JAXP Pipeline Examples

The following sample programs demonstrate various aspects of JAXP pipeline construction.  They are appropriate for under-the-hood use of the DeltaXML software, however following the introduction of the DeltaXML Core 3.0 release and the PipelinedComparator class, we would suggest DeltaXML users consider using this class before using JAXP and the techniques described here.

This is a work in progress with the aim of providing a wide range of examples and descriptive text to form a complete tutorial for this topic.

1. Dependencies

Some of the example pipelines use the DeltaXML XMLComparator class  from com.deltaxml.api package in DeltaXML Core; if you do not already have this software a time-limited evalution can be requested. The concepts in these examples are also relevant to both JAXP and compatible products such as the DeltaXML Core API; some examples use JAXP without DeltaXML software.  A column in the table below indicates if the example requires DeltaXML or JAXP.

The examples can be compiled and run using a J2SE 1.4 or later JDK (a javac compiler is needed, so the JRE is not appropriate).
The JDK for a number of platforms is available from http://java.sun.com.

Ant build scripts are provided to make compiling and running the examples easier.  While not compulsory we would recommend using Ant. It is available from: http://ant.apache.org.

Finally, some of the example work best using the Saxon XSLT processor. We recommend using version 6.5.3 available from: http://saxon.sourceforge.net/.

2. Pipeline paper

A paper which was developed in conjunction with these examples was presented at XML2004 (Washington DC, November 2004).  Please see our articles and papers page for further details.

3. The Examples

You can explore the individual examples listed below by downloading pipelines.zip and unpacking all the examples in a single bundle.

Example Description Core API required
Stream File The Stream File example is one of the simplest possible pipelines. It demonstrates how a single JAXP pipeline component can utilize File IO. Yes
Stream String This example shows how IO can be performed to and from Java Strings. This may be useful when strings are associated with an internal storage buffer for a database or GUI widget. Yes
Parser Control Parsers used at the start of a pipeline may need special configuration of certain types of XML processing. This example shows how SAX parsers can be configured and integrated into a pipeline. This example also shows some aspects of pull-mode processing. Yes
Output Control While it is possible to statically configure the final stage of an XSLT based pipeline component, using the <xsl:output> statement, it may be necessary to configure aspects such as indentation at runtime. This example demonstrates a technique which can be used with triggered and push-mode pipeline components. No
Five Stage This example demonstrates how pull-mode, triggered and push-mode pipeline components can be integrated together. The pipeline consists of a sequence of 5 XSLT processing stages. No
Doctype Preservation Some pipeline components are not DOCTYPE aware. This example demonstrate a technique for determining the input DOCTYPE (using pre-parsing) and then reinstating it at the end of the pipeline. It borrows some techniques from the Output and Parser Control examples. No