| Home | Trees | Indices | Help |
|
|---|
|
|
1 #
2 # ElementTree
3 # $Id: ElementInclude.py 1862 2004-06-18 07:31:02Z Fredrik $
4 #
5 # limited xinclude support for element trees
6 #
7 # history:
8 # 2003-08-15 fl created
9 # 2003-11-14 fl fixed default loader
10 #
11 # Copyright (c) 2003-2004 by Fredrik Lundh. All rights reserved.
12 #
13 # fredrik@pythonware.com
14 # http://www.pythonware.com
15 #
16 # --------------------------------------------------------------------
17 # The ElementTree toolkit is
18 #
19 # Copyright (c) 1999-2004 by Fredrik Lundh
20 #
21 # By obtaining, using, and/or copying this software and/or its
22 # associated documentation, you agree that you have read, understood,
23 # and will comply with the following terms and conditions:
24 #
25 # Permission to use, copy, modify, and distribute this software and
26 # its associated documentation for any purpose and without fee is
27 # hereby granted, provided that the above copyright notice appears in
28 # all copies, and that both that copyright notice and this permission
29 # notice appear in supporting documentation, and that the name of
30 # Secret Labs AB or the author not be used in advertising or publicity
31 # pertaining to distribution of the software without specific, written
32 # prior permission.
33 #
34 # SECRET LABS AB AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD
35 # TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANT-
36 # ABILITY AND FITNESS. IN NO EVENT SHALL SECRET LABS AB OR THE AUTHOR
37 # BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY
38 # DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
39 # WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
40 # ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
41 # OF THIS SOFTWARE.
42 # --------------------------------------------------------------------
43
44 """
45 Limited XInclude support for the ElementTree package.
46
47 While lxml.etree has full support for XInclude (see
48 `etree.ElementTree.xinclude()`), this module provides a simpler, pure
49 Python, ElementTree compatible implementation that supports a simple
50 form of custom URL resolvers.
51 """
52
53 import copy, etree
54 from urlparse import urljoin
55 from urllib2 import urlopen
56
57 try:
58 set
59 except NameError:
60 from sets import Set as set
61
62 XINCLUDE = "{http://www.w3.org/2001/XInclude}"
63
64 XINCLUDE_INCLUDE = XINCLUDE + "include"
65 XINCLUDE_FALLBACK = XINCLUDE + "fallback"
66
67 ##
68 # Fatal include error.
69
72
73 ##
74 # ET compatible default loader.
75 # This loader reads an included resource from disk.
76 #
77 # @param href Resource reference.
78 # @param parse Parse mode. Either "xml" or "text".
79 # @param encoding Optional text encoding.
80 # @return The expanded resource. If the parse mode is "xml", this
81 # is an ElementTree instance. If the parse mode is "text", this
82 # is a Unicode string. If the loader fails, it can return None
83 # or raise an IOError exception.
84 # @throws IOError If the loader fails to load the resource.
85
87 file = open(href)
88 if parse == "xml":
89 data = etree.parse(file).getroot()
90 else:
91 data = file.read()
92 if encoding:
93 data = data.decode(encoding)
94 file.close()
95 return data
96
97 ##
98 # Default loader used by lxml.etree - handles custom resolvers properly
99 #
100
102 if parse == "xml":
103 data = etree.parse(href, parser).getroot()
104 else:
105 if "://" in href:
106 f = urlopen(href)
107 else:
108 f = open(href)
109 data = f.read()
110 f.close()
111 if encoding:
112 data = data.decode(encoding)
113 return data
114
115 ##
116 # Wrapper for ET compatibility - drops the parser
117
121 return load
122
123
124 ##
125 # Expand XInclude directives.
126 #
127 # @param elem Root element.
128 # @param loader Optional resource loader. If omitted, it defaults
129 # to {@link default_loader}. If given, it should be a callable
130 # that implements the same interface as <b>default_loader</b>.
131 # @throws FatalIncludeError If the function fails to include a given
132 # resource, or if the tree contains malformed XInclude elements.
133 # @throws IOError If the function fails to load a given resource.
134 # @returns the node or its replacement if it was an XInclude node
135
137 if base_url is None:
138 if hasattr(elem, 'getroot'):
139 tree = elem
140 elem = elem.getroot()
141 else:
142 tree = elem.getroottree()
143 if hasattr(tree, 'docinfo'):
144 base_url = tree.docinfo.URL
145 elif hasattr(elem, 'getroot'):
146 elem = elem.getroot()
147 _include(elem, loader, base_url=base_url)
148
150 if loader is not None:
151 load_include = _wrap_et_loader(loader)
152 else:
153 load_include = _lxml_default_loader
154
155 if _parent_hrefs is None:
156 _parent_hrefs = set()
157
158 parser = elem.getroottree().parser
159
160 include_elements = list(
161 elem.getiterator('{http://www.w3.org/2001/XInclude}*'))
162
163 for e in include_elements:
164 if e.tag == XINCLUDE_INCLUDE:
165 # process xinclude directive
166 href = urljoin(base_url, e.get("href"))
167 parse = e.get("parse", "xml")
168 parent = e.getparent()
169 if parse == "xml":
170 if href in _parent_hrefs:
171 raise FatalIncludeError(
172 "recursive include of %r detected" % href
173 )
174 _parent_hrefs.add(href)
175 node = load_include(href, parse, parser=parser)
176 if node is None:
177 raise FatalIncludeError(
178 "cannot load %r as %r" % (href, parse)
179 )
180 node = _include(node, loader, _parent_hrefs)
181 if e.tail:
182 node.tail = (node.tail or "") + e.tail
183 if parent is None:
184 return node # replaced the root node!
185 parent.replace(e, node)
186 elif parse == "text":
187 text = load_include(href, parse, encoding=e.get("encoding"))
188 if text is None:
189 raise FatalIncludeError(
190 "cannot load %r as %r" % (href, parse)
191 )
192 predecessor = e.getprevious()
193 if predecessor is not None:
194 predecessor.tail = (predecessor.tail or "") + text
195 elif parent is None:
196 return text # replaced the root node!
197 else:
198 parent.text = (parent.text or "") + text + (e.tail or "")
199 parent.remove(e)
200 else:
201 raise FatalIncludeError(
202 "unknown parse type in xi:include tag (%r)" % parse
203 )
204 elif e.tag == XINCLUDE_FALLBACK:
205 parent = e.getparent()
206 if parent is not None and parent.tag != XINCLUDE_INCLUDE:
207 raise FatalIncludeError(
208 "xi:fallback tag must be child of xi:include (%r)" % e.tag
209 )
210 else:
211 raise FatalIncludeError(
212 "Invalid element found in XInclude namespace (%r)" % e.tag
213 )
214 return elem
215
| Home | Trees | Indices | Help |
|
|---|
| Generated by Epydoc 3.0 on Wed Mar 26 22:13:10 2008 | http://epydoc.sourceforge.net |