libs/elementtree/SimpleXMLTreeBuilder.py

   1 #
   2 # ElementTree
   3 # $Id: SimpleXMLTreeBuilder.py 1862 2004-06-18 07:31:02Z Fredrik $
   4 #
   5 # A simple XML tree builder, based on Python's xmllib
   6 #
   7 # Note that due to bugs in xmllib, this builder does not fully support
   8 # namespaces (unqualified attributes are put in the default namespace,
   9 # instead of being left as is).  Run this module as a script to find
  10 # out if this affects your Python version.
  11 #
  12 # history:
  13 # 2001-10-20 fl   created
  14 # 2002-05-01 fl   added namespace support for xmllib
  15 # 2002-08-17 fl   added xmllib sanity test
  16 #
  17 # Copyright (c) 1999-2004 by Fredrik Lundh.  All rights reserved.
  18 #
  19 # fredrik@pythonware.com
  20 # http://www.pythonware.com
  21 #
  22 # --------------------------------------------------------------------
  23 # The ElementTree toolkit is
  24 #
  25 # Copyright (c) 1999-2004 by Fredrik Lundh
  26 #
  27 # By obtaining, using, and/or copying this software and/or its
  28 # associated documentation, you agree that you have read, understood,
  29 # and will comply with the following terms and conditions:
  30 #
  31 # Permission to use, copy, modify, and distribute this software and
  32 # its associated documentation for any purpose and without fee is
  33 # hereby granted, provided that the above copyright notice appears in
  34 # all copies, and that both that copyright notice and this permission
  35 # notice appear in supporting documentation, and that the name of
  36 # Secret Labs AB or the author not be used in advertising or publicity
  37 # pertaining to distribution of the software without specific, written
  38 # prior permission.
  39 #
  40 # SECRET LABS AB AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD
  41 # TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANT-
  42 # ABILITY AND FITNESS.  IN NO EVENT SHALL SECRET LABS AB OR THE AUTHOR
  43 # BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY
  44 # DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
  45 # WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
  46 # ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
  47 # OF THIS SOFTWARE.
  48 # --------------------------------------------------------------------
  49
  50 ##
  51 # Tools to build element trees from XML files, using <b>xmllib</b>.
  52 # This module can be used instead of the standard tree builder, for
  53 # Python versions where "expat" is not available (such as 1.5.2).
  54 # <p>
  55 # Note that due to bugs in <b>xmllib</b>, the namespace support is
  56 # not reliable (you can run the module as a script to find out exactly
  57 # how unreliable it is on your Python version).
  58 ##
  59
  60 import xmllib, string
  61
  62 import ElementTree
  63
  64 ##
  65 # ElementTree builder for XML source data.
  66 #
  67 # @see elementtree.ElementTree
  68
  69 class TreeBuilder(xmllib.XMLParser):
  70
  71     def __init__(self, html=0):
  72         self.__builder = ElementTree.TreeBuilder()
  73         if html:
  74             import htmlentitydefs
  75             self.entitydefs.update(htmlentitydefs.entitydefs)
  76         xmllib.XMLParser.__init__(self)
  77
  78     ##
  79     # Feeds data to the parser.
  80     #
  81     # @param data Encoded data.
  82
  83     def feed(self, data):
  84         xmllib.XMLParser.feed(self, data)
  85
  86     ##
  87     # Finishes feeding data to the parser.
  88     #
  89     # @return An element structure.
  90     # @defreturn Element
  91
  92     def close(self):
  93         xmllib.XMLParser.close(self)
  94         return self.__builder.close()
  95
  96     def handle_data(self, data):
  97         self.__builder.data(data)
  98
  99     handle_cdata = handle_data
 100
 101     def unknown_starttag(self, tag, attrs):
 102         attrib = {}
 103         for key, value in attrs.items():
 104             attrib[fixname(key)] = value
 105         self.__builder.start(fixname(tag), attrib)
 106
 107     def unknown_endtag(self, tag):
 108         self.__builder.end(fixname(tag))
 109
 110
 111 def fixname(name, split=string.split):
 112     # xmllib in 2.0 and later provides limited (and slightly broken)
 113     # support for XML namespaces.
 114     if " " not in name:
 115         return name
 116     return "{%s}%s" % tuple(split(name, " ", 1))
 117
 118
 119 if __name__ == "__main__":
 120     import sys
 121     # sanity check: look for known namespace bugs in xmllib
 122     p = TreeBuilder()
 123     text = """\
 124     <root xmlns='default'>
 125        <tag attribute='value' />
 126     </root>
 127     """
 128     p.feed(text)
 129     tree = p.close()
 130     status = []
 131     # check for bugs in the xmllib implementation
 132     tag = tree.find("{default}tag")
 133     if tag is None:
 134         status.append("namespaces not supported")
 135     if tag is not None and tag.get("{default}attribute"):
 136         status.append("default namespace applied to unqualified attribute")
 137     # report bugs
 138     if status:
 139         print "xmllib doesn't work properly in this Python version:"
 140         for bug in status:
 141             print "-", bug
 142     else:
 143         print "congratulations; no problems found in xmllib"
 144