misc/xmlwrapper.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3 #
   4 # Copyright 2004, 2005 Zuza Software Foundation
   5 #
   6 # This file is part of translate.
   7 #
   8 # translate is free software; you can redistribute it and/or modify
   9 # it under the terms of the GNU General Public License as published by
  10 # the Free Software Foundation; either version 2 of the License, or
  11 # (at your option) any later version.
  12 #
  13 # translate is distributed in the hope that it will be useful,
  14 # but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 # GNU General Public License for more details.
  17 #
  18 # You should have received a copy of the GNU General Public License
  19 # along with translate; if not, write to the Free Software
  20 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  21
  22 """simpler wrapper to the elementtree XML parser"""
  23
  24 try:
  25     from xml.etree import ElementTree
  26 except ImportError:
  27     from elementtree import ElementTree
  28 # this is needed to prevent expat-version conflicts with wx >= 2.5.2.2
  29 from xml.parsers import expat
  30
  31 basicfixtag = ElementTree.fixtag
  32
  33 def makefixtagproc(namespacemap):
  34     """this constructs an alternative fixtag procedure that will use appropriate names for namespaces..."""
  35     def fixtag(tag, namespaces):
  36         """given a decorated tag (of the form {uri}tag), return prefixed tag and namespace declaration, if any"""
  37         if isinstance(tag, ElementTree.QName):
  38             tag = tag.text
  39         namespace_uri, tag = tag[1:].split("}", 1)
  40         prefix = namespaces.get(namespace_uri)
  41         if prefix is None:
  42             if namespace_uri in namespacemap:
  43                 prefix = namespacemap[namespace_uri]
  44             else:
  45                 prefix = "ns%d" % len(namespaces)
  46             namespaces[namespace_uri] = prefix
  47             xmlns = ("xmlns:%s" % prefix, namespace_uri)
  48         else:
  49             xmlns = None
  50         return "%s:%s" % (prefix, tag), xmlns
  51     return fixtag
  52
  53 def splitnamespace(fulltag):
  54     if '{' in fulltag:
  55         namespace = fulltag[fulltag.find('{'):fulltag.find('}')+1]
  56     else:
  57         namespace = ""
  58     tag = fulltag.replace(namespace, "", 1)
  59     return namespace, tag
  60
  61 class XMLWrapper:
  62     """simple wrapper for xml objects"""
  63     def __init__(self,obj):
  64         """construct object from the elementtree item"""
  65         self.obj = obj
  66         self.namespace, self.tag = splitnamespace(self.obj.tag)
  67         self.attrib = {}
  68         for fullkey, value in self.obj.attrib.iteritems():
  69             namespace, key = splitnamespace(fullkey)
  70             self.attrib[key] = value
  71     def getchild(self, searchtag, tagclass=None):
  72         """get a child with the given tag name"""
  73         if tagclass is None: tagclass = XMLWrapper
  74         for childobj in self.obj.getiterator():
  75             # getiterator() includes self...
  76             if childobj == self.obj: continue
  77             childns, childtag = splitnamespace(childobj.tag)
  78             if childtag == searchtag:
  79                 child = tagclass(childobj)
  80                 return child
  81         raise KeyError("could not find child with tag %r" % searchtag)
  82     def getchildren(self, searchtag, tagclass=None, excludetags=[]):
  83         """get all children with the given tag name"""
  84         if tagclass is None: tagclass = XMLWrapper
  85         childobjects = []
  86         for childobj in self.obj.getiterator():
  87             # getiterator() includes self...
  88             if childobj == self.obj: continue
  89             childns, childtag = splitnamespace(childobj.tag)
  90             if childtag == searchtag:
  91                 childobjects.append(childobj)
  92         children = [tagclass(childobj) for childobj in childobjects]
  93         return children
  94     def gettext(self, searchtag):
  95         """get some contained text"""
  96         return self.getchild(searchtag).obj.text
  97     def getxml(self, encoding=None):
  98         return ElementTree.tostring(self.obj, encoding)
  99     def getplaintext(self, excludetags=[]):
 100         text = ""
 101         if self.obj.text != None: text += self.obj.text
 102         for child in self.obj._children:
 103             simplechild = XMLWrapper(child)
 104             if simplechild.tag not in excludetags:
 105                 text += simplechild.getplaintext(excludetags)
 106         if self.obj.tail != None: text += self.obj.tail
 107         return text
 108     def getvalues(self, searchtag):
 109         """get some contained values..."""
 110         values = [child.obj.text for child in self.getchildren(searchtag)]
 111         return values
 112     def __repr__(self):
 113         """return a representation of the object"""
 114         return self.tag+':'+repr(self.__dict__)
 115     def getattr(self, attrname):
 116         """gets an attribute of the tag"""
 117         return self.attrib[attrname]
 118     def write(self, file, encoding="UTF-8"):
 119         """writes the object as XML to a file..."""
 120         e = ElementTree.ElementTree(self.obj)
 121         e.write(file, encoding)
 122
 123 def BuildTree(xmlstring):
 124     parser = ElementTree.XMLTreeBuilder()
 125     parser.feed(xmlstring)
 126     return parser.close()
 127
 128 def MakeElement(tag, attrib={}, **extraargs):
 129     return ElementTree.Element(tag, attrib, **extraargs)
 130