XIST

A new API for iterating through XML trees has been
added. It's possible to pass iterators and
generators to Frag and Element constructors now.
Parsing broken HTML is now done with the HTML
parser from libxml2. The parser now has the option
to ignore illegal elements, attributes, processing
instructions, and entities. Other minor changes
and updates were made.

An extensible validation architecture has been implemented. The SVG module has been updated to 1.1. The docbook module has been updated to 4.3. Experimental support for XPython has been added. Passing global attributes has been simplified. CharRef handling has been simplified. There are many other small enhancements and bugfixes.

Parsers and publishers have been made reusable, and the parsing and publishing APIs have been simplified. Converter contexts have been changed to work better with subclasses. A new find function that also searches attributes has been added. Duplicate character references have been removed from the iHTML namespace.

Namespace handling has been rewritten to be more standards compliant (no more namespace prefixes for entity references or processing instructions). Global attributes will now always generate the appropriate xmlns attributes. Support for uTidylib has been added, and arguments can be passed to tidy. The HTMLParser can handle global attributes. When parsing from a URL, the base URL will be correct even if the request gets redirected.

Python 2.3 is required now. Namespaces have been enhanced and are now subclassable, so it's possible to overwrite elements. Tree traversal has been rewritten to be more flexible. A namespace for XSL-FO has been added and ll.xist.ns.doc supports conversion to XSL-FO now. The HTML handling of the HTMLParser has been improved. XIST DOMs now support copy.copy() and copy.deepcopy(). There are many other small bugfixes and enhancements.