XSLT

Index > Books > Declarative Development Guide > Technology Tour > Core XML > XSLT

Introduction

XSLT transforms are the most common means of transforming XML. NetKernel makes it simple to apply XSL transforms and capture the results for additional processing. XSLTs can be used as modular components of an application. Pipelined transforms make it easy to create filters, generate views on data, or style data for presentation.

XSLT 1.0

NetKernel optimizes XSLT by transparently caching transforms. In addition it seamlessly provides the ability to use interpreted or compiled transforms. XSLT v1.0 transformations are performed with the xslt or xsltc accessors.

Here's an example showing a pipelined filter-presentation pattern. First a transform is applied to extract Act 1 from King Lear. The extracted act is then transformed to XHTML and presented.

<idoc>  <seq>
    <instr>
      <comment> ********************************* Select Act1 from the complete play and store as variable ********************************* </comment>
      <type>xslt</type>
      <operand>../content/lear.xml</operand>
      <operator>act1.xsl</operator>
      <target>var:act1</target>
    </instr>
    <instr>
      <comment> ********************************* Transform act1 to XHTML and send to response ********************************* </comment>
      <type>xslt</type>
      <operand>var:act1</operand>
      <operator>act_to_html.xsl</operator>
      <target>this:response</target>
    </instr>
  </seq>
</idoc>

Here are the documents used

The source data (since this is very large and intensive for browsers to render this is only showing the Act 1 Scene 1 section) view
The act1 filter transform view
The XHTML presentation transform view

XSLT 2.0

Also provided is the xslt2 accessor which supports the draft W3C XSLT v2.0 standard. XSLT v2.0 provides many enhancements to the language and uses XPath 2.0 for path and functions.

As with the xslt v1.0 accessor all XSLT 2.0 transforms are transparently cached - furthermore the xslt2 accessor is fully backwards compatible with v1.0 transforms. Your choice of which xslt transform accessor to use will depend on a trade-off between feature set, performance and standards maturity.