ScaleDOM

A lazy-loading DOM implementation for processing huge XML documents.

Synopsis

ScaleDOM is a Xerces-based XML DOM parser which has a small memory footprint due to lazy loading of XML nodes. It only keeps a portion of the XML document in memory and re-loads nodes from the source file when necessary.

If you run into "OutOfMemoryError" using your standard DOM parser, ScaleDOM may be just the right solution for you.

Usage

Please refer to the folder "example" for a small sample project. The class ScaleDomParsingTest illustrates how to dynamically enable/disable ScaleDOM parsing using the corresponding system property:

System.setProperty(
    "javax.xml.parsers.DocumentBuilderFactory", 
    ScaleDomDocumentBuilderFactory.class.getName()
);

To run the sample project, first build and install ScaleDOM using

mvn install

and then run the tests in the "example" project:

cd example
mvn test

Project Details

For detailed information, please refer to the document doc/ScaleDOM.pdf, which also contains a small performance evaluation of ScaleDOM.

Change Log

• 2013-09-14: v1.2
  • parse/traverse XML directly from a URL connection. Portions of the document are lazily loaded using the "Range=startByte-endByte" HTTP header.
• 2013-09-13: v1.1
  • allow dynamic switching between ScaleDOM and Xerces using system property "javax.xml.parsers.DocumentBuilderFactory"
  • add isScaleDomEnabled() method to the patched Xerces classes (CoreDocumentImpl, DocumentFragmentImpl, ElementImpl, EntityReferenceImpl, ParentNode) to dynamically detect whether we are in "ScaleDOM mode" or not.
  • add example project to illustrate the use of ScaleDOM
  • source code compatibility with Java 1.6 (removed 1.7 specific code)
• 2013-08-29: v1.0
  • initial release with base functionality

Developers

• Dominik Rauch (e0825084@student.tuwien.ac.at)
• Waldemar Hummer (hummer@infosys.tuwien.ac.at)

License

ScaleDOM is published open-source under the Apache License 2.0.