5 # light-weight XML support for Python 1.5.2 and later.
8 # 2001-10-20 fl created (from various sources)
9 # 2001-11-01 fl return root from parse method
10 # 2002-02-16 fl sort attributes in lexical order
11 # 2002-04-06 fl TreeBuilder refactoring, added PythonDoc markup
12 # 2002-05-01 fl finished TreeBuilder refactoring
13 # 2002-07-14 fl added basic namespace support to ElementTree.write
14 # 2002-07-25 fl added QName attribute support
15 # 2002-10-20 fl fixed encoding in write
16 # 2002-11-24 fl changed default encoding to ascii; fixed attribute encoding
17 # 2002-11-27 fl accept file objects or file names for parse/write
18 # 2002-12-04 fl moved XMLTreeBuilder back to this module
19 # 2003-01-11 fl fixed entity encoding glitch for us-ascii
20 # 2003-02-13 fl added XML literal factory
21 # 2003-02-21 fl added ProcessingInstruction/PI factory
22 # 2003-05-11 fl added tostring/fromstring helpers
23 # 2003-05-26 fl added ElementPath support
24 # 2003-07-05 fl added makeelement factory method
25 # 2003-07-28 fl added more well-known namespace prefixes
26 # 2003-08-15 fl fixed typo in ElementTree.findtext (Thomas Dartsch)
27 # 2003-09-04 fl fall back on emulator if ElementPath is not installed
28 # 2003-10-31 fl markup updates
29 # 2003-11-15 fl fixed nested namespace bug
30 # 2004-03-28 fl added XMLID helper
31 # 2004-06-02 fl added default support to findtext
32 # 2004-06-08 fl fixed encoding of non-ascii element/attribute names
33 # 2004-08-23 fl take advantage of post-2.1 expat features
34 # 2005-02-01 fl added iterparse implementation
35 # 2005-03-02 fl fixed iterparse support for pre-2.2 versions
37 # Copyright (c) 1999-2005 by Fredrik Lundh. All rights reserved.
39 # fredrik@pythonware.com
40 # http://www.pythonware.com
42 # --------------------------------------------------------------------
43 # The ElementTree toolkit is
45 # Copyright (c) 1999-2005 by Fredrik Lundh
47 # By obtaining, using, and/or copying this software and/or its
48 # associated documentation, you agree that you have read, understood,
49 # and will comply with the following terms and conditions:
51 # Permission to use, copy, modify, and distribute this software and
52 # its associated documentation for any purpose and without fee is
53 # hereby granted, provided that the above copyright notice appears in
54 # all copies, and that both that copyright notice and this permission
55 # notice appear in supporting documentation, and that the name of
56 # Secret Labs AB or the author not be used in advertising or publicity
57 # pertaining to distribution of the software without specific, written
60 # SECRET LABS AB AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD
61 # TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANT-
62 # ABILITY AND FITNESS. IN NO EVENT SHALL SECRET LABS AB OR THE AUTHOR
63 # BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY
64 # DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS,
65 # WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
66 # ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
68 # --------------------------------------------------------------------
74 "Element", "ElementTree",
76 "iselement", "iterparse",
78 "PI", "ProcessingInstruction",
88 # The <b>Element</b> type is a flexible container object, designed to
89 # store hierarchical data structures in memory. The type can be
90 # described as a cross between a list and a dictionary.
92 # Each element has a number of properties associated with it:
94 # <li>a <i>tag</i>. This is a string identifying what kind of data
95 # this element represents (the element type, in other words).</li>
96 # <li>a number of <i>attributes</i>, stored in a Python dictionary.</li>
97 # <li>a <i>text</i> string.</li>
98 # <li>an optional <i>tail</i> string.</li>
99 # <li>a number of <i>child elements</i>, stored in a Python sequence</li>
102 # To create an element instance, use the {@link #Element} or {@link
103 # #SubElement} factory functions.
105 # The {@link #ElementTree} class can be used to wrap an element
106 # structure, and convert it from and to XML.
109 import string, sys, re
111 class _SimpleElementPath:
112 # emulate pre-1.2 find/findtext/findall behaviour
113 def find(self, element, tag):
118 def findtext(self, element, tag, default=None):
121 return elem.text or ""
123 def findall(self, element, tag):
125 return element.getiterator(tag[3:])
135 # FIXME: issue warning in this case?
136 ElementPath = _SimpleElementPath()
138 # TODO: add support for custom namespace resolvers/default namespaces
139 # TODO: add improved support for incremental parsing
144 # Internal element class. This class defines the Element interface,
145 # and provides a reference implementation of this interface.
147 # You should not create instances of this class directly. Use the
148 # appropriate factory functions instead, such as {@link #Element}
149 # and {@link #SubElement}.
154 # @see ProcessingInstruction
156 class _ElementInterface:
157 # <tag attrib>text<child/>...</tag>tail
160 # (Attribute) Element tag.
165 # (Attribute) Element attribute dictionary. Where possible, use
166 # {@link #_ElementInterface.get},
167 # {@link #_ElementInterface.set},
168 # {@link #_ElementInterface.keys}, and
169 # {@link #_ElementInterface.items} to access
170 # element attributes.
175 # (Attribute) Text before first subelement. This is either a
176 # string or the value None, if there was no text.
181 # (Attribute) Text after this element's end tag, but before the
182 # next sibling element's start tag. This is either a string or
183 # the value None, if there was no text.
185 tail = None # text after end tag, if any
187 def __init__(self, tag, attrib):
193 return "<Element %s at %x>" % (self.tag, id(self))
196 # Creates a new element object of the same type as this element.
198 # @param tag Element tag.
199 # @param attrib Element attributes, given as a dictionary.
200 # @return A new element instance.
202 def makeelement(self, tag, attrib):
203 return Element(tag, attrib)
206 # Returns the number of subelements.
208 # @return The number of subelements.
211 return len(self._children)
214 # Returns the given subelement.
216 # @param index What subelement to return.
217 # @return The given subelement.
218 # @exception IndexError If the given element does not exist.
220 def __getitem__(self, index):
221 return self._children[index]
224 # Replaces the given subelement.
226 # @param index What subelement to replace.
227 # @param element The new element value.
228 # @exception IndexError If the given element does not exist.
229 # @exception AssertionError If element is not a valid object.
231 def __setitem__(self, index, element):
232 assert iselement(element)
233 self._children[index] = element
236 # Deletes the given subelement.
238 # @param index What subelement to delete.
239 # @exception IndexError If the given element does not exist.
241 def __delitem__(self, index):
242 del self._children[index]
245 # Returns a list containing subelements in the given range.
247 # @param start The first subelement to return.
248 # @param stop The first subelement that shouldn't be returned.
249 # @return A sequence object containing subelements.
251 def __getslice__(self, start, stop):
252 return self._children[start:stop]
255 # Replaces a number of subelements with elements from a sequence.
257 # @param start The first subelement to replace.
258 # @param stop The first subelement that shouldn't be replaced.
259 # @param elements A sequence object with zero or more elements.
260 # @exception AssertionError If a sequence member is not a valid object.
262 def __setslice__(self, start, stop, elements):
263 for element in elements:
264 assert iselement(element)
265 self._children[start:stop] = list(elements)
268 # Deletes a number of subelements.
270 # @param start The first subelement to delete.
271 # @param stop The first subelement to leave in there.
273 def __delslice__(self, start, stop):
274 del self._children[start:stop]
277 # Adds a subelement to the end of this element.
279 # @param element The element to add.
280 # @exception AssertionError If a sequence member is not a valid object.
282 def append(self, element):
283 assert iselement(element)
284 self._children.append(element)
287 # Inserts a subelement at the given position in this element.
289 # @param index Where to insert the new subelement.
290 # @exception AssertionError If the element is not a valid object.
292 def insert(self, index, element):
293 assert iselement(element)
294 self._children.insert(index, element)
297 # Removes a matching subelement. Unlike the <b>find</b> methods,
298 # this method compares elements based on identity, not on tag
301 # @param element What element to remove.
302 # @exception ValueError If a matching element could not be found.
303 # @exception AssertionError If the element is not a valid object.
305 def remove(self, element):
306 assert iselement(element)
307 self._children.remove(element)
310 # Returns all subelements. The elements are returned in document
313 # @return A list of subelements.
314 # @defreturn list of Element instances
316 def getchildren(self):
317 return self._children
320 # Finds the first matching subelement, by tag name or path.
322 # @param path What element to look for.
323 # @return The first matching element, or None if no element was found.
324 # @defreturn Element or None
326 def find(self, path):
327 return ElementPath.find(self, path)
330 # Finds text for the first matching subelement, by tag name or path.
332 # @param path What element to look for.
333 # @param default What to return if the element was not found.
334 # @return The text content of the first matching element, or the
335 # default value no element was found. Note that if the element
336 # has is found, but has no text content, this method returns an
340 def findtext(self, path, default=None):
341 return ElementPath.findtext(self, path, default)
344 # Finds all matching subelements, by tag name or path.
346 # @param path What element to look for.
347 # @return A list or iterator containing all matching elements,
349 # @defreturn list of Element instances
351 def findall(self, path):
352 return ElementPath.findall(self, path)
355 # Resets an element. This function removes all subelements, clears
356 # all attributes, and sets the text and tail attributes to None.
361 self.text = self.tail = None
364 # Gets an element attribute.
366 # @param key What attribute to look for.
367 # @param default What to return if the attribute was not found.
368 # @return The attribute value, or the default value, if the
369 # attribute was not found.
370 # @defreturn string or None
372 def get(self, key, default=None):
373 return self.attrib.get(key, default)
376 # Sets an element attribute.
378 # @param key What attribute to set.
379 # @param value The attribute value.
381 def set(self, key, value):
382 self.attrib[key] = value
385 # Gets a list of attribute names. The names are returned in an
386 # arbitrary order (just like for an ordinary Python dictionary).
388 # @return A list of element attribute names.
389 # @defreturn list of strings
392 return self.attrib.keys()
395 # Gets element attributes, as a sequence. The attributes are
396 # returned in an arbitrary order.
398 # @return A list of (name, value) tuples for all attributes.
399 # @defreturn list of (string, string) tuples
402 return self.attrib.items()
405 # Creates a tree iterator. The iterator loops over this element
406 # and all subelements, in document order, and returns all elements
407 # with a matching tag.
409 # If the tree structure is modified during iteration, the result
412 # @param tag What tags to look for (default is to return all elements).
413 # @return A list or iterator containing all the matching elements.
414 # @defreturn list or iterator
416 def getiterator(self, tag=None):
420 if tag is None or self.tag == tag:
422 for node in self._children:
423 nodes.extend(node.getiterator(tag))
427 _Element = _ElementInterface
430 # Element factory. This function returns an object implementing the
431 # standard Element interface. The exact class or type of that object
432 # is implementation dependent, but it will always be compatible with
433 # the {@link #_ElementInterface} class in this module.
435 # The element name, attribute names, and attribute values can be
436 # either 8-bit ASCII strings or Unicode strings.
438 # @param tag The element name.
439 # @param attrib An optional dictionary, containing element attributes.
440 # @param **extra Additional attributes, given as keyword arguments.
441 # @return An element instance.
444 def Element(tag, attrib={}, **extra):
445 attrib = attrib.copy()
447 return _ElementInterface(tag, attrib)
450 # Subelement factory. This function creates an element instance, and
451 # appends it to an existing element.
453 # The element name, attribute names, and attribute values can be
454 # either 8-bit ASCII strings or Unicode strings.
456 # @param parent The parent element.
457 # @param tag The subelement name.
458 # @param attrib An optional dictionary, containing element attributes.
459 # @param **extra Additional attributes, given as keyword arguments.
460 # @return An element instance.
463 def SubElement(parent, tag, attrib={}, **extra):
464 attrib = attrib.copy()
466 element = parent.makeelement(tag, attrib)
467 parent.append(element)
471 # Comment element factory. This factory function creates a special
472 # element that will be serialized as an XML comment.
474 # The comment string can be either an 8-bit ASCII string or a Unicode
477 # @param text A string containing the comment string.
478 # @return An element instance, representing a comment.
481 def Comment(text=None):
482 element = Element(Comment)
487 # PI element factory. This factory function creates a special element
488 # that will be serialized as an XML processing instruction.
490 # @param target A string containing the PI target.
491 # @param text A string containing the PI contents, if any.
492 # @return An element instance, representing a PI.
495 def ProcessingInstruction(target, text=None):
496 element = Element(ProcessingInstruction)
497 element.text = target
499 element.text = element.text + " " + text
502 PI = ProcessingInstruction
505 # QName wrapper. This can be used to wrap a QName attribute value, in
506 # order to get proper namespace handling on output.
508 # @param text A string containing the QName value, in the form {uri}local,
509 # or, if the tag argument is given, the URI part of a QName.
510 # @param tag Optional tag. If given, the first argument is interpreted as
511 # an URI, and this argument is interpreted as a local name.
512 # @return An opaque object, representing the QName.
515 def __init__(self, text_or_uri, tag=None):
517 text_or_uri = "{%s}%s" % (text_or_uri, tag)
518 self.text = text_or_uri
522 return hash(self.text)
523 def __cmp__(self, other):
524 if isinstance(other, QName):
525 return cmp(self.text, other.text)
526 return cmp(self.text, other)
529 # ElementTree wrapper class. This class represents an entire element
530 # hierarchy, and adds some extra support for serialization to and from
533 # @param element Optional root element.
534 # @keyparam file Optional file handle or name. If given, the
535 # tree is initialized with the contents of this XML file.
539 def __init__(self, element=None, file=None):
540 assert element is None or iselement(element)
541 self._root = element # first node
546 # Gets the root element for this tree.
548 # @return An element instance.
555 # Replaces the root element for this tree. This discards the
556 # current contents of the tree, and replaces it with the given
557 # element. Use with care.
559 # @param element An element instance.
561 def _setroot(self, element):
562 assert iselement(element)
566 # Loads an external XML document into this element tree.
568 # @param source A file name or file object.
569 # @param parser An optional parser instance. If not given, the
570 # standard {@link XMLTreeBuilder} parser is used.
571 # @return The document root element.
574 def parse(self, source, parser=None):
575 if not hasattr(source, "read"):
576 source = open(source, "rb")
578 parser = XMLTreeBuilder()
580 data = source.read(32768)
584 self._root = parser.close()
588 # Creates a tree iterator for the root element. The iterator loops
589 # over all elements in this tree, in document order.
591 # @param tag What tags to look for (default is to return all elements)
592 # @return An iterator.
593 # @defreturn iterator
595 def getiterator(self, tag=None):
596 assert self._root is not None
597 return self._root.getiterator(tag)
600 # Finds the first toplevel element with given tag.
601 # Same as getroot().find(path).
603 # @param path What element to look for.
604 # @return The first matching element, or None if no element was found.
605 # @defreturn Element or None
607 def find(self, path):
608 assert self._root is not None
611 return self._root.find(path)
614 # Finds the element text for the first toplevel element with given
615 # tag. Same as getroot().findtext(path).
617 # @param path What toplevel element to look for.
618 # @param default What to return if the element was not found.
619 # @return The text content of the first matching element, or the
620 # default value no element was found. Note that if the element
621 # has is found, but has no text content, this method returns an
625 def findtext(self, path, default=None):
626 assert self._root is not None
629 return self._root.findtext(path, default)
632 # Finds all toplevel elements with the given tag.
633 # Same as getroot().findall(path).
635 # @param path What element to look for.
636 # @return A list or iterator containing all matching elements,
638 # @defreturn list of Element instances
640 def findall(self, path):
641 assert self._root is not None
644 return self._root.findall(path)
647 # Writes the element tree to a file, as XML.
649 # @param file A file name, or a file object opened for writing.
650 # @param encoding Optional output encoding (default is US-ASCII).
652 def write(self, file, encoding="us-ascii"):
653 assert self._root is not None
654 if not hasattr(file, "write"):
655 file = open(file, "wb")
657 encoding = "us-ascii"
658 elif encoding != "utf-8" and encoding != "us-ascii":
659 file.write("<?xml version='1.0' encoding='%s'?>\n" % encoding)
660 self._write(file, self._root, encoding, {})
662 def _write(self, file, node, encoding, namespaces):
666 file.write("<!-- %s -->" % _escape_cdata(node.text, encoding))
667 elif tag is ProcessingInstruction:
668 file.write("<?%s?>" % _escape_cdata(node.text, encoding))
671 xmlns_items = [] # new namespaces in this scope
673 if isinstance(tag, QName) or tag[:1] == "{":
674 tag, xmlns = fixtag(tag, namespaces)
675 if xmlns: xmlns_items.append(xmlns)
677 _raise_serialization_error(tag)
678 file.write("<" + _encode(tag, encoding))
679 if items or xmlns_items:
680 items.sort() # lexical order
683 if isinstance(k, QName) or k[:1] == "{":
684 k, xmlns = fixtag(k, namespaces)
685 if xmlns: xmlns_items.append(xmlns)
687 _raise_serialization_error(k)
689 if isinstance(v, QName):
690 v, xmlns = fixtag(v, namespaces)
691 if xmlns: xmlns_items.append(xmlns)
693 _raise_serialization_error(v)
694 file.write(" %s=\"%s\"" % (_encode(k, encoding),
695 _escape_attrib(v, encoding)))
696 for k, v in xmlns_items:
697 file.write(" %s=\"%s\"" % (_encode(k, encoding),
698 _escape_attrib(v, encoding)))
699 if node.text or len(node):
702 file.write(_escape_cdata(node.text, encoding))
704 self._write(file, n, encoding, namespaces)
705 file.write("</" + _encode(tag, encoding) + ">")
708 for k, v in xmlns_items:
711 file.write(_escape_cdata(node.tail, encoding))
713 # --------------------------------------------------------------------
717 # Checks if an object appears to be a valid element object.
719 # @param An element instance.
720 # @return A true value if this is an element object.
723 def iselement(element):
724 # FIXME: not sure about this; might be a better idea to look
725 # for tag/attrib/text attributes
726 return isinstance(element, _ElementInterface) or hasattr(element, "tag")
729 # Writes an element tree or element structure to sys.stdout. This
730 # function should be used for debugging only.
732 # The exact output format is implementation dependent. In this
733 # version, it's written as an ordinary XML file.
735 # @param elem An element tree or an individual element.
739 if not isinstance(elem, ElementTree):
740 elem = ElementTree(elem)
741 elem.write(sys.stdout)
742 tail = elem.getroot().tail
743 if not tail or tail[-1] != "\n":
744 sys.stdout.write("\n")
746 def _encode(s, encoding):
748 return s.encode(encoding)
749 except AttributeError:
750 return s # 1.5.2: assume the string uses the right encoding
752 if sys.version[:3] == "1.5":
753 _escape = re.compile(r"[&<>\"\x80-\xff]+") # 1.5.2
755 _escape = re.compile(eval(r'u"[&<>\"\u0080-\uffff]+"'))
765 # "well-known" namespace prefixes
766 "http://www.w3.org/XML/1998/namespace": "xml",
767 "http://www.w3.org/1999/xhtml": "html",
768 "http://www.w3.org/1999/02/22-rdf-syntax-ns#": "rdf",
769 "http://schemas.xmlsoap.org/wsdl/": "wsdl",
772 def _raise_serialization_error(text):
774 "cannot serialize %r (type %s)" % (text, type(text).__name__)
777 def _encode_entity(text, pattern=_escape):
778 # map reserved and non-ascii characters to numerical entities
779 def escape_entities(m, map=_escape_map):
782 for char in m.group():
785 text = "&#%d;" % ord(char)
787 return string.join(out, "")
789 return _encode(pattern.sub(escape_entities, text), "ascii")
791 _raise_serialization_error(text)
794 # the following functions assume an ascii-compatible encoding
797 def _escape_cdata(text, encoding=None, replace=string.replace):
798 # escape character data
802 text = _encode(text, encoding)
804 return _encode_entity(text)
805 text = replace(text, "&", "&")
806 text = replace(text, "<", "<")
807 text = replace(text, ">", ">")
809 except (TypeError, AttributeError):
810 _raise_serialization_error(text)
812 def _escape_attrib(text, encoding=None, replace=string.replace):
813 # escape attribute value
817 text = _encode(text, encoding)
819 return _encode_entity(text)
820 text = replace(text, "&", "&")
821 text = replace(text, "'", "'") # FIXME: overkill
822 text = replace(text, "\"", """)
823 text = replace(text, "<", "<")
824 text = replace(text, ">", ">")
826 except (TypeError, AttributeError):
827 _raise_serialization_error(text)
829 def fixtag(tag, namespaces):
830 # given a decorated tag (of the form {uri}tag), return prefixed
831 # tag and namespace declaration, if any
832 if isinstance(tag, QName):
834 namespace_uri, tag = string.split(tag[1:], "}", 1)
835 prefix = namespaces.get(namespace_uri)
837 prefix = _namespace_map.get(namespace_uri)
839 prefix = "ns%d" % len(namespaces)
840 namespaces[namespace_uri] = prefix
844 xmlns = ("xmlns:%s" % prefix, namespace_uri)
847 return "%s:%s" % (prefix, tag), xmlns
850 # Parses an XML document into an element tree.
852 # @param source A filename or file object containing XML data.
853 # @param parser An optional parser instance. If not given, the
854 # standard {@link XMLTreeBuilder} parser is used.
855 # @return An ElementTree instance
857 def parse(source, parser=None):
859 tree.parse(source, parser)
863 # Parses an XML document into an element tree incrementally, and reports
864 # what's going on to the user.
866 # @param source A filename or file object containing XML data.
867 # @param events A list of events to report back. If omitted, only "end"
868 # events are reported.
869 # @return A (event, elem) iterator.
873 def __init__(self, source, events=None):
874 if not hasattr(source, "read"):
875 source = open(source, "rb")
879 self.root = self._root = None
880 self._parser = XMLTreeBuilder()
881 # wire up the parser for event reporting
882 parser = self._parser._parser
883 append = self._events.append
889 parser.ordered_attributes = 1
890 parser.specified_attributes = 1
891 def handler(tag, attrib_in, event=event, append=append,
892 start=self._parser._start_list):
893 append((event, start(tag, attrib_in)))
894 parser.StartElementHandler = handler
895 except AttributeError:
896 def handler(tag, attrib_in, event=event, append=append,
897 start=self._parser._start):
898 append((event, start(tag, attrib_in)))
899 parser.StartElementHandler = handler
901 def handler(tag, event=event, append=append,
902 end=self._parser._end):
903 append((event, end(tag)))
904 parser.EndElementHandler = handler
905 elif event == "start-ns":
906 def handler(prefix, uri, event=event, append=append):
908 uri = _encode(uri, "ascii")
911 append((event, (prefix or "", uri)))
912 parser.StartNamespaceDeclHandler = handler
913 elif event == "end-ns":
914 def handler(prefix, event=event, append=append):
915 append((event, None))
916 parser.EndNamespaceDeclHandler = handler
921 item = self._events[self._index]
923 if self._parser is None:
924 self.root = self._root
932 data = self._file.read(16384)
934 self._parser.feed(data)
936 self._root = self._parser.close()
939 self._index = self._index + 1
947 def __getitem__(self, index):
951 # Parses an XML document from a string constant. This function can
952 # be used to embed "XML literals" in Python code.
954 # @param source A string containing XML data.
955 # @return An Element instance.
959 parser = XMLTreeBuilder()
961 return parser.close()
964 # Parses an XML document from a string constant, and also returns
965 # a dictionary which maps from element id:s to elements.
967 # @param source A string containing XML data.
968 # @return A tuple containing an Element instance and a dictionary.
969 # @defreturn (Element, dictionary)
972 parser = XMLTreeBuilder()
974 tree = parser.close()
976 for elem in tree.getiterator():
983 # Parses an XML document from a string constant. Same as {@link #XML}.
985 # @def fromstring(text)
986 # @param source A string containing XML data.
987 # @return An Element instance.
993 # Generates a string representation of an XML element, including all
996 # @param element An Element instance.
997 # @return An encoded string containing the XML data.
1000 def tostring(element, encoding=None):
1005 file.write = data.append
1006 ElementTree(element).write(file, encoding)
1007 return string.join(data, "")
1010 # Generic element structure builder. This builder converts a sequence
1011 # of {@link #TreeBuilder.start}, {@link #TreeBuilder.data}, and {@link
1012 # #TreeBuilder.end} method calls to a well-formed element structure.
1014 # You can use this class to build an element structure using a custom XML
1015 # parser, or a parser for some other XML-like format.
1017 # @param element_factory Optional element factory. This factory
1018 # is called to create new Element instances, as necessary.
1022 def __init__(self, element_factory=None):
1023 self._data = [] # data collector
1024 self._elem = [] # element stack
1025 self._last = None # last element
1026 self._tail = None # true if we're after an end tag
1027 if element_factory is None:
1028 element_factory = _ElementInterface
1029 self._factory = element_factory
1032 # Flushes the parser buffers, and returns the toplevel documen
1035 # @return An Element instance.
1036 # @defreturn Element
1039 assert len(self._elem) == 0, "missing end tags"
1040 assert self._last != None, "missing toplevel element"
1045 if self._last is not None:
1046 text = string.join(self._data, "")
1048 assert self._last.tail is None, "internal error (tail)"
1049 self._last.tail = text
1051 assert self._last.text is None, "internal error (text)"
1052 self._last.text = text
1056 # Adds text to the current element.
1058 # @param data A string. This should be either an 8-bit string
1059 # containing ASCII text, or a Unicode string.
1061 def data(self, data):
1062 self._data.append(data)
1065 # Opens a new element.
1067 # @param tag The element name.
1068 # @param attrib A dictionary containing element attributes.
1069 # @return The opened element.
1070 # @defreturn Element
1072 def start(self, tag, attrs):
1074 self._last = elem = self._factory(tag, attrs)
1076 self._elem[-1].append(elem)
1077 self._elem.append(elem)
1082 # Closes the current element.
1084 # @param tag The element name.
1085 # @return The closed element.
1086 # @defreturn Element
1090 self._last = self._elem.pop()
1091 assert self._last.tag == tag,\
1092 "end tag mismatch (expected %s, got %s)" % (
1093 self._last.tag, tag)
1098 # Element structure builder for XML source data, based on the
1099 # <b>expat</b> parser.
1101 # @keyparam target Target object. If omitted, the builder uses an
1102 # instance of the standard {@link #TreeBuilder} class.
1103 # @keyparam html Predefine HTML entities. This flag is not supported
1104 # by the current implementation.
1108 class XMLTreeBuilder:
1110 def __init__(self, html=0, target=None):
1112 from xml.parsers import expat
1115 "No module named expat; use SimpleXMLTreeBuilder instead"
1117 self._parser = parser = expat.ParserCreate(None, "}")
1119 target = TreeBuilder()
1120 self._target = target
1121 self._names = {} # name memo cache
1123 parser.DefaultHandlerExpand = self._default
1124 parser.StartElementHandler = self._start
1125 parser.EndElementHandler = self._end
1126 parser.CharacterDataHandler = self._data
1127 # let expat do the buffering, if supported
1129 self._parser.buffer_text = 1
1130 except AttributeError:
1132 # use new-style attribute handling, if supported
1134 self._parser.ordered_attributes = 1
1135 self._parser.specified_attributes = 1
1136 parser.StartElementHandler = self._start_list
1137 except AttributeError:
1140 if not parser.returns_unicode:
1142 # target.xml(encoding, None)
1143 self._doctype = None
1146 def _fixtext(self, text):
1147 # convert text string to ascii, if possible
1149 return _encode(text, "ascii")
1150 except UnicodeError:
1153 def _fixname(self, key):
1154 # expand qname, and convert name string to ascii, if possible
1156 name = self._names[key]
1161 self._names[key] = name = self._fixtext(name)
1164 def _start(self, tag, attrib_in):
1165 fixname = self._fixname
1168 for key, value in attrib_in.items():
1169 attrib[fixname(key)] = self._fixtext(value)
1170 return self._target.start(tag, attrib)
1172 def _start_list(self, tag, attrib_in):
1173 fixname = self._fixname
1177 for i in range(0, len(attrib_in), 2):
1178 attrib[fixname(attrib_in[i])] = self._fixtext(attrib_in[i+1])
1179 return self._target.start(tag, attrib)
1181 def _data(self, text):
1182 return self._target.data(self._fixtext(text))
1184 def _end(self, tag):
1185 return self._target.end(self._fixname(tag))
1187 def _default(self, text):
1190 # deal with undefined entities
1192 self._target.data(self.entity[text[1:-1]])
1194 from xml.parsers import expat
1196 "undefined entity %s: line %d, column %d" %
1197 (text, self._parser.ErrorLineNumber,
1198 self._parser.ErrorColumnNumber)
1200 elif prefix == "<" and text[:9] == "<!DOCTYPE":
1201 self._doctype = [] # inside a doctype declaration
1202 elif self._doctype is not None:
1203 # parse doctype contents
1205 self._doctype = None
1207 text = string.strip(text)
1210 self._doctype.append(text)
1211 n = len(self._doctype)
1213 type = self._doctype[1]
1214 if type == "PUBLIC" and n == 4:
1215 name, type, pubid, system = self._doctype
1216 elif type == "SYSTEM" and n == 3:
1217 name, type, system = self._doctype
1223 self.doctype(name, pubid, system[1:-1])
1224 self._doctype = None
1227 # Handles a doctype declaration.
1229 # @param name Doctype name.
1230 # @param pubid Public identifier.
1231 # @param system System identifier.
1233 def doctype(self, name, pubid, system):
1237 # Feeds data to the parser.
1239 # @param data Encoded data.
1241 def feed(self, data):
1242 self._parser.Parse(data, 0)
1245 # Finishes feeding data to the parser.
1247 # @return An element structure.
1248 # @defreturn Element
1251 self._parser.Parse("", 1) # end of data
1252 tree = self._target.close()
1253 del self._target, self._parser # get rid of circular references