The ElementTree Library
$Id: CHANGES 2326 2005-03-17 07:45:21Z fredrik $

*** Changes from release 1.1 to 1.2 ***

(1.2.6 released)

- Fixed handling of entities defined in internal DTD's (reported
  by Greg Wilson).

- Fixed serialization under non-standard default encodings (but
  using non-standard default encodings is still a lousy idea ;-)

(1.2.5 released)

- Added 'iterparse' implementation.  This is similar to 'parse', but
  returns a stream of events while it builds the tree.  By default,
  the parser only returns "end" events (for completed elements):

	for event, elem in iterparse(source):
	    ...

  To get other events, use the "events" option to pass in a tuple
  containing the events you want:

	for event, elem in iterparse(source, events=(...)):
	    ...

  The event tuple can contain one or more of:

  "start"
    generated for start tags, after the element has been created
    (but before the current element has been fully populated)
  "end"
    generated for end tags, after all element children has been
    created.
  "start-ns"
    generated when a new namespace scope is opened.  for this event,
    the elem value is a (prefix, url) tuple.
  "end-ns"
    generated when the current namespace scope is closed.  elem
    is None.

  Events arrive asynchronously; the tree is usually more complete
  than the events indicate, but this is nothing you can rely on.

  The iterable itself contains context information.  In the current
  release, the only public context attribute is "root", which is set
  to the root element when parsing is finished.  To access the con-
  text, assign the iterable to a variable before looping over it:

	context = iterparse(source)
	for event, elem in context:
	    ...
	root = context.root

(1.2.4 released)

- Fixed another FancyTreeBuilder bug on Python 2.3.

(1.2.3 released)

- Fixed the FancyTreeBuilder class, which was broken in 1.2.1
  and 1.2.2 (broken for some Python versions, at least).

(1.2.2 released)

- Fixed some ASCII/Unicode issues in the HTML parser.  You can now
  use the parser on documents that mixes encoded 8-bit data with
  character references outside the ASCII range. (backported from 1.3)

(1.2.1 released)

- Changed XMLTreeBuilder to take advantage of new expat features, if
  present.  This speeds up parsing quite a bit. (backported from 1.3)

(1.2c1 released; 1.2 final released)

- Added 'docs' directory, with PythonDoc documentation for the
  ElementTree library.  See docs/index.html for an overview.

(1.2b4 released)

- Fixed encoding of Unicode element names and attribute names
  (reported by Ken Rimey).

(1.2b3 released)

- Added default argument to 'findtext'.  Note that 'findtext' now
  always returns an empty string if a matching element is found, but
  has no text content.  None is only returned if no element is found,
  and no default value is specified.

- Make sure 'dump' adds a trailing linefeed.

(1.2b2 released)

- Added optional tree builder argument to the HTMLTreeBuilder class.

(1.2b1 released)

- Added XMLID() helper.  This is similar to XML(), but returns both
  the root element and a dictionary mapping ID attributes to elements.

- Added simple SgmlopXMLTreeBuilder module.  This is a very fast
  parser, but it doesn't yet support namespaces.  To use this parser,
  you need the sgmlop driver:

      http://effbot.org/zone/sgmlop-index.htm

- Fixed exception in test suite; the TidyHTMLTreeBuilder class
  now raises a RuntimeError exception if the _elementidy module
  is not available.

(1.2a5 released)

- Fixed problem that could result in repeated use of the same
  namespace prefix in the same element (!).

- Fixed import error in ElementInclude, when using the default
  loader (Gustavo Niemeyer).

(1.2a4 released)

- Fixed exception when .//tag fails to find matching elements
  (reported by Mike Kent) (@XMLTOOLKIT28)

- Fall back on pre-1.2 find/findtext/findall behaviour if the
  ElementPath module is not installed.  If you don't need path
  support, you can simply copy the ElementTree module to your
  own project.

(1.2a3 released)

- Added experimental support for XInclude-style preprocessing.  The
  ElementInclude module expands xi:include elements, using a custom
  resolver.  The current release ignores xi:fallback elements.

- Fixed typo in ElementTree.findtext (reported by Thomas Dartsch)
  (@XMLTOOLKIT25)

- Fixed parsing of periods in element names (reported by Brian
  Vicente) (@XMLTOOLKIT27)

(1.2a2 released)

- Fixed serialization of elements and attributes in the XML default
  namespace (http://www.w3.org/XML/1998/namespace).  Added "rdf" to
  the set of "well-known" namespace prefixes.

- Added 'makeelement' factory method.  Added 'target' argument to
  XMLTreeBuilder class.

(1.2a1 released)

- Added support for a very limited subset of the abbreviated XPath
  syntax.  The following location paths are supported:

	tag -- select all subelements with the given tag
	. -- select this element
	* -- select all subelements
	// (empty path) -- select all subelements, on all levels

  Examples:

	p -- select all p subelements
	.//a -- select all a sublements, at all sublevels
	*/img -- select all img grandchildren
	ul/li -- select all li elements that are children of ul elements
	.//ul/li -- same, but select elements anywhere in the subtree

  Absolute paths (paths starting with a slash) can only be used on
  ElementTree instances.  To use // on an Element instance, add a
  leading period (.).

*** Changes from release 1.0 to 1.1 ***

(1.1 final released)

- Added 'fromstring' and 'tostring' helpers.  The 'XML' function is
  an alias for 'fromstring', and provides a convenient way to add XML
  literals to source code:

	from elementtree.ElementTree import XML

	element = XML('<element>content</element>')

- Moved XMLTreeBuilder functionality into the ElementTree module.  If
  all you need is basic XML support, you can simply copy the ElementTree
  module to your own project.

- Added SimpleXMLWriter module.

(1.1b2 released)

- Changed default encoding to US-ASCII.  Use tree.write(file, "utf-8")
  to get the old behaviour.  If the tree contains text that cannot be
  encoded using the given encoding, the writer uses numerical entities
  for all non-ASCII characters in that text segment.

(1.1b1 released)

- Map tags and attribute names having the same value to the same
  object.  This saves space when reading large XML trees, and also
  gives a small speedup (less than 10%).

- Added benchmark script.  This script takes a filename argument, and
  loads the given file into memory using the XML and SimpleXML tree
  builders.  For each parser, it reports the document size and the
  time needed to parse the document.