming the web</point>
<point level="1">XML provides the ability to add value to our content</point>
<point level="1">But we still need to deliver it as HTML</point>
<point level="1">AxKit can give you the best of both worlds</point>
</slide>
<slide>
<title>Alternative Media</title>
<point level="1">The web is no longer just web browsers on PCs</point>
<point level="1">TV is becoming huge, Phone/Handheld is already huge</point>
<point level="1">But the cost of entry into different media types is still high</point>
<point level="1">Using XML for our content reduces that cost of entry</point>
<point level="2">And it will reduce your long term costs</point>
</slide>
<slide>
<title>Media vs Styles</title>
<point level="1">Same concept, different reasons</point>
<point level="1">Use different styles for the same media device</point>
<point level="1">e.g. No graphics HTML version, printable version,
PDF version. All delivered to browsers on PCs</point>
<point level="1">Some applications in personalisation</point>
</slide>
<slide>
<title>The Key - Stylesheets</title>
<point level="1">Stylesheets translate XML on the fly to different formats</point>
<point level="1">Different stylesheets used for different media types</point>
<point level="1">... and for different styles of the same media type</point>
<point level="1">Generally, content is cached for increased performance</point>
</slide>
<slide>
<title>What is the benefit to you?</title>
<point level="1">Web Developers benefit because of separation of concerns</point>
<point level="1">Content</point>
<point level="1">Presentation</point>
<point level="1">Logic</point>
<point level="1">Site Management</point>
</slide>
<slide>
<title>Development Scalability</title>
<point level="1">Long term site management is a mess without separation of concerns</point>
<point level="1">Templates vs print print print</point>
<point level="2">But we can do better</point>
<point level="1">Different devices throws a spanner in the works</point>
<point level="1">Ideal is for content creators to focus on content, developers to focus on logic, designers to focus on presentation, and webmasters to focus on site management</point>
<point level="1">Unfortunately this is still idealistic, but much better than print print print</point>
</slide>
<slide>
<title>Content Authoring</title>
<point level="1">XML Editors</point>
<point level="1">WYSIWYM Editors (What You See Is What You Mean)</point>
<point level="1">foo2xml can also be used (e.g. pod2xml)</point>
<point level="1">Mostly commercial offerings: Adept, XMeTaL, XMLSpy</point>
<point level="1">Open source options: OpenOffice, Emacs+psgml, jedit</point>
<point level="1">Some re-training may be required</point>
</slide>
<slide>
<title>Presentation Layer</title>
<point level="1">Currently the hardest part of converting from static HTML sites</point>
<point level="1">Some commercial tools available for writing XSLT stylesheets</point>
<point level="1">Sometimes templates generated using HTML tools (e.g. Frontpage) possible</point>
<point level="1">Often best way is to get a design, then have a stylesheet guru convert that</point>
</slide>
<slide>
<title>Business Logic</title>
<point level="1">AxKit uses taglib concept</point>
<point level="1">Logic wrapped in XML tags</point>
<source_code><order_fluglebar quantity="1"/></source_code>
<point level="1">Designers can drop these tags into their XML</point>
<point level="1">Logic tags output XML, not HTML</point>
<point level="1">Thus logic can be used for any output media/style</point>
<point level="1">See next section (Dynamic Content)</point>
</slide>
<slide>
<title>Site Management</title>
<point level="1">AxKit uses Apache config directives in httpd.conf/.htaccess</point>
<point level="1">Allows you to use your web master skills</point>
<point level="1">Also allows flexible configuration based on <Location>, <Files>, and <Directory> directives</point>
</slide>
<slide>
<title>XSLT Overview</title>
<point level="1">Reformation of transformation part of DSSSL into XML (from Scheme)</point>
<point level="1">Functional language (no side effects)</point>
<point level="1">Rules based ("apply this template when we see this element")</point>
<point level="1">Implemented in AxKit via:</point>
<point level="2">XML::LibXSLT (very fast compliant processor from Red Hat Labs</point>
<point level="2">XML::Sablotron (lightweight processor from Ginger Alliance</point>
<point level="2">Others: XML::Xalan, XML::XSLT, XML::Transformiix</point>
</slide>
<slide>
<title>XPathScript Overview</title>
<point level="1">Combination of:</point>
<point level="2">Perl</point>
<point level="2">ASP <% %> delimiters</point>
<point level="2">XML::XPath for locating nodes in the source XML</point>
<point level="2">Declarative (Rules based) processing</point>
<point level="1">Not a functional language</point>
<point level="1">Side effects allowed</point>
<point level="1">Full access to Apache API</point>
</slide>
<slide>
<title>XPathScript example (xml)</title>
<source_code><![CDATA[
<employees>
<employee>
<name>
<firstname>Roger</firstname>
<lastname>Rabbit</lastname>
</name>
<department>Humour</department>
</employee>
<employee>
<name>
<firstname>Jessica</firstname>
<lastname>Rabbit</lastname>
</name>
<department>Modelling</department>
</employee>
</employees>
]]></source_code>
</slide>
<slide>
<title>XPathScript example (stylesheet)</title>
<source_code><![CDATA[
<html>
<head><title>Employee List</title></head>
<body>
<h1>Employees at Acme Corp.</h1>
<% foreach my $employee (findnodes('/employees/employee')) { %>
<b><%= $employee->findvalue("name/lastname") %>,
<%= $employee->findvalue("name/firstname") %></b>
works in the <%= $employee->findvalue('department') %>
department.
<hr>
<% } %>
</body>
</html>
]]></source_code>
</slide>
<slide>
<title>XPathScript example (results)</title>
<source_code><![CDATA[
<html>
<head><title>Employee List</title></head>
<body>
<h1>Employees at Acme Corp.</h1>
<b>Rabbit, Roger</b> works in the Humour department
<hr>
<b>Rabbit, Jessica</b> works in the Modelling department
<hr>
</body>
</html>
]]></source_code>
</slide>
<slide>
<title>XPathScript - Processing Documents</title>
<point level="1">Transforming Data is very different to transforming Documents</point>
<point level="1">Documents have mixed content</point>
<point level="1">Needs rules based processing</point>
<point level="1">XPathScript implements a feature rich declarative processing system</point>
</slide>
<slide>
<title>XPathScript - Declarative Templates</title>
<point level="1">The $t hash reference</point>
<source_code><![CDATA[
$t->{'para'}{pre} = '<p>';
$t->{'para'}{post} = '</p>';
]]></source_code>
<point level="1">Matches element names (unlike XSLT)</point>
<point level="1">XPathScript applies the rules as it traverses the document tree</point>
</slide>
<slide>
<title>XPathScript - $t subkeys</title>
<point level="1">Sub-keys of $t specify what to do</point>
<point level="1">Keys who's value is something to add to the ouput:</point>
<point level="2">pre</point>
<point level="2">post</point>
<point level="2">prechildren</point>
<point level="2">postchildren</point>
<point level="2">prechild</point>
<point level="2">postchild</point>
<point level="1">Other keys:</point>
<point level="2">showtag</point>
<point level="2">testcode</point>
</slide>
<slide>
<title>XPathScript - testcode</title>
<point level="1">testcode value is a code ref</point>
<point level="1">Executed every time an element with that name is visited</point>
<point level="1">Gives access to XML::XPath API on that node, and a local copy of $t</point>
<source_code><![CDATA[
$t->{'a'}{testcode} = sub {
my ($node, $t2) = @_;
if ($node->findvalue('@name')) {
# process as anchor
}
else {
# process as link
}
return DO_SELF_AND_KIDS;
};
]]></source_code>
</slide>
<slide>
<title>XPathScript tech details</title>
<point level="1">Script is compiled into Perl code</point>
<point level="1">Compiled once on first hit. Only re-compiled if changed</point>
<point level="1">Imports some functions: apply_templates(), findnodes(), findvalue()</point>
<point level="1">Fast enough for most real-time transformation needs (e.g. for dynamic content</point>
</slide>
<slide>
<title>Complete Example</title>
<point level="1">Take23.org - mod_perl news and resources</point>
<point level="1">Uses RSS 1.0 for all headlines and syndication</point>
<point level="1">Static Content, but looks dynamic</point>
<point level="1">Accepts content for articles in:</point>
<point level="2">XHTML</point>
<point level="2">DocBook</point>
<point level="2">POD (via pod2xml)</point>
<point level="2">OpenOffice 614 XML</point>
</slide>
<slide>
<title>Take23 - how it works</title>
<point level="1">httpd.conf contains AxAddRootProcessor directives</point>
<point level="1">Stylesheets written in XPathScript</point>
<point level="1">Stylesheets componentized to maximize reuse</point>
<point level="1">Uses OpenOffice provider module from AxKit.com to read OpenOffice files as XML</point>
<point level="1">News items created using AxKit::NewsMaker, a lightweight RSS/News content management system</point>
<point level="1">Cached content delivered at static HTML speeds</point>
</slide>
</slideset>
<slideset>
<title>AxKit for Dynamic Content</title>
<slide>
<title>XSP is the Key</title>
<point level="1">XSP is an XML language invented by the Cocoon Project at Apache</point>
<point level="1">XSP is implementation language agnostic</point>
<point level="2">Cocoon implements using Java</point>
<point level="2">AxKit implements using Perl</point>
<point level="1">XSP allows you to embed code in XML</point>
<point level="1">XSP is a page-based language like PHP and ASP</point>
<point level="1">... but XSP is better because you generate XML, maintaining separation of concerns</point>
</slide>
<slide>
<title>XSP Example</title>
<source_code><![CDATA[
<xsp:page xmlns:xsp="http://www.apache.org/1999/XSP/Core"
language="Perl">
<time>
<xsp:expr>localtime(time)</xsp:expr>
</time>
</xsp:page>
]]></source_code>
<point level="1">Which generates:</point>
<source_code><![CDATA[
<time>
Tue May 8 10:20:03 2001
</time>
]]></source_code>
<point level="1">Note: we then transform to HTML in the next part of the pipeline</point>
</slide>
<slide>
<title>XSP Lowdown</title>
<point level="1">XSP implements 11 core tags</point>
<point level="2">page</point>
<point level="2">structure</point>
<point level="2">import</point>
<point level="2">content</point>
<point level="2">logic</point>
<point level="2">element</point>
<point level="2">attribute</point>
<point level="2">pi</point>
<point level="2">comment</point>
<point level="2">text</point>
<point level="2">expr</point>
<point level="1">(plus a couple of other minor/supporting tags)</point>
</slide>
<slide>
<title>XSP and Namespaces</title>
<point level="1">XML Namespaces are key to how XSP works</point>
<source_code>
<prefix:tag xmlns:prefix="URI">
</source_code>
<point level="1">Each namespace can either implement custom functionality</point>
<point level="1">... or just be part of the output</point>
<point level="1">We call the namespaces "Tag Libraries", or taglibs for short</point>
</slide>
<slide>
<title>More XSP Examples</title>
<source_code><![CDATA[
<xsp:page xmlns:xsp="http://www.apache.org/1999/XSP/Core"
language="Perl">
<xsp:structure>
<xsp:import>Time::Piece</xsp:import>
</xsp:structure>
<xsp:logic>
sub afternoon {
return localtime->hour > 12;
}
</xsp:logic>
<page>
<xsp:logic>
if (afternoon()) {
<xsp:content>
Good <em>Afternoon!!!</em>
</xsp:content>
}
else {
<xsp:content>
Good <em>Morning!!!</em>
</xsp:content>
}
</xsp:logic>
</page>
</xsp:page>
]]></source_code>
</slide>
<slide>
<title>XSP Implementation</title>
<point level="1">SAX Parser, callbacks generate Perl code</point>
<point level="1">Perl compiled into a unique class</point>
<point level="1">handler() method wraps non-XSP top level tags (<page> tag in previous example)</point>
<point level="1">xsp:logic outside of top level tags can generate local functions</point>
<point level="2">mod_perl developers will know how handy this is to avoid closure problems</point>
<point level="1">Parser code implements "namespace dispatch"</point>
<point level="1">taglib implementations register with the core XSP parser engine's namespace dispatch system</point>
</slide>
<slide>
<title>XSP Taglibs</title>
<point level="1">Taglibs implement custom functionality</point>
<point level="1">Usually this is business logic</point>
<point level="1">Some low level taglibs exist on CPAN:</point>
<point level="2">Param</point>
<point level="2">Utils</point>
<point level="2">Cookie</point>
<point level="2">ParamAttrib</point>
<point level="2">SendMail</point>
<point level="2">Exceptions</point>
<point level="2">ESQL</point>
</slide>
<slide>
<title>Implementing Taglibs</title>
<point level="1">3 ways to implement taglibs (in decreasing levels of difficulty)</point>
<point level="1">1. Write a Stylesheet</point>
<point level="2">this transforms the taglib tags into core XSP tags</point>
<point level="1">2. Write a SAX parser</point>
<point level="2">here the parser has to write perl code</point>
<point level="1">3. Use the AxKit TaglibHelper module</point>
<point level="2">here you just write a Perl module, naming the functions after the tags you want to implement</point>
<point level="1">Most of the taglibs on CPAN use method 2 at the moment, simply because of when TaglibHelper appeared</point>
</slide>
<slide>
<title>Param Taglib</title>
<source_code><![CDATA[
<xsp:page
xmlns:xsp="http://apache.org/xsp/core/v1"
xmlns:param="http://axkit.org/NS/xsp/param/v1"
language="Perl"
>
<page>
<xsp:logic>
if (<param:name/>) {
<xsp:content>
Your name is: <param:name/>
</xsp:content>
}
else {
<xsp:content>
<form>
Enter your name: <input type="text" name="name" />
<input type="submit"/>
</form>
</xsp:content>
}
</xsp:logic>
</page>
</xsp:page>
]]></source_code>
</slide>
<slide>
<title>Things to note...</title>
<point level="1">Freely mixed Perl and XML</point>
<point level="2">No <% ... %> to introduce one or the other</point>
<point level="2">An XSP page is pure XML</point>
<point level="1">That has consequences:</point>
<source_code><![CDATA[
<xsp:logic>
my $page = <param:page/>
if ($page < 3) { # INVALID CODE. < not allowed in XML
...
}
</xsp:logic>
]]></source_code>
<point level="1">Could iritate some people</point>
<point level="1">Easy to fix this case. Use 3 > $page</point>
</slide>
<slide>
<title>Cookie Taglib</title>
<source_code><![CDATA[
<xsp:page
xmlns:xsp="http://apache.org/xsp/core/v1"
xmlns:cookie="http://axkit.org/NS/xsp/cookie/v1"
language="Perl"
>
<page>
<xsp:logic>
my $value;
if ($value = <cookie:fetch name="count"/>) {
$value++;
}
else {
$value = 1;
}
</xsp:logic>
<cookie:create name="count">
<cookie:value><xsp:expr>$value</xsp:expr></cookie:value>
</cookie:create>
<p>Cookie value: <xsp:expr>$value</xsp:expr></p>
</page>
</xsp:page>
]]></source_code>
</slide>
<slide>
<title>More things to note</title>
<point level="1">XSP extends Perl's notion of DWIM (Do What I Mean)</point>
<point level="1">Note how the cookie:value is taken from a Perl expression</point>
<point level="1">It could equally have been hard coded. Witness these three versions:</point>
<source_code><![CDATA[
<cookie:value>3</cookie:value>
<cookie:value><xsp:expr>2 + 1</xsp:expr></cookie:value>
<cookie:value><param:cookie_value/></cookie:value>
]]></source_code>
<point level="1">All produce valid code</point>
</slide>
<slide>
<title>SendMail Taglib</title>
<source_code><![CDATA[
<xsp:page (namespace stuff removed for space issues)>
<page>
<xsp:logic>
if (!<param:email/>) {
<p>You forgot to supply an email address!</p>
}
else {
if (<param:subopt/> eq "sub") {
$to = "axkit-users-subscribe@axkit.org";
}
elsif (<param:subopt/> eq "unsub") {
$to = "axkit-users-unsubscribe@axkit.org";
}
<mail:send-mail>
<mail:from><param:user_email/></mail:from>
<mail:to><xsp:expr>$to</xsp:expr></mail:to>
<mail:body>Subscribe or Unsubscribe <param:user_email/></mail:body>
</mail:send-mail>
<p>(un)subscription request sent</p>
}
</xsp:logic>
</page>
</xsp:page>
]]></source_code>
</slide>
<slide>
<title>Exceptions Taglib</title>
<point level="1">Previous slides lacking error handling</point>
<point level="1">All AxKit taglibs throw exceptions on error</point>
<point level="1">Need Exception taglib to catch them</point>
<source_code><![CDATA[
<except:try>
... # code that throws exceptions
<except:catch>
<error>An Error Occured: <except:message/></error>
</except:catch>
</except:try>
]]></source_code>
</slide>
<slide>
<title>ESQL Taglib</title>
<point level="1">The mother of all taglibs</point>
<point level="1">Executes SQL against DBI connections</point>
<point level="1">Add Apache::DBI in the mix for cached connections</point>
<point level="1">Options for SQL with or without results</point>
<point level="1">Columns can be retrieved all at once, or one at a time</point>
<point level="1">Allows emulation of nested queries (e.g. for MySQL)</point>
</slide>
<slide>
<title>ESQL Example</title>
<source_code><![CDATA[
<xsp:page
language="Perl"
xmlns:xsp="http://apache.org/xsp/core/v1"
xmlns:esql="http://apache.org/xsp/SQL/v2"
xmlns:except="http://axkit.org/NS/xsp/exception/v1"
xmlns:param="http://axkit.org/NS/xsp/param/v1"
indent-result="no"
>
<addresses>
<esql:connection>
<!-- connection parameters (normally use entities for these) -->
<esql:driver>Pg</esql:driver>
<esql:dburl>dbname=phonebook</esql:dburl>
<esql:username>postgres</esql:username>
<esql:password></esql:password>
<except:try>
<esql:execute-query>
]]></source_code>
</slide>
<slide>
<title>ESQL Example (cont)</title>
<source_code><![CDATA[
<xsp:logic>
if (<param:address_id/>) {
<esql:query>
SELECT * FROM address WHERE id =
<esql:parameter><param:address_id/></esql:parameter>
</esql:query>
}
else {
<esql:query>
SELECT * FROM address
</esql:query>
}
</xsp:logic>
<esql:results>
<esql:row-results>
<address>
<esql:get-columns/>
</address>
</esql:row-results>
</esql:results>
</esql:execute-query>
]]></source_code>
</slide>
<slide>
<title>ESQL Example (cont)</title>
<source_code><![CDATA[
<except:catch>
Error Occured: <except:message/>
</except:catch>
</except:try>
</esql:connection>
</addresses>
</xsp:page>
]]></source_code>
</slide>
<slide>
<title>ESQL Example (results)</title>
<source_code><![CDATA[
<addresses>
<address>
<id>2</id>
<last_name>Sergeant</last_name>
<first_name>Matt</first_name>
<title>Mr</title>
<company>AxKit.com Ltd</company>
<email>matt@axkit.com</email>
<classification_id>1</classification_id>
</address>
</addresses>
]]></source_code>
</slide>
<slide>
<title>PerForm</title>
<point>Easy Form handling</point>
<point>Should allow for same form to be presented to WML or other formats. But HTML is the priority</point>
<point>Callback based</point>
<point>Supports validation, loading, start/end form</point>
<point>Generates XML abstract form, not HTML or XHTML form</point>
<point>Example XSLT supplied for transforming to HTML</point>
</slide>
<slide>
<title>PerForm (example)</title>
<source_code><![CDATA[
<xsp:page>
<xsp:logic>
...
</xsl:logic>
<html>
<head><title>Index</title></head>
<body>
<f:form name="view">
<h2>View Items</h2>
<f:single-select name="types"/>
<f:submit name="go" value="Go!"/>
</f:form>
</body>
</html>
</xsp:page>
]]></source_code>
<point>Callbacks implemented in top-level <xsp:logic> section</point>
</slide>
<slide>
<title>PerForm (example cont)</title>
<point>Callbacks:</point>
<source_code><![CDATA[
sub load_types {
my ($ctxt, $selected) = @_;
my $db = AxKit::RHNews::DB->instance();
my @types = $db->get_asset_types();
return $selected, "All Types" => "0", map { $_->[1] => $_->[1] } @types;
}
use Apache::Util ();
sub submit_go {
my ($ctxt) = @_;
return "view_asset.xsp?type=" . Apache::Util::escape_uri($ctxt->{Form}{types});
}
]]></source_code>
</slide>
<slide>
<title>Building your XSP Based App</title>
<point level="1">Hide logic in taglibs</point>
<point level="1">Or hide logic in Perl modules called from <xsp:logic> and <xsp:expr> sections</point>
<point level="1">Generate XML, not HTML</point>
<point level="1">Process to HTML in a further pipeline transformation</point>
</slide>
</slideset>
<slideset>
<title>Advanced AxKit</title>
<slide>
<title>Alternate Providers</title>
<point>Builtins: File, Filter, Scalar</point>
<point>Others: e.g. OpenOffice</point>
<point>API:</point>
<point level="2">init( %params )</point>
<point level="2">key() = string</point>
<point level="2">exists() = ?</point>
<point level="2">process() = ?</point>
<point level="2">decline(%args)</point>
<point level="2">mtime() = int</point>
<point level="2">get_fh() = FH</point>
<point level="2">get_strref() = strref</point>
<point level="2">has_changed($mtime) = ?</point>
<point level="2">get_ext_ent_handler() = subref</point>
<point level="2">get_styles($media, $style)</point>
</slide>
<slide>
<title>Example: Ever expiring file</title>
<source_code><![CDATA[
package Apache::AxKit::Provider::ExpireFile;
use Apache::AxKit::Provider::File;
@ISA = ('Apache::AxKit::Provider::File');
sub has_changed {
my ($self, $mtime) = @_;
return 1;
}
1;
]]></source_code>
<point>httpd.conf</point>
<source_code><![CDATA[
AxProvider Apache::AxKit::Provider::ExpireFile
]]></source_code>
</slide>
<slide>
<title>Alternate ConfigReaders</title>
<point level="2">new($r)</point>
<point level="2">StyleMap() = hashref</point>
<point level="2">CacheDir() = string</point>
<point level="2">ProviderClass() = string</point>
<point level="2">PreferredStyle() = string</point>
<point level="2">PreferredMedia() = string</point>
<point level="2">CacheModule() = string</point>
<point level="2">DebugLevel() = int</point>
<point level="2">StackTrace() = ?</point>
<point level="2">LogDeclines() = ?</point>
<point level="2">OutputCharset() = string</point>
<point level="2">ErrorStyles() = complex</point>
<point level="2">GzipOutput() = ?</point>
<point level="2">DoGzip() = ?</point>
<point level="2">GetMatchingProcessors() = complex</point>
<point level="2">XSPTaglibs() = list</point>
</slide>
<slide>
<title>Alternate Cache Module</title>
<point>Possibly store cache in a database</point>
<point>API:</point>
<point level="2">new($r, $xmlfile, @parts)</point>
<point level="2">write($string)</point>
<point level="2">read() = string</point>
<point level="2">get_fh() = FH</point>
<point level="2">deliver()</point>
<point level="2">reset()</point>
<point level="2">has_changed($mtime) = ?</point>
<point level="2">mtime()</point>
<point level="2">exists() = ?</point>
<point level="2">key() = string</point>
<point level="2">no_cache($on)</point>
</slide>
</slideset>
<slide>
<title>AxKit Future Plans</title>
<point level="1">Content Management System, for easier site building and maintainence</point>
<point level="1">Apache 2.0 Port</point>
<point level="1">More SAX work, more C work</point>
<point level="1">Allow JSP taglibs to work</point>
</slide>
<slide>
<title>Conclusions</title>
<point level="1">Perl and XML are a powerful combination</point>
<point level="1">XPath and XSLT add to the mix...</point>
<point level="1">AxKit can reduce your long term costs</point>
<point level="2">In site re-design</point>
<point level="2">and in content re-purposing</point>
<point level="1">Open Source equal to commercial alternatives</point>
</slide>
<slide>
<title>Resources and contact</title>
<point level="1">AxKit: http://axkit.org/</point>
<point level="1">CPAN: http://search.cpan.org</point>
<point level="1">libxml and libxslt: http://www.xmlsoft.org</point>
<point level="1">Sablotron: http://www.gingerall.com</point>
<point level="1">XPath and XSLT Tutorials: http://zvon.org</point>
</slide>
</slideshow>