The Element type is a simple but flexible container object, designed to store hierarchical data structures, such as simplified XML infosets, in memory. The ElementTree package provides a Python implementation of this type, plus code to serialize element trees to and from XML files. The 1.2 alpha 2 release fixes some minor issues (most notably, nicer serialization of the "xml" default namespace). You can get the ElementTree toolkit from: http://effbot.org/downloads Brief documentation and some code samples (including an XML-RPC unmarshaller in 16 lines) are available from: http://effbot.org/zone/element-index.htm If you want more sample code, my blog currently features a number of short articles on parsing stuff with element trees (the first four all deal with RSS and RSS-like formats): http://online.effbot.org/2003_07_01_archive.htm#element-tricks In a related release, the ElementTidy library provides an alternative tree builder that can read (almost) arbitrary HTML, and turn it into well-formed XHTML element trees. The ElementTidy library uses a library versions of Dave Raggett's HTML Tidy utility to do the clean- up (source code is included), and does not rely on external utilities. http://effbot.org/downloads http://effbot.org/zone/element-tidylib.htm enjoy /F