Python xml.parsers.expat.ParserCreate() Examples

def test_ignore_use_foreign_dtd(self):
        """
        If UseForeignDTD is passed True and a document with an external
        entity reference is parsed, ExternalEntityRefHandler is called with
        the public and system ids from the document.
        """
        handler_call_args = []
        def resolve_entity(context, base, system_id, public_id):
            handler_call_args.append((public_id, system_id))
            return 1

        parser = expat.ParserCreate()
        parser.UseForeignDTD(True)
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse(
            "<?xml version='1.0'?><!DOCTYPE foo PUBLIC 'bar' 'baz'><element/>")
        self.assertEqual(handler_call_args, [("bar", "baz")]) 

def test_unchanged_size(self):
        xml1 = ("<?xml version='1.0' encoding='iso8859'?><s>%s" % ('a' * 512))
        xml2 = 'a'*512 + '</s>'
        parser = expat.ParserCreate()
        parser.CharacterDataHandler = self.counting_handler
        parser.buffer_size = 512
        parser.buffer_text = 1

        # Feed 512 bytes of character data: the handler should be called
        # once.
        self.n = 0
        parser.Parse(xml1)
        self.assertEqual(self.n, 1)

        # Reassign to buffer_size, but assign the same size.
        parser.buffer_size = parser.buffer_size
        self.assertEqual(self.n, 1)

        # Try parsing rest of the document
        parser.Parse(xml2)
        self.assertEqual(self.n, 2) 

def test_ignore_use_foreign_dtd(self):
        """
        If UseForeignDTD is passed True and a document with an external
        entity reference is parsed, ExternalEntityRefHandler is called with
        the public and system ids from the document.
        """
        handler_call_args = []
        def resolve_entity(context, base, system_id, public_id):
            handler_call_args.append((public_id, system_id))
            return 1

        parser = expat.ParserCreate()
        parser.UseForeignDTD(True)
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse(
            "<?xml version='1.0'?><!DOCTYPE foo PUBLIC 'bar' 'baz'><element/>")
        self.assertEqual(handler_call_args, [("bar", "baz")]) 

def test_parse_only_xml_data(self):
        # http://python.org/sf/1296433
        #
        xml = "<?xml version='1.0' encoding='iso8859'?><s>%s</s>" % ('a' * 1025)
        # this one doesn't crash
        #xml = "<?xml version='1.0'?><s>%s</s>" % ('a' * 10000)

        class SpecificException(Exception):
            pass

        def handler(text):
            raise SpecificException

        parser = expat.ParserCreate()
        parser.CharacterDataHandler = handler

        # https://github.com/IronLanguages/ironpython2/issues/464
        if sys.platform == 'cli':
            self.assertRaises(Exception, parser.Parse, xml, True)
        else:
            self.assertRaises(Exception, parser.Parse, xml) 

def loadXML(self, fileName):
        """
        Loads the crawler parameters from an XML file.
        @param fileName The file from where is loaded the crawler data
        """
        self._parser = expat.ParserCreate("UTF-8")
        self._parser.StartElementHandler = self.__start_element
        self._parser.EndElementHandler = self.__end_element
        self._parser.CharacterDataHandler = self.__char_data
        self._parser.returns_unicode = False

        f = None
        try:
            f = open(fileName)
            content = f.read()
            self.__feed(content.replace("\n", ""))
        finally:
            if f is not None:
                f.close() 

def test_unchanged_size(self):
        xml1 = ("<?xml version='1.0' encoding='iso8859'?><s>%s" % ('a' * 512))
        xml2 = 'a'*512 + '</s>'
        parser = expat.ParserCreate()
        parser.CharacterDataHandler = self.counting_handler
        parser.buffer_size = 512
        parser.buffer_text = 1

        # Feed 512 bytes of character data: the handler should be called
        # once.
        self.n = 0
        parser.Parse(xml1)
        self.assertEqual(self.n, 1)

        # Reassign to buffer_size, but assign the same size.
        parser.buffer_size = parser.buffer_size
        self.assertEqual(self.n, 1)

        # Try parsing rest of the document
        parser.Parse(xml2)
        self.assertEqual(self.n, 2) 

def loadXML(self, fileName):
        """
        Loads the crawler parameters from an XML file.
        @param fileName The file from where is loaded the crawler data
        """
        self._parser = expat.ParserCreate("UTF-8")
        self._parser.StartElementHandler = self.__start_element
        self._parser.EndElementHandler = self.__end_element
        self._parser.CharacterDataHandler = self.__char_data
        self._parser.returns_unicode = False

        f = None
        try:
            f = open(fileName)
            content = f.read()
            self.__feed(content.replace("\n", ""))
        finally:
            if f is not None:
                f.close() 

def test_change_size_1(self):
        xml1 = "<?xml version='1.0' encoding='iso8859'?><a><s>%s" % ('a' * 1024)
        xml2 = "aaa</s><s>%s</s></a>" % ('a' * 1025)
        parser = expat.ParserCreate()
        parser.CharacterDataHandler = self.counting_handler
        parser.buffer_text = 1
        parser.buffer_size = 1024
        self.assertEqual(parser.buffer_size, 1024)

        self.n = 0
        parser.Parse(xml1, 0)
        parser.buffer_size *= 2
        self.assertEqual(parser.buffer_size, 2048)
        parser.Parse(xml2, 1)
        self.assertEqual(self.n, 2) 

def reset(self):
        if self._namespaces:
            self._parser = expat.ParserCreate(self._source.getEncoding(), " ",
                                              intern=self._interning)
            self._parser.namespace_prefixes = 1
            self._parser.StartElementHandler = self.start_element_ns
            self._parser.EndElementHandler = self.end_element_ns
        else:
            self._parser = expat.ParserCreate(self._source.getEncoding(),
                                              intern = self._interning)
            self._parser.StartElementHandler = self.start_element
            self._parser.EndElementHandler = self.end_element

        self._reset_cont_handler()
        self._parser.UnparsedEntityDeclHandler = self.unparsed_entity_decl
        self._parser.NotationDeclHandler = self.notation_decl
        self._parser.StartNamespaceDeclHandler = self.start_namespace_decl
        self._parser.EndNamespaceDeclHandler = self.end_namespace_decl

        self._decl_handler_prop = None
        if self._lex_handler_prop:
            self._reset_lex_handler_prop()
#         self._parser.DefaultHandler =
#         self._parser.DefaultHandlerExpand =
#         self._parser.NotStandaloneHandler =
        self._parser.ExternalEntityRefHandler = self.external_entity_ref
        try:
            self._parser.SkippedEntityHandler = self.skipped_entity_handler
        except AttributeError:
            # This pyexpat does not support SkippedEntity
            pass
        self._parser.SetParamEntityParsing(
            expat.XML_PARAM_ENTITY_PARSING_UNLESS_STANDALONE)

        self._parsing = 0
        self._entity_stack = []

    # Locator methods 

def parse(self, fileobj):
        from xml.parsers.expat import ParserCreate
        parser = ParserCreate()
        parser.StartElementHandler = self.handleBeginElement
        parser.EndElementHandler = self.handleEndElement
        parser.CharacterDataHandler = self.handleData
        parser.ParseFile(fileobj)
        return self.root 

def test_utf8(self):

        out = self.Outputter()
        parser = expat.ParserCreate(namespace_separator='!')
        for name in self.handler_names:
            setattr(parser, name, getattr(out, name))
        parser.returns_unicode = 0
        parser.Parse(data, 1)

        # Verify output
        op = out.out
        self.assertEqual(op[0], 'PI: \'xml-stylesheet\' \'href="stylesheet.css"\'')
        self.assertEqual(op[1], "Comment: ' comment data '")
        self.assertEqual(op[2], "Notation declared: ('notation', None, 'notation.jpeg', None)")
        self.assertEqual(op[3], "Unparsed entity decl: ('unparsed_entity', None, 'entity.file', None, 'notation')")
        self.assertEqual(op[4], "Start element: 'root' {'attr1': 'value1', 'attr2': 'value2\\xe1\\xbd\\x80'}")
        self.assertEqual(op[5], "NS decl: 'myns' 'http://www.python.org/namespace'")
        self.assertEqual(op[6], "Start element: 'http://www.python.org/namespace!subelement' {}")
        self.assertEqual(op[7], "Character data: 'Contents of subelements'")
        self.assertEqual(op[8], "End element: 'http://www.python.org/namespace!subelement'")
        self.assertEqual(op[9], "End of NS decl: 'myns'")
        self.assertEqual(op[10], "Start element: 'sub2' {}")
        self.assertEqual(op[11], 'Start of CDATA section')
        self.assertEqual(op[12], "Character data: 'contents of CDATA section'")
        self.assertEqual(op[13], 'End of CDATA section')
        self.assertEqual(op[14], "End element: 'sub2'")
        self.assertEqual(op[15], "External entity ref: (None, 'entity.file', None)")
        self.assertEqual(op[16], "End element: 'root'") 

def test_unicode(self):
        # Try the parse again, this time producing Unicode output
        out = self.Outputter()
        parser = expat.ParserCreate(namespace_separator='!')
        parser.returns_unicode = 1
        for name in self.handler_names:
            setattr(parser, name, getattr(out, name))

        parser.Parse(data, 1)

        op = out.out
        self.assertEqual(op[0], 'PI: u\'xml-stylesheet\' u\'href="stylesheet.css"\'')
        self.assertEqual(op[1], "Comment: u' comment data '")
        self.assertEqual(op[2], "Notation declared: (u'notation', None, u'notation.jpeg', None)")
        self.assertEqual(op[3], "Unparsed entity decl: (u'unparsed_entity', None, u'entity.file', None, u'notation')")
        self.assertEqual(op[4], "Start element: u'root' {u'attr1': u'value1', u'attr2': u'value2\\u1f40'}")
        self.assertEqual(op[5], "NS decl: u'myns' u'http://www.python.org/namespace'")
        self.assertEqual(op[6], "Start element: u'http://www.python.org/namespace!subelement' {}")
        self.assertEqual(op[7], "Character data: u'Contents of subelements'")
        self.assertEqual(op[8], "End element: u'http://www.python.org/namespace!subelement'")
        self.assertEqual(op[9], "End of NS decl: u'myns'")
        self.assertEqual(op[10], "Start element: u'sub2' {}")
        self.assertEqual(op[11], 'Start of CDATA section')
        self.assertEqual(op[12], "Character data: u'contents of CDATA section'")
        self.assertEqual(op[13], 'End of CDATA section')
        self.assertEqual(op[14], "End element: u'sub2'")
        self.assertEqual(op[15], "External entity ref: (None, u'entity.file', None)")
        self.assertEqual(op[16], "End element: u'root'") 

def test_use_foreign_dtd(self):
        """
        If UseForeignDTD is passed True and a document without an external
        entity reference is parsed, ExternalEntityRefHandler is first called
        with None for the public and system ids.
        """
        handler_call_args = []
        def resolve_entity(context, base, system_id, public_id):
            handler_call_args.append((public_id, system_id))
            return 1

        parser = expat.ParserCreate()
        parser.UseForeignDTD(True)
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse("<?xml version='1.0'?><element/>")
        self.assertEqual(handler_call_args, [(None, None)])

        # test UseForeignDTD() is equal to UseForeignDTD(True)
        handler_call_args[:] = []

        parser = expat.ParserCreate()
        parser.UseForeignDTD()
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse("<?xml version='1.0'?><element/>")
        self.assertEqual(handler_call_args, [(None, None)]) 

def small_buffer_test(self, buffer_len):
        xml = "<?xml version='1.0' encoding='iso8859'?><s>%s</s>" % ('a' * buffer_len)
        parser = expat.ParserCreate()
        parser.CharacterDataHandler = self.counting_handler
        parser.buffer_size = 1024
        parser.buffer_text = 1

        self.n = 0
        parser.Parse(xml)
        return self.n 

def createParser(self):
        """Create a new namespace-handling parser."""
        parser = expat.ParserCreate(namespace_separator=" ")
        parser.namespace_prefixes = True
        return parser 

def test1(self):
        xml = "\0\r\n"
        parser = expat.ParserCreate()
        try:
            parser.Parse(xml, True)
            self.fail()
        except expat.ExpatError as e:
            self.assertEqual(str(e), 'unclosed token: line 2, column 0') 

def test2(self):
        xml = "<?xml version\xc2\x85='1.0'?>\r\n"
        parser = expat.ParserCreate()
        try:
            parser.Parse(xml, True)
            self.fail()
        except expat.ExpatError as e:
            self.assertEqual(str(e), 'XML declaration not well-formed: line 1, column 14') 

def createParser(self):
        """Create a new parser object."""
        return expat.ParserCreate() 

def test_illegal(self):
        try:
            expat.ParserCreate(namespace_separator=42)
            self.fail()
        except TypeError, e:
            self.assertEqual(str(e),
                'ParserCreate() argument 2 must be string or None, not int') 

def createParser(self):
        """Create a new namespace-handling parser."""
        parser = expat.ParserCreate(namespace_separator=" ")
        parser.namespace_prefixes = True
        return parser 

def createParser(self):
        """Create a new parser object."""
        return expat.ParserCreate() 

def _get_plist(self, s):
        p = {}
        try:
            if sys.version_info >= (3, 0):
                p = plistlib.loads(s.encode("utf-8"))
            else:
                # p = plistlib.readPlistFromString(s)
                # We avoid using readPlistFromString() as that uses
                # cStringIO and fails when Unicode strings are detected
                # Don't subclass - keep the parser local
                from xml.parsers.expat import ParserCreate
                # Create a new PlistParser object - then we need to set up
                # the values and parse.
                pa = plistlib.PlistParser()
                # We also monkey patch this to encode unicode as utf-8
                def end_string():
                    d = pa.getData()
                    if isinstance(d,unicode):
                        d = d.encode("utf-8")
                    pa.addObject(d)
                pa.end_string = end_string
                parser = ParserCreate()
                parser.StartElementHandler = pa.handleBeginElement
                parser.EndElementHandler = pa.handleEndElement
                parser.CharacterDataHandler = pa.handleData
                if isinstance(s, unicode):
                    # Encode unicode -> string; use utf-8 for safety
                    s = s.encode("utf-8")
                # Parse the string
                parser.Parse(s, 1)
                p = pa.root
        except Exception as e:
            print(e)
            pass
        return p 

def __init__(self):
        self._parser = expat.ParserCreate()
        self._parser.StartElementHandler = self.start_element
        self._parser.EndElementHandler = self.end_element
        self._parser.CharacterDataHandler = self.char_data
        self.vulnerabilities = []
        self.vul = None
        self.references = {}
        self.title = ""
        self.url = ""
        self.tag = "" 

def __init__(self):
        self._parser = expat.ParserCreate()
        self._parser.StartElementHandler = self.start_element
        self._parser.EndElementHandler = self.end_element
        self._parser.CharacterDataHandler = self.char_data
        self.anomalies = []
        self.anom = None
        self.references = {}
        self.title = ""
        self.url = ""
        self.tag = "" 

def __init__(self):
        self._parser = expat.ParserCreate()
        self._parser.StartElementHandler = self.start_element
        self._parser.EndElementHandler = self.end_element
        self._parser.CharacterDataHandler = self.char_data
        self.reportGenerators = []
        self.repGen = None
        self.tag = "" 

def __init__(self, html=0, target=None):
        try:
            from xml.parsers import expat
        except ImportError:
            raise ImportError(
                "No module named expat; use SimpleXMLTreeBuilder instead"
            )
        self._parser = parser = expat.ParserCreate(None, "}")
        if target is None:
            target = TreeBuilder()
        self._target = target
        self._names = {}  # name memo cache
        # callbacks
        parser.DefaultHandlerExpand = self._default
        parser.StartElementHandler = self._start
        parser.EndElementHandler = self._end
        parser.CharacterDataHandler = self._data
        # let expat do the buffering, if supported
        try:
            self._parser.buffer_text = 1
        except AttributeError:
            pass
        # use new-style attribute handling, if supported
        try:
            self._parser.ordered_attributes = 1
            self._parser.specified_attributes = 1
            parser.StartElementHandler = self._start_list
        except AttributeError:
            pass
        encoding = None
        if hasattr(parser, "returns_unicode") and not parser.returns_unicode:
            encoding = "utf-8"
        # target.xml(encoding, None)
        self._doctype = None
        self.entity = {} 

def parse(xml_input, encoding=None, expat=expat, process_namespaces=False,
          namespace_separator=':', **kwargs):
    handler = _DictSAXHandler(namespace_separator=namespace_separator,
                              **kwargs)
    if isinstance(xml_input, _unicode):
        if not encoding:
            encoding = 'utf-8'
        xml_input = xml_input.encode(encoding)
    if not process_namespaces:
        namespace_separator = None
    parser = expat.ParserCreate(
        encoding,
        namespace_separator
    )
    try:
        parser.ordered_attributes = True
    except AttributeError:
        # Jython's expat does not support ordered_attributes
        pass
    parser.StartElementHandler = handler.startElement
    parser.EndElementHandler = handler.endElement
    parser.CharacterDataHandler = handler.characters
    parser.buffer_text = True
    try:
        parser.ParseFile(xml_input)
    except (TypeError, AttributeError):
        parser.Parse(xml_input, True)
    return handler.item 

def test_use_foreign_dtd(self):
        """
        If UseForeignDTD is passed True and a document without an external
        entity reference is parsed, ExternalEntityRefHandler is first called
        with None for the public and system ids.
        """
        handler_call_args = []
        def resolve_entity(context, base, system_id, public_id):
            handler_call_args.append((public_id, system_id))
            return 1

        parser = expat.ParserCreate()
        parser.UseForeignDTD(True)
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse("<?xml version='1.0'?><element/>")
        self.assertEqual(handler_call_args, [(None, None)])

        # test UseForeignDTD() is equal to UseForeignDTD(True)
        handler_call_args[:] = []

        parser = expat.ParserCreate()
        parser.UseForeignDTD()
        parser.SetParamEntityParsing(expat.XML_PARAM_ENTITY_PARSING_ALWAYS)
        parser.ExternalEntityRefHandler = resolve_entity
        parser.Parse("<?xml version='1.0'?><element/>")
        self.assertEqual(handler_call_args, [(None, None)]) 

def test2(self):
        xml = "<?xml version\xc2\x85='1.0'?>\r\n"
        parser = expat.ParserCreate()
        err_pattern = r'XML declaration not well-formed: line 1, column \d+'
        with self.assertRaisesRegexp(expat.ExpatError, err_pattern):
            parser.Parse(xml, True) 

def test1(self):
        xml = "\0\r\n"
        parser = expat.ParserCreate()
        try:
            parser.Parse(xml, True)
            self.fail()
        except expat.ExpatError as e:
            self.assertEqual(str(e), 'unclosed token: line 2, column 0')