python-3.6.zip added from Github

README.cosmo contains the necessary links.

third_party/python/Lib/xml/dom/NodeFilter.py

@@ -0,0 +1,27 @@
+# This is the Python mapping for interface NodeFilter from
+# DOM2-Traversal-Range. It contains only constants.
+
+class NodeFilter:
+    """
+    This is the DOM2 NodeFilter interface. It contains only constants.
+    """
+    FILTER_ACCEPT = 1
+    FILTER_REJECT = 2
+    FILTER_SKIP   = 3
+
+    SHOW_ALL                    = 0xFFFFFFFF
+    SHOW_ELEMENT                = 0x00000001
+    SHOW_ATTRIBUTE              = 0x00000002
+    SHOW_TEXT                   = 0x00000004
+    SHOW_CDATA_SECTION          = 0x00000008
+    SHOW_ENTITY_REFERENCE       = 0x00000010
+    SHOW_ENTITY                 = 0x00000020
+    SHOW_PROCESSING_INSTRUCTION = 0x00000040
+    SHOW_COMMENT                = 0x00000080
+    SHOW_DOCUMENT               = 0x00000100
+    SHOW_DOCUMENT_TYPE          = 0x00000200
+    SHOW_DOCUMENT_FRAGMENT      = 0x00000400
+    SHOW_NOTATION               = 0x00000800
+
+    def acceptNode(self, node):
+        raise NotImplementedError

third_party/python/Lib/xml/dom/__init__.py

@@ -0,0 +1,140 @@
+"""W3C Document Object Model implementation for Python.
+
+The Python mapping of the Document Object Model is documented in the
+Python Library Reference in the section on the xml.dom package.
+
+This package contains the following modules:
+
+minidom -- A simple implementation of the Level 1 DOM with namespace
+           support added (based on the Level 2 specification) and other
+           minor Level 2 functionality.
+
+pulldom -- DOM builder supporting on-demand tree-building for selected
+           subtrees of the document.
+
+"""
+
+
+class Node:
+    """Class giving the NodeType constants."""
+    __slots__ = ()
+
+    # DOM implementations may use this as a base class for their own
+    # Node implementations.  If they don't, the constants defined here
+    # should still be used as the canonical definitions as they match
+    # the values given in the W3C recommendation.  Client code can
+    # safely refer to these values in all tests of Node.nodeType
+    # values.
+
+    ELEMENT_NODE                = 1
+    ATTRIBUTE_NODE              = 2
+    TEXT_NODE                   = 3
+    CDATA_SECTION_NODE          = 4
+    ENTITY_REFERENCE_NODE       = 5
+    ENTITY_NODE                 = 6
+    PROCESSING_INSTRUCTION_NODE = 7
+    COMMENT_NODE                = 8
+    DOCUMENT_NODE               = 9
+    DOCUMENT_TYPE_NODE          = 10
+    DOCUMENT_FRAGMENT_NODE      = 11
+    NOTATION_NODE               = 12
+
+
+#ExceptionCode
+INDEX_SIZE_ERR                 = 1
+DOMSTRING_SIZE_ERR             = 2
+HIERARCHY_REQUEST_ERR          = 3
+WRONG_DOCUMENT_ERR             = 4
+INVALID_CHARACTER_ERR          = 5
+NO_DATA_ALLOWED_ERR            = 6
+NO_MODIFICATION_ALLOWED_ERR    = 7
+NOT_FOUND_ERR                  = 8
+NOT_SUPPORTED_ERR              = 9
+INUSE_ATTRIBUTE_ERR            = 10
+INVALID_STATE_ERR              = 11
+SYNTAX_ERR                     = 12
+INVALID_MODIFICATION_ERR       = 13
+NAMESPACE_ERR                  = 14
+INVALID_ACCESS_ERR             = 15
+VALIDATION_ERR                 = 16
+
+
+class DOMException(Exception):
+    """Abstract base class for DOM exceptions.
+    Exceptions with specific codes are specializations of this class."""
+
+    def __init__(self, *args, **kw):
+        if self.__class__ is DOMException:
+            raise RuntimeError(
+                "DOMException should not be instantiated directly")
+        Exception.__init__(self, *args, **kw)
+
+    def _get_code(self):
+        return self.code
+
+
+class IndexSizeErr(DOMException):
+    code = INDEX_SIZE_ERR
+
+class DomstringSizeErr(DOMException):
+    code = DOMSTRING_SIZE_ERR
+
+class HierarchyRequestErr(DOMException):
+    code = HIERARCHY_REQUEST_ERR
+
+class WrongDocumentErr(DOMException):
+    code = WRONG_DOCUMENT_ERR
+
+class InvalidCharacterErr(DOMException):
+    code = INVALID_CHARACTER_ERR
+
+class NoDataAllowedErr(DOMException):
+    code = NO_DATA_ALLOWED_ERR
+
+class NoModificationAllowedErr(DOMException):
+    code = NO_MODIFICATION_ALLOWED_ERR
+
+class NotFoundErr(DOMException):
+    code = NOT_FOUND_ERR
+
+class NotSupportedErr(DOMException):
+    code = NOT_SUPPORTED_ERR
+
+class InuseAttributeErr(DOMException):
+    code = INUSE_ATTRIBUTE_ERR
+
+class InvalidStateErr(DOMException):
+    code = INVALID_STATE_ERR
+
+class SyntaxErr(DOMException):
+    code = SYNTAX_ERR
+
+class InvalidModificationErr(DOMException):
+    code = INVALID_MODIFICATION_ERR
+
+class NamespaceErr(DOMException):
+    code = NAMESPACE_ERR
+
+class InvalidAccessErr(DOMException):
+    code = INVALID_ACCESS_ERR
+
+class ValidationErr(DOMException):
+    code = VALIDATION_ERR
+
+class UserDataHandler:
+    """Class giving the operation constants for UserDataHandler.handle()."""
+
+    # Based on DOM Level 3 (WD 9 April 2002)
+
+    NODE_CLONED   = 1
+    NODE_IMPORTED = 2
+    NODE_DELETED  = 3
+    NODE_RENAMED  = 4
+
+XML_NAMESPACE = "http://www.w3.org/XML/1998/namespace"
+XMLNS_NAMESPACE = "http://www.w3.org/2000/xmlns/"
+XHTML_NAMESPACE = "http://www.w3.org/1999/xhtml"
+EMPTY_NAMESPACE = None
+EMPTY_PREFIX = None
+
+from .domreg import getDOMImplementation, registerDOMImplementation

third_party/python/Lib/xml/dom/domreg.py

@@ -0,0 +1,99 @@
+"""Registration facilities for DOM. This module should not be used
+directly. Instead, the functions getDOMImplementation and
+registerDOMImplementation should be imported from xml.dom."""
+
+# This is a list of well-known implementations.  Well-known names
+# should be published by posting to xml-sig@python.org, and are
+# subsequently recorded in this file.
+
+import sys
+
+well_known_implementations = {
+    'minidom':'xml.dom.minidom',
+    '4DOM': 'xml.dom.DOMImplementation',
+    }
+
+# DOM implementations not officially registered should register
+# themselves with their
+
+registered = {}
+
+def registerDOMImplementation(name, factory):
+    """registerDOMImplementation(name, factory)
+
+    Register the factory function with the name. The factory function
+    should return an object which implements the DOMImplementation
+    interface. The factory function can either return the same object,
+    or a new one (e.g. if that implementation supports some
+    customization)."""
+
+    registered[name] = factory
+
+def _good_enough(dom, features):
+    "_good_enough(dom, features) -> Return 1 if the dom offers the features"
+    for f,v in features:
+        if not dom.hasFeature(f,v):
+            return 0
+    return 1
+
+def getDOMImplementation(name=None, features=()):
+    """getDOMImplementation(name = None, features = ()) -> DOM implementation.
+
+    Return a suitable DOM implementation. The name is either
+    well-known, the module name of a DOM implementation, or None. If
+    it is not None, imports the corresponding module and returns
+    DOMImplementation object if the import succeeds.
+
+    If name is not given, consider the available implementations to
+    find one with the required feature set. If no implementation can
+    be found, raise an ImportError. The features list must be a sequence
+    of (feature, version) pairs which are passed to hasFeature."""
+
+    import os
+    creator = None
+    mod = well_known_implementations.get(name)
+    if mod:
+        mod = __import__(mod, {}, {}, ['getDOMImplementation'])
+        return mod.getDOMImplementation()
+    elif name:
+        return registered[name]()
+    elif not sys.flags.ignore_environment and "PYTHON_DOM" in os.environ:
+        return getDOMImplementation(name = os.environ["PYTHON_DOM"])
+
+    # User did not specify a name, try implementations in arbitrary
+    # order, returning the one that has the required features
+    if isinstance(features, str):
+        features = _parse_feature_string(features)
+    for creator in registered.values():
+        dom = creator()
+        if _good_enough(dom, features):
+            return dom
+
+    for creator in well_known_implementations.keys():
+        try:
+            dom = getDOMImplementation(name = creator)
+        except Exception: # typically ImportError, or AttributeError
+            continue
+        if _good_enough(dom, features):
+            return dom
+
+    raise ImportError("no suitable DOM implementation found")
+
+def _parse_feature_string(s):
+    features = []
+    parts = s.split()
+    i = 0
+    length = len(parts)
+    while i < length:
+        feature = parts[i]
+        if feature[0] in "0123456789":
+            raise ValueError("bad feature name: %r" % (feature,))
+        i = i + 1
+        version = None
+        if i < length:
+            v = parts[i]
+            if v[0] in "0123456789":
+                i = i + 1
+                version = v
+        features.append((feature, version))
+    return tuple(features)

third_party/python/Lib/xml/dom/expatbuilder.py

@@ -0,0 +1,965 @@
+"""Facility to use the Expat parser to load a minidom instance
+from a string or file.
+
+This avoids all the overhead of SAX and pulldom to gain performance.
+"""
+
+# Warning!
+#
+# This module is tightly bound to the implementation details of the
+# minidom DOM and can't be used with other DOM implementations.  This
+# is due, in part, to a lack of appropriate methods in the DOM (there is
+# no way to create Entity and Notation nodes via the DOM Level 2
+# interface), and for performance.  The latter is the cause of some fairly
+# cryptic code.
+#
+# Performance hacks:
+#
+#   -  .character_data_handler() has an extra case in which continuing
+#      data is appended to an existing Text node; this can be a
+#      speedup since pyexpat can break up character data into multiple
+#      callbacks even though we set the buffer_text attribute on the
+#      parser.  This also gives us the advantage that we don't need a
+#      separate normalization pass.
+#
+#   -  Determining that a node exists is done using an identity comparison
+#      with None rather than a truth test; this avoids searching for and
+#      calling any methods on the node object if it exists.  (A rather
+#      nice speedup is achieved this way as well!)
+
+from xml.dom import xmlbuilder, minidom, Node
+from xml.dom import EMPTY_NAMESPACE, EMPTY_PREFIX, XMLNS_NAMESPACE
+from xml.parsers import expat
+from xml.dom.minidom import _append_child, _set_attribute_node
+from xml.dom.NodeFilter import NodeFilter
+
+TEXT_NODE = Node.TEXT_NODE
+CDATA_SECTION_NODE = Node.CDATA_SECTION_NODE
+DOCUMENT_NODE = Node.DOCUMENT_NODE
+
+FILTER_ACCEPT = xmlbuilder.DOMBuilderFilter.FILTER_ACCEPT
+FILTER_REJECT = xmlbuilder.DOMBuilderFilter.FILTER_REJECT
+FILTER_SKIP = xmlbuilder.DOMBuilderFilter.FILTER_SKIP
+FILTER_INTERRUPT = xmlbuilder.DOMBuilderFilter.FILTER_INTERRUPT
+
+theDOMImplementation = minidom.getDOMImplementation()
+
+# Expat typename -> TypeInfo
+_typeinfo_map = {
+    "CDATA":    minidom.TypeInfo(None, "cdata"),
+    "ENUM":     minidom.TypeInfo(None, "enumeration"),
+    "ENTITY":   minidom.TypeInfo(None, "entity"),
+    "ENTITIES": minidom.TypeInfo(None, "entities"),
+    "ID":       minidom.TypeInfo(None, "id"),
+    "IDREF":    minidom.TypeInfo(None, "idref"),
+    "IDREFS":   minidom.TypeInfo(None, "idrefs"),
+    "NMTOKEN":  minidom.TypeInfo(None, "nmtoken"),
+    "NMTOKENS": minidom.TypeInfo(None, "nmtokens"),
+    }
+
+class ElementInfo(object):
+    __slots__ = '_attr_info', '_model', 'tagName'
+
+    def __init__(self, tagName, model=None):
+        self.tagName = tagName
+        self._attr_info = []
+        self._model = model
+
+    def __getstate__(self):
+        return self._attr_info, self._model, self.tagName
+
+    def __setstate__(self, state):
+        self._attr_info, self._model, self.tagName = state
+
+    def getAttributeType(self, aname):
+        for info in self._attr_info:
+            if info[1] == aname:
+                t = info[-2]
+                if t[0] == "(":
+                    return _typeinfo_map["ENUM"]
+                else:
+                    return _typeinfo_map[info[-2]]
+        return minidom._no_type
+
+    def getAttributeTypeNS(self, namespaceURI, localName):
+        return minidom._no_type
+
+    def isElementContent(self):
+        if self._model:
+            type = self._model[0]
+            return type not in (expat.model.XML_CTYPE_ANY,
+                                expat.model.XML_CTYPE_MIXED)
+        else:
+            return False
+
+    def isEmpty(self):
+        if self._model:
+            return self._model[0] == expat.model.XML_CTYPE_EMPTY
+        else:
+            return False
+
+    def isId(self, aname):
+        for info in self._attr_info:
+            if info[1] == aname:
+                return info[-2] == "ID"
+        return False
+
+    def isIdNS(self, euri, ename, auri, aname):
+        # not sure this is meaningful
+        return self.isId((auri, aname))
+
+def _intern(builder, s):
+    return builder._intern_setdefault(s, s)
+
+def _parse_ns_name(builder, name):
+    assert ' ' in name
+    parts = name.split(' ')
+    intern = builder._intern_setdefault
+    if len(parts) == 3:
+        uri, localname, prefix = parts
+        prefix = intern(prefix, prefix)
+        qname = "%s:%s" % (prefix, localname)
+        qname = intern(qname, qname)
+        localname = intern(localname, localname)
+    elif len(parts) == 2:
+        uri, localname = parts
+        prefix = EMPTY_PREFIX
+        qname = localname = intern(localname, localname)
+    else:
+        raise ValueError("Unsupported syntax: spaces in URIs not supported: %r" % name)
+    return intern(uri, uri), localname, prefix, qname
+
+
+class ExpatBuilder:
+    """Document builder that uses Expat to build a ParsedXML.DOM document
+    instance."""
+
+    def __init__(self, options=None):
+        if options is None:
+            options = xmlbuilder.Options()
+        self._options = options
+        if self._options.filter is not None:
+            self._filter = FilterVisibilityController(self._options.filter)
+        else:
+            self._filter = None
+            # This *really* doesn't do anything in this case, so
+            # override it with something fast & minimal.
+            self._finish_start_element = id
+        self._parser = None
+        self.reset()
+
+    def createParser(self):
+        """Create a new parser object."""
+        return expat.ParserCreate()
+
+    def getParser(self):
+        """Return the parser object, creating a new one if needed."""
+        if not self._parser:
+            self._parser = self.createParser()
+            self._intern_setdefault = self._parser.intern.setdefault
+            self._parser.buffer_text = True
+            self._parser.ordered_attributes = True
+            self._parser.specified_attributes = True
+            self.install(self._parser)
+        return self._parser
+
+    def reset(self):
+        """Free all data structures used during DOM construction."""
+        self.document = theDOMImplementation.createDocument(
+            EMPTY_NAMESPACE, None, None)
+        self.curNode = self.document
+        self._elem_info = self.document._elem_info
+        self._cdata = False
+
+    def install(self, parser):
+        """Install the callbacks needed to build the DOM into the parser."""
+        # This creates circular references!
+        parser.StartDoctypeDeclHandler = self.start_doctype_decl_handler
+        parser.StartElementHandler = self.first_element_handler
+        parser.EndElementHandler = self.end_element_handler
+        parser.ProcessingInstructionHandler = self.pi_handler
+        if self._options.entities:
+            parser.EntityDeclHandler = self.entity_decl_handler
+        parser.NotationDeclHandler = self.notation_decl_handler
+        if self._options.comments:
+            parser.CommentHandler = self.comment_handler
+        if self._options.cdata_sections:
+            parser.StartCdataSectionHandler = self.start_cdata_section_handler
+            parser.EndCdataSectionHandler = self.end_cdata_section_handler
+            parser.CharacterDataHandler = self.character_data_handler_cdata
+        else:
+            parser.CharacterDataHandler = self.character_data_handler
+        parser.ExternalEntityRefHandler = self.external_entity_ref_handler
+        parser.XmlDeclHandler = self.xml_decl_handler
+        parser.ElementDeclHandler = self.element_decl_handler
+        parser.AttlistDeclHandler = self.attlist_decl_handler
+
+    def parseFile(self, file):
+        """Parse a document from a file object, returning the document
+        node."""
+        parser = self.getParser()
+        first_buffer = True
+        try:
+            while 1:
+                buffer = file.read(16*1024)
+                if not buffer:
+                    break
+                parser.Parse(buffer, 0)
+                if first_buffer and self.document.documentElement:
+                    self._setup_subset(buffer)
+                first_buffer = False
+            parser.Parse("", True)
+        except ParseEscape:
+            pass
+        doc = self.document
+        self.reset()
+        self._parser = None
+        return doc
+
+    def parseString(self, string):
+        """Parse a document from a string, returning the document node."""
+        parser = self.getParser()
+        try:
+            parser.Parse(string, True)
+            self._setup_subset(string)
+        except ParseEscape:
+            pass
+        doc = self.document
+        self.reset()
+        self._parser = None
+        return doc
+
+    def _setup_subset(self, buffer):
+        """Load the internal subset if there might be one."""
+        if self.document.doctype:
+            extractor = InternalSubsetExtractor()
+            extractor.parseString(buffer)
+            subset = extractor.getSubset()
+            self.document.doctype.internalSubset = subset
+
+    def start_doctype_decl_handler(self, doctypeName, systemId, publicId,
+                                   has_internal_subset):
+        doctype = self.document.implementation.createDocumentType(
+            doctypeName, publicId, systemId)
+        doctype.ownerDocument = self.document
+        _append_child(self.document, doctype)
+        self.document.doctype = doctype
+        if self._filter and self._filter.acceptNode(doctype) == FILTER_REJECT:
+            self.document.doctype = None
+            del self.document.childNodes[-1]
+            doctype = None
+            self._parser.EntityDeclHandler = None
+            self._parser.NotationDeclHandler = None
+        if has_internal_subset:
+            if doctype is not None:
+                doctype.entities._seq = []
+                doctype.notations._seq = []
+            self._parser.CommentHandler = None
+            self._parser.ProcessingInstructionHandler = None
+            self._parser.EndDoctypeDeclHandler = self.end_doctype_decl_handler
+
+    def end_doctype_decl_handler(self):
+        if self._options.comments:
+            self._parser.CommentHandler = self.comment_handler
+        self._parser.ProcessingInstructionHandler = self.pi_handler
+        if not (self._elem_info or self._filter):
+            self._finish_end_element = id
+
+    def pi_handler(self, target, data):
+        node = self.document.createProcessingInstruction(target, data)
+        _append_child(self.curNode, node)
+        if self._filter and self._filter.acceptNode(node) == FILTER_REJECT:
+            self.curNode.removeChild(node)
+
+    def character_data_handler_cdata(self, data):
+        childNodes = self.curNode.childNodes
+        if self._cdata:
+            if (  self._cdata_continue
+                  and childNodes[-1].nodeType == CDATA_SECTION_NODE):
+                childNodes[-1].appendData(data)
+                return
+            node = self.document.createCDATASection(data)
+            self._cdata_continue = True
+        elif childNodes and childNodes[-1].nodeType == TEXT_NODE:
+            node = childNodes[-1]
+            value = node.data + data
+            node.data = value
+            return
+        else:
+            node = minidom.Text()
+            node.data = data
+            node.ownerDocument = self.document
+        _append_child(self.curNode, node)
+
+    def character_data_handler(self, data):
+        childNodes = self.curNode.childNodes
+        if childNodes and childNodes[-1].nodeType == TEXT_NODE:
+            node = childNodes[-1]
+            node.data = node.data + data
+            return
+        node = minidom.Text()
+        node.data = node.data + data
+        node.ownerDocument = self.document
+        _append_child(self.curNode, node)
+
+    def entity_decl_handler(self, entityName, is_parameter_entity, value,
+                            base, systemId, publicId, notationName):
+        if is_parameter_entity:
+            # we don't care about parameter entities for the DOM
+            return
+        if not self._options.entities:
+            return
+        node = self.document._create_entity(entityName, publicId,
+                                            systemId, notationName)
+        if value is not None:
+            # internal entity
+            # node *should* be readonly, but we'll cheat
+            child = self.document.createTextNode(value)
+            node.childNodes.append(child)
+        self.document.doctype.entities._seq.append(node)
+        if self._filter and self._filter.acceptNode(node) == FILTER_REJECT:
+            del self.document.doctype.entities._seq[-1]
+
+    def notation_decl_handler(self, notationName, base, systemId, publicId):
+        node = self.document._create_notation(notationName, publicId, systemId)
+        self.document.doctype.notations._seq.append(node)
+        if self._filter and self._filter.acceptNode(node) == FILTER_ACCEPT:
+            del self.document.doctype.notations._seq[-1]
+
+    def comment_handler(self, data):
+        node = self.document.createComment(data)
+        _append_child(self.curNode, node)
+        if self._filter and self._filter.acceptNode(node) == FILTER_REJECT:
+            self.curNode.removeChild(node)
+
+    def start_cdata_section_handler(self):
+        self._cdata = True
+        self._cdata_continue = False
+
+    def end_cdata_section_handler(self):
+        self._cdata = False
+        self._cdata_continue = False
+
+    def external_entity_ref_handler(self, context, base, systemId, publicId):
+        return 1
+
+    def first_element_handler(self, name, attributes):
+        if self._filter is None and not self._elem_info:
+            self._finish_end_element = id
+        self.getParser().StartElementHandler = self.start_element_handler
+        self.start_element_handler(name, attributes)
+
+    def start_element_handler(self, name, attributes):
+        node = self.document.createElement(name)
+        _append_child(self.curNode, node)
+        self.curNode = node
+
+        if attributes:
+            for i in range(0, len(attributes), 2):
+                a = minidom.Attr(attributes[i], EMPTY_NAMESPACE,
+                                 None, EMPTY_PREFIX)
+                value = attributes[i+1]
+                a.value = value
+                a.ownerDocument = self.document
+                _set_attribute_node(node, a)
+
+        if node is not self.document.documentElement:
+            self._finish_start_element(node)
+
+    def _finish_start_element(self, node):
+        if self._filter:
+            # To be general, we'd have to call isSameNode(), but this
+            # is sufficient for minidom:
+            if node is self.document.documentElement:
+                return
+            filt = self._filter.startContainer(node)
+            if filt == FILTER_REJECT:
+                # ignore this node & all descendents
+                Rejecter(self)
+            elif filt == FILTER_SKIP:
+                # ignore this node, but make it's children become
+                # children of the parent node
+                Skipper(self)
+            else:
+                return
+            self.curNode = node.parentNode
+            node.parentNode.removeChild(node)
+            node.unlink()
+
+    # If this ever changes, Namespaces.end_element_handler() needs to
+    # be changed to match.
+    #
+    def end_element_handler(self, name):
+        curNode = self.curNode
+        self.curNode = curNode.parentNode
+        self._finish_end_element(curNode)
+
+    def _finish_end_element(self, curNode):
+        info = self._elem_info.get(curNode.tagName)
+        if info:
+            self._handle_white_text_nodes(curNode, info)
+        if self._filter:
+            if curNode is self.document.documentElement:
+                return
+            if self._filter.acceptNode(curNode) == FILTER_REJECT:
+                self.curNode.removeChild(curNode)
+                curNode.unlink()
+
+    def _handle_white_text_nodes(self, node, info):
+        if (self._options.whitespace_in_element_content
+            or not info.isElementContent()):
+            return
+
+        # We have element type information and should remove ignorable
+        # whitespace; identify for text nodes which contain only
+        # whitespace.
+        L = []
+        for child in node.childNodes:
+            if child.nodeType == TEXT_NODE and not child.data.strip():
+                L.append(child)
+
+        # Remove ignorable whitespace from the tree.
+        for child in L:
+            node.removeChild(child)
+
+    def element_decl_handler(self, name, model):
+        info = self._elem_info.get(name)
+        if info is None:
+            self._elem_info[name] = ElementInfo(name, model)
+        else:
+            assert info._model is None
+            info._model = model
+
+    def attlist_decl_handler(self, elem, name, type, default, required):
+        info = self._elem_info.get(elem)
+        if info is None:
+            info = ElementInfo(elem)
+            self._elem_info[elem] = info
+        info._attr_info.append(
+            [None, name, None, None, default, 0, type, required])
+
+    def xml_decl_handler(self, version, encoding, standalone):
+        self.document.version = version
+        self.document.encoding = encoding
+        # This is still a little ugly, thanks to the pyexpat API. ;-(
+        if standalone >= 0:
+            if standalone:
+                self.document.standalone = True
+            else:
+                self.document.standalone = False
+
+
+# Don't include FILTER_INTERRUPT, since that's checked separately
+# where allowed.
+_ALLOWED_FILTER_RETURNS = (FILTER_ACCEPT, FILTER_REJECT, FILTER_SKIP)
+
+class FilterVisibilityController(object):
+    """Wrapper around a DOMBuilderFilter which implements the checks
+    to make the whatToShow filter attribute work."""
+
+    __slots__ = 'filter',
+
+    def __init__(self, filter):
+        self.filter = filter
+
+    def startContainer(self, node):
+        mask = self._nodetype_mask[node.nodeType]
+        if self.filter.whatToShow & mask:
+            val = self.filter.startContainer(node)
+            if val == FILTER_INTERRUPT:
+                raise ParseEscape
+            if val not in _ALLOWED_FILTER_RETURNS:
+                raise ValueError(
+                      "startContainer() returned illegal value: " + repr(val))
+            return val
+        else:
+            return FILTER_ACCEPT
+
+    def acceptNode(self, node):
+        mask = self._nodetype_mask[node.nodeType]
+        if self.filter.whatToShow & mask:
+            val = self.filter.acceptNode(node)
+            if val == FILTER_INTERRUPT:
+                raise ParseEscape
+            if val == FILTER_SKIP:
+                # move all child nodes to the parent, and remove this node
+                parent = node.parentNode
+                for child in node.childNodes[:]:
+                    parent.appendChild(child)
+                # node is handled by the caller
+                return FILTER_REJECT
+            if val not in _ALLOWED_FILTER_RETURNS:
+                raise ValueError(
+                      "acceptNode() returned illegal value: " + repr(val))
+            return val
+        else:
+            return FILTER_ACCEPT
+
+    _nodetype_mask = {
+        Node.ELEMENT_NODE:                NodeFilter.SHOW_ELEMENT,
+        Node.ATTRIBUTE_NODE:              NodeFilter.SHOW_ATTRIBUTE,
+        Node.TEXT_NODE:                   NodeFilter.SHOW_TEXT,
+        Node.CDATA_SECTION_NODE:          NodeFilter.SHOW_CDATA_SECTION,
+        Node.ENTITY_REFERENCE_NODE:       NodeFilter.SHOW_ENTITY_REFERENCE,
+        Node.ENTITY_NODE:                 NodeFilter.SHOW_ENTITY,
+        Node.PROCESSING_INSTRUCTION_NODE: NodeFilter.SHOW_PROCESSING_INSTRUCTION,
+        Node.COMMENT_NODE:                NodeFilter.SHOW_COMMENT,
+        Node.DOCUMENT_NODE:               NodeFilter.SHOW_DOCUMENT,
+        Node.DOCUMENT_TYPE_NODE:          NodeFilter.SHOW_DOCUMENT_TYPE,
+        Node.DOCUMENT_FRAGMENT_NODE:      NodeFilter.SHOW_DOCUMENT_FRAGMENT,
+        Node.NOTATION_NODE:               NodeFilter.SHOW_NOTATION,
+        }
+
+
+class FilterCrutch(object):
+    __slots__ = '_builder', '_level', '_old_start', '_old_end'
+
+    def __init__(self, builder):
+        self._level = 0
+        self._builder = builder
+        parser = builder._parser
+        self._old_start = parser.StartElementHandler
+        self._old_end = parser.EndElementHandler
+        parser.StartElementHandler = self.start_element_handler
+        parser.EndElementHandler = self.end_element_handler
+
+class Rejecter(FilterCrutch):
+    __slots__ = ()
+
+    def __init__(self, builder):
+        FilterCrutch.__init__(self, builder)
+        parser = builder._parser
+        for name in ("ProcessingInstructionHandler",
+                     "CommentHandler",
+                     "CharacterDataHandler",
+                     "StartCdataSectionHandler",
+                     "EndCdataSectionHandler",
+                     "ExternalEntityRefHandler",
+                     ):
+            setattr(parser, name, None)
+
+    def start_element_handler(self, *args):
+        self._level = self._level + 1
+
+    def end_element_handler(self, *args):
+        if self._level == 0:
+            # restore the old handlers
+            parser = self._builder._parser
+            self._builder.install(parser)
+            parser.StartElementHandler = self._old_start
+            parser.EndElementHandler = self._old_end
+        else:
+            self._level = self._level - 1
+
+class Skipper(FilterCrutch):
+    __slots__ = ()
+
+    def start_element_handler(self, *args):
+        node = self._builder.curNode
+        self._old_start(*args)
+        if self._builder.curNode is not node:
+            self._level = self._level + 1
+
+    def end_element_handler(self, *args):
+        if self._level == 0:
+            # We're popping back out of the node we're skipping, so we
+            # shouldn't need to do anything but reset the handlers.
+            self._builder._parser.StartElementHandler = self._old_start
+            self._builder._parser.EndElementHandler = self._old_end
+            self._builder = None
+        else:
+            self._level = self._level - 1
+            self._old_end(*args)
+
+
+# framework document used by the fragment builder.
+# Takes a string for the doctype, subset string, and namespace attrs string.
+
+_FRAGMENT_BUILDER_INTERNAL_SYSTEM_ID = \
+    "http://xml.python.org/entities/fragment-builder/internal"
+
+_FRAGMENT_BUILDER_TEMPLATE = (
+    '''\
+<!DOCTYPE wrapper
+  %%s [
+  <!ENTITY fragment-builder-internal
+    SYSTEM "%s">
+%%s
+]>
+<wrapper %%s
+>&fragment-builder-internal;</wrapper>'''
+    % _FRAGMENT_BUILDER_INTERNAL_SYSTEM_ID)
+
+
+class FragmentBuilder(ExpatBuilder):
+    """Builder which constructs document fragments given XML source
+    text and a context node.
+
+    The context node is expected to provide information about the
+    namespace declarations which are in scope at the start of the
+    fragment.
+    """
+
+    def __init__(self, context, options=None):
+        if context.nodeType == DOCUMENT_NODE:
+            self.originalDocument = context
+            self.context = context
+        else:
+            self.originalDocument = context.ownerDocument
+            self.context = context
+        ExpatBuilder.__init__(self, options)
+
+    def reset(self):
+        ExpatBuilder.reset(self)
+        self.fragment = None
+
+    def parseFile(self, file):
+        """Parse a document fragment from a file object, returning the
+        fragment node."""
+        return self.parseString(file.read())
+
+    def parseString(self, string):
+        """Parse a document fragment from a string, returning the
+        fragment node."""
+        self._source = string
+        parser = self.getParser()
+        doctype = self.originalDocument.doctype
+        ident = ""
+        if doctype:
+            subset = doctype.internalSubset or self._getDeclarations()
+            if doctype.publicId:
+                ident = ('PUBLIC "%s" "%s"'
+                         % (doctype.publicId, doctype.systemId))
+            elif doctype.systemId:
+                ident = 'SYSTEM "%s"' % doctype.systemId
+        else:
+            subset = ""
+        nsattrs = self._getNSattrs() # get ns decls from node's ancestors
+        document = _FRAGMENT_BUILDER_TEMPLATE % (ident, subset, nsattrs)
+        try:
+            parser.Parse(document, 1)
+        except:
+            self.reset()
+            raise
+        fragment = self.fragment
+        self.reset()
+##         self._parser = None
+        return fragment
+
+    def _getDeclarations(self):
+        """Re-create the internal subset from the DocumentType node.
+
+        This is only needed if we don't already have the
+        internalSubset as a string.
+        """
+        doctype = self.context.ownerDocument.doctype
+        s = ""
+        if doctype:
+            for i in range(doctype.notations.length):
+                notation = doctype.notations.item(i)
+                if s:
+                    s = s + "\n  "
+                s = "%s<!NOTATION %s" % (s, notation.nodeName)
+                if notation.publicId:
+                    s = '%s PUBLIC "%s"\n             "%s">' \
+                        % (s, notation.publicId, notation.systemId)
+                else:
+                    s = '%s SYSTEM "%s">' % (s, notation.systemId)
+            for i in range(doctype.entities.length):
+                entity = doctype.entities.item(i)
+                if s:
+                    s = s + "\n  "
+                s = "%s<!ENTITY %s" % (s, entity.nodeName)
+                if entity.publicId:
+                    s = '%s PUBLIC "%s"\n             "%s"' \
+                        % (s, entity.publicId, entity.systemId)
+                elif entity.systemId:
+                    s = '%s SYSTEM "%s"' % (s, entity.systemId)
+                else:
+                    s = '%s "%s"' % (s, entity.firstChild.data)
+                if entity.notationName:
+                    s = "%s NOTATION %s" % (s, entity.notationName)
+                s = s + ">"
+        return s
+
+    def _getNSattrs(self):
+        return ""
+
+    def external_entity_ref_handler(self, context, base, systemId, publicId):
+        if systemId == _FRAGMENT_BUILDER_INTERNAL_SYSTEM_ID:
+            # this entref is the one that we made to put the subtree
+            # in; all of our given input is parsed in here.
+            old_document = self.document
+            old_cur_node = self.curNode
+            parser = self._parser.ExternalEntityParserCreate(context)
+            # put the real document back, parse into the fragment to return
+            self.document = self.originalDocument
+            self.fragment = self.document.createDocumentFragment()
+            self.curNode = self.fragment
+            try:
+                parser.Parse(self._source, 1)
+            finally:
+                self.curNode = old_cur_node
+                self.document = old_document
+                self._source = None
+            return -1
+        else:
+            return ExpatBuilder.external_entity_ref_handler(
+                self, context, base, systemId, publicId)
+
+
+class Namespaces:
+    """Mix-in class for builders; adds support for namespaces."""
+
+    def _initNamespaces(self):
+        # list of (prefix, uri) ns declarations.  Namespace attrs are
+        # constructed from this and added to the element's attrs.
+        self._ns_ordered_prefixes = []
+
+    def createParser(self):
+        """Create a new namespace-handling parser."""
+        parser = expat.ParserCreate(namespace_separator=" ")
+        parser.namespace_prefixes = True
+        return parser
+
+    def install(self, parser):
+        """Insert the namespace-handlers onto the parser."""
+        ExpatBuilder.install(self, parser)
+        if self._options.namespace_declarations:
+            parser.StartNamespaceDeclHandler = (
+                self.start_namespace_decl_handler)
+
+    def start_namespace_decl_handler(self, prefix, uri):
+        """Push this namespace declaration on our storage."""
+        self._ns_ordered_prefixes.append((prefix, uri))
+
+    def start_element_handler(self, name, attributes):
+        if ' ' in name:
+            uri, localname, prefix, qname = _parse_ns_name(self, name)
+        else:
+            uri = EMPTY_NAMESPACE
+            qname = name
+            localname = None
+            prefix = EMPTY_PREFIX
+        node = minidom.Element(qname, uri, prefix, localname)
+        node.ownerDocument = self.document
+        _append_child(self.curNode, node)
+        self.curNode = node
+
+        if self._ns_ordered_prefixes:
+            for prefix, uri in self._ns_ordered_prefixes:
+                if prefix:
+                    a = minidom.Attr(_intern(self, 'xmlns:' + prefix),
+                                     XMLNS_NAMESPACE, prefix, "xmlns")
+                else:
+                    a = minidom.Attr("xmlns", XMLNS_NAMESPACE,
+                                     "xmlns", EMPTY_PREFIX)
+                a.value = uri
+                a.ownerDocument = self.document
+                _set_attribute_node(node, a)
+            del self._ns_ordered_prefixes[:]
+
+        if attributes:
+            node._ensure_attributes()
+            _attrs = node._attrs
+            _attrsNS = node._attrsNS
+            for i in range(0, len(attributes), 2):
+                aname = attributes[i]
+                value = attributes[i+1]
+                if ' ' in aname:
+                    uri, localname, prefix, qname = _parse_ns_name(self, aname)
+                    a = minidom.Attr(qname, uri, localname, prefix)
+                    _attrs[qname] = a
+                    _attrsNS[(uri, localname)] = a
+                else:
+                    a = minidom.Attr(aname, EMPTY_NAMESPACE,
+                                     aname, EMPTY_PREFIX)
+                    _attrs[aname] = a
+                    _attrsNS[(EMPTY_NAMESPACE, aname)] = a
+                a.ownerDocument = self.document
+                a.value = value
+                a.ownerElement = node
+
+    if __debug__:
+        # This only adds some asserts to the original
+        # end_element_handler(), so we only define this when -O is not
+        # used.  If changing one, be sure to check the other to see if
+        # it needs to be changed as well.
+        #
+        def end_element_handler(self, name):
+            curNode = self.curNode
+            if ' ' in name:
+                uri, localname, prefix, qname = _parse_ns_name(self, name)
+                assert (curNode.namespaceURI == uri
+                        and curNode.localName == localname
+                        and curNode.prefix == prefix), \
+                        "element stack messed up! (namespace)"
+            else:
+                assert curNode.nodeName == name, \
+                       "element stack messed up - bad nodeName"
+                assert curNode.namespaceURI == EMPTY_NAMESPACE, \
+                       "element stack messed up - bad namespaceURI"
+            self.curNode = curNode.parentNode
+            self._finish_end_element(curNode)
+
+
+class ExpatBuilderNS(Namespaces, ExpatBuilder):
+    """Document builder that supports namespaces."""
+
+    def reset(self):
+        ExpatBuilder.reset(self)
+        self._initNamespaces()
+
+
+class FragmentBuilderNS(Namespaces, FragmentBuilder):
+    """Fragment builder that supports namespaces."""
+
+    def reset(self):
+        FragmentBuilder.reset(self)
+        self._initNamespaces()
+
+    def _getNSattrs(self):
+        """Return string of namespace attributes from this element and
+        ancestors."""
+        # XXX This needs to be re-written to walk the ancestors of the
+        # context to build up the namespace information from
+        # declarations, elements, and attributes found in context.
+        # Otherwise we have to store a bunch more data on the DOM
+        # (though that *might* be more reliable -- not clear).
+        attrs = ""
+        context = self.context
+        L = []
+        while context:
+            if hasattr(context, '_ns_prefix_uri'):
+                for prefix, uri in context._ns_prefix_uri.items():
+                    # add every new NS decl from context to L and attrs string
+                    if prefix in L:
+                        continue
+                    L.append(prefix)
+                    if prefix:
+                        declname = "xmlns:" + prefix
+                    else:
+                        declname = "xmlns"
+                    if attrs:
+                        attrs = "%s\n    %s='%s'" % (attrs, declname, uri)
+                    else:
+                        attrs = " %s='%s'" % (declname, uri)
+            context = context.parentNode
+        return attrs
+
+
+class ParseEscape(Exception):
+    """Exception raised to short-circuit parsing in InternalSubsetExtractor."""
+    pass
+
+class InternalSubsetExtractor(ExpatBuilder):
+    """XML processor which can rip out the internal document type subset."""
+
+    subset = None
+
+    def getSubset(self):
+        """Return the internal subset as a string."""
+        return self.subset
+
+    def parseFile(self, file):
+        try:
+            ExpatBuilder.parseFile(self, file)
+        except ParseEscape:
+            pass
+
+    def parseString(self, string):
+        try:
+            ExpatBuilder.parseString(self, string)
+        except ParseEscape:
+            pass
+
+    def install(self, parser):
+        parser.StartDoctypeDeclHandler = self.start_doctype_decl_handler
+        parser.StartElementHandler = self.start_element_handler
+
+    def start_doctype_decl_handler(self, name, publicId, systemId,
+                                   has_internal_subset):
+        if has_internal_subset:
+            parser = self.getParser()
+            self.subset = []
+            parser.DefaultHandler = self.subset.append
+            parser.EndDoctypeDeclHandler = self.end_doctype_decl_handler
+        else:
+            raise ParseEscape()
+
+    def end_doctype_decl_handler(self):
+        s = ''.join(self.subset).replace('\r\n', '\n').replace('\r', '\n')
+        self.subset = s
+        raise ParseEscape()
+
+    def start_element_handler(self, name, attrs):
+        raise ParseEscape()
+
+
+def parse(file, namespaces=True):
+    """Parse a document, returning the resulting Document node.
+
+    'file' may be either a file name or an open file object.
+    """
+    if namespaces:
+        builder = ExpatBuilderNS()
+    else:
+        builder = ExpatBuilder()
+
+    if isinstance(file, str):
+        with open(file, 'rb') as fp:
+            result = builder.parseFile(fp)
+    else:
+        result = builder.parseFile(file)
+    return result
+
+
+def parseString(string, namespaces=True):
+    """Parse a document from a string, returning the resulting
+    Document node.
+    """
+    if namespaces:
+        builder = ExpatBuilderNS()
+    else:
+        builder = ExpatBuilder()
+    return builder.parseString(string)
+
+
+def parseFragment(file, context, namespaces=True):
+    """Parse a fragment of a document, given the context from which it
+    was originally extracted.  context should be the parent of the
+    node(s) which are in the fragment.
+
+    'file' may be either a file name or an open file object.
+    """
+    if namespaces:
+        builder = FragmentBuilderNS(context)
+    else:
+        builder = FragmentBuilder(context)
+
+    if isinstance(file, str):
+        with open(file, 'rb') as fp:
+            result = builder.parseFile(fp)
+    else:
+        result = builder.parseFile(file)
+    return result
+
+
+def parseFragmentString(string, context, namespaces=True):
+    """Parse a fragment of a document from a string, given the context
+    from which it was originally extracted.  context should be the
+    parent of the node(s) which are in the fragment.
+    """
+    if namespaces:
+        builder = FragmentBuilderNS(context)
+    else:
+        builder = FragmentBuilder(context)
+    return builder.parseString(string)
+
+
+def makeBuilder(options):
+    """Create a builder based on an Options object."""
+    if options.namespaces:
+        return ExpatBuilderNS(options)
+    else:
+        return ExpatBuilder(options)

third_party/python/Lib/xml/dom/minicompat.py

@@ -0,0 +1,109 @@
+"""Python version compatibility support for minidom.
+
+This module contains internal implementation details and
+should not be imported; use xml.dom.minidom instead.
+"""
+
+# This module should only be imported using "import *".
+#
+# The following names are defined:
+#
+#   NodeList      -- lightest possible NodeList implementation
+#
+#   EmptyNodeList -- lightest possible NodeList that is guaranteed to
+#                    remain empty (immutable)
+#
+#   StringTypes   -- tuple of defined string types
+#
+#   defproperty   -- function used in conjunction with GetattrMagic;
+#                    using these together is needed to make them work
+#                    as efficiently as possible in both Python 2.2+
+#                    and older versions.  For example:
+#
+#                        class MyClass(GetattrMagic):
+#                            def _get_myattr(self):
+#                                return something
+#
+#                        defproperty(MyClass, "myattr",
+#                                    "return some value")
+#
+#                    For Python 2.2 and newer, this will construct a
+#                    property object on the class, which avoids
+#                    needing to override __getattr__().  It will only
+#                    work for read-only attributes.
+#
+#                    For older versions of Python, inheriting from
+#                    GetattrMagic will use the traditional
+#                    __getattr__() hackery to achieve the same effect,
+#                    but less efficiently.
+#
+#                    defproperty() should be used for each version of
+#                    the relevant _get_<property>() function.
+
+__all__ = ["NodeList", "EmptyNodeList", "StringTypes", "defproperty"]
+
+import xml.dom
+
+StringTypes = (str,)
+
+
+class NodeList(list):
+    __slots__ = ()
+
+    def item(self, index):
+        if 0 <= index < len(self):
+            return self[index]
+
+    def _get_length(self):
+        return len(self)
+
+    def _set_length(self, value):
+        raise xml.dom.NoModificationAllowedErr(
+            "attempt to modify read-only attribute 'length'")
+
+    length = property(_get_length, _set_length,
+                      doc="The number of nodes in the NodeList.")
+
+    # For backward compatibility
+    def __setstate__(self, state):
+        if state is None:
+            state = []
+        self[:] = state
+
+
+class EmptyNodeList(tuple):
+    __slots__ = ()
+
+    def __add__(self, other):
+        NL = NodeList()
+        NL.extend(other)
+        return NL
+
+    def __radd__(self, other):
+        NL = NodeList()
+        NL.extend(other)
+        return NL
+
+    def item(self, index):
+        return None
+
+    def _get_length(self):
+        return 0
+
+    def _set_length(self, value):
+        raise xml.dom.NoModificationAllowedErr(
+            "attempt to modify read-only attribute 'length'")
+
+    length = property(_get_length, _set_length,
+                      doc="The number of nodes in the NodeList.")
+
+
+def defproperty(klass, name, doc):
+    get = getattr(klass, ("_get_" + name))
+    def set(self, value, name=name):
+        raise xml.dom.NoModificationAllowedErr(
+            "attempt to modify read-only attribute " + repr(name))
+    assert not hasattr(klass, "_set_" + name), \
+           "expected not to find _set_" + name
+    prop = property(get, set, doc=doc)
+    setattr(klass, name, prop)

third_party/python/Lib/xml/dom/pulldom.py

@@ -0,0 +1,342 @@
+import xml.sax
+import xml.sax.handler
+
+START_ELEMENT = "START_ELEMENT"
+END_ELEMENT = "END_ELEMENT"
+COMMENT = "COMMENT"
+START_DOCUMENT = "START_DOCUMENT"
+END_DOCUMENT = "END_DOCUMENT"
+PROCESSING_INSTRUCTION = "PROCESSING_INSTRUCTION"
+IGNORABLE_WHITESPACE = "IGNORABLE_WHITESPACE"
+CHARACTERS = "CHARACTERS"
+
+class PullDOM(xml.sax.ContentHandler):
+    _locator = None
+    document = None
+
+    def __init__(self, documentFactory=None):
+        from xml.dom import XML_NAMESPACE
+        self.documentFactory = documentFactory
+        self.firstEvent = [None, None]
+        self.lastEvent = self.firstEvent
+        self.elementStack = []
+        self.push = self.elementStack.append
+        try:
+            self.pop = self.elementStack.pop
+        except AttributeError:
+            # use class' pop instead
+            pass
+        self._ns_contexts = [{XML_NAMESPACE:'xml'}] # contains uri -> prefix dicts
+        self._current_context = self._ns_contexts[-1]
+        self.pending_events = []
+
+    def pop(self):
+        result = self.elementStack[-1]
+        del self.elementStack[-1]
+        return result
+
+    def setDocumentLocator(self, locator):
+        self._locator = locator
+
+    def startPrefixMapping(self, prefix, uri):
+        if not hasattr(self, '_xmlns_attrs'):
+            self._xmlns_attrs = []
+        self._xmlns_attrs.append((prefix or 'xmlns', uri))
+        self._ns_contexts.append(self._current_context.copy())
+        self._current_context[uri] = prefix or None
+
+    def endPrefixMapping(self, prefix):
+        self._current_context = self._ns_contexts.pop()
+
+    def startElementNS(self, name, tagName , attrs):
+        # Retrieve xml namespace declaration attributes.
+        xmlns_uri = 'http://www.w3.org/2000/xmlns/'
+        xmlns_attrs = getattr(self, '_xmlns_attrs', None)
+        if xmlns_attrs is not None:
+            for aname, value in xmlns_attrs:
+                attrs._attrs[(xmlns_uri, aname)] = value
+            self._xmlns_attrs = []
+        uri, localname = name
+        if uri:
+            # When using namespaces, the reader may or may not
+            # provide us with the original name. If not, create
+            # *a* valid tagName from the current context.
+            if tagName is None:
+                prefix = self._current_context[uri]
+                if prefix:
+                    tagName = prefix + ":" + localname
+                else:
+                    tagName = localname
+            if self.document:
+                node = self.document.createElementNS(uri, tagName)
+            else:
+                node = self.buildDocument(uri, tagName)
+        else:
+            # When the tagname is not prefixed, it just appears as
+            # localname
+            if self.document:
+                node = self.document.createElement(localname)
+            else:
+                node = self.buildDocument(None, localname)
+
+        for aname,value in attrs.items():
+            a_uri, a_localname = aname
+            if a_uri == xmlns_uri:
+                if a_localname == 'xmlns':
+                    qname = a_localname
+                else:
+                    qname = 'xmlns:' + a_localname
+                attr = self.document.createAttributeNS(a_uri, qname)
+                node.setAttributeNodeNS(attr)
+            elif a_uri:
+                prefix = self._current_context[a_uri]
+                if prefix:
+                    qname = prefix + ":" + a_localname
+                else:
+                    qname = a_localname
+                attr = self.document.createAttributeNS(a_uri, qname)
+                node.setAttributeNodeNS(attr)
+            else:
+                attr = self.document.createAttribute(a_localname)
+                node.setAttributeNode(attr)
+            attr.value = value
+
+        self.lastEvent[1] = [(START_ELEMENT, node), None]
+        self.lastEvent = self.lastEvent[1]
+        self.push(node)
+
+    def endElementNS(self, name, tagName):
+        self.lastEvent[1] = [(END_ELEMENT, self.pop()), None]
+        self.lastEvent = self.lastEvent[1]
+
+    def startElement(self, name, attrs):
+        if self.document:
+            node = self.document.createElement(name)
+        else:
+            node = self.buildDocument(None, name)
+
+        for aname,value in attrs.items():
+            attr = self.document.createAttribute(aname)
+            attr.value = value
+            node.setAttributeNode(attr)
+
+        self.lastEvent[1] = [(START_ELEMENT, node), None]
+        self.lastEvent = self.lastEvent[1]
+        self.push(node)
+
+    def endElement(self, name):
+        self.lastEvent[1] = [(END_ELEMENT, self.pop()), None]
+        self.lastEvent = self.lastEvent[1]
+
+    def comment(self, s):
+        if self.document:
+            node = self.document.createComment(s)
+            self.lastEvent[1] = [(COMMENT, node), None]
+            self.lastEvent = self.lastEvent[1]
+        else:
+            event = [(COMMENT, s), None]
+            self.pending_events.append(event)
+
+    def processingInstruction(self, target, data):
+        if self.document:
+            node = self.document.createProcessingInstruction(target, data)
+            self.lastEvent[1] = [(PROCESSING_INSTRUCTION, node), None]
+            self.lastEvent = self.lastEvent[1]
+        else:
+            event = [(PROCESSING_INSTRUCTION, target, data), None]
+            self.pending_events.append(event)
+
+    def ignorableWhitespace(self, chars):
+        node = self.document.createTextNode(chars)
+        self.lastEvent[1] = [(IGNORABLE_WHITESPACE, node), None]
+        self.lastEvent = self.lastEvent[1]
+
+    def characters(self, chars):
+        node = self.document.createTextNode(chars)
+        self.lastEvent[1] = [(CHARACTERS, node), None]
+        self.lastEvent = self.lastEvent[1]
+
+    def startDocument(self):
+        if self.documentFactory is None:
+            import xml.dom.minidom
+            self.documentFactory = xml.dom.minidom.Document.implementation
+
+    def buildDocument(self, uri, tagname):
+        # Can't do that in startDocument, since we need the tagname
+        # XXX: obtain DocumentType
+        node = self.documentFactory.createDocument(uri, tagname, None)
+        self.document = node
+        self.lastEvent[1] = [(START_DOCUMENT, node), None]
+        self.lastEvent = self.lastEvent[1]
+        self.push(node)
+        # Put everything we have seen so far into the document
+        for e in self.pending_events:
+            if e[0][0] == PROCESSING_INSTRUCTION:
+                _,target,data = e[0]
+                n = self.document.createProcessingInstruction(target, data)
+                e[0] = (PROCESSING_INSTRUCTION, n)
+            elif e[0][0] == COMMENT:
+                n = self.document.createComment(e[0][1])
+                e[0] = (COMMENT, n)
+            else:
+                raise AssertionError("Unknown pending event ",e[0][0])
+            self.lastEvent[1] = e
+            self.lastEvent = e
+        self.pending_events = None
+        return node.firstChild
+
+    def endDocument(self):
+        self.lastEvent[1] = [(END_DOCUMENT, self.document), None]
+        self.pop()
+
+    def clear(self):
+        "clear(): Explicitly release parsing structures"
+        self.document = None
+
+class ErrorHandler:
+    def warning(self, exception):
+        print(exception)
+    def error(self, exception):
+        raise exception
+    def fatalError(self, exception):
+        raise exception
+
+class DOMEventStream:
+    def __init__(self, stream, parser, bufsize):
+        self.stream = stream
+        self.parser = parser
+        self.bufsize = bufsize
+        if not hasattr(self.parser, 'feed'):
+            self.getEvent = self._slurp
+        self.reset()
+
+    def reset(self):
+        self.pulldom = PullDOM()
+        # This content handler relies on namespace support
+        self.parser.setFeature(xml.sax.handler.feature_namespaces, 1)
+        self.parser.setContentHandler(self.pulldom)
+
+    def __getitem__(self, pos):
+        rc = self.getEvent()
+        if rc:
+            return rc
+        raise IndexError
+
+    def __next__(self):
+        rc = self.getEvent()
+        if rc:
+            return rc
+        raise StopIteration
+
+    def __iter__(self):
+        return self
+
+    def expandNode(self, node):
+        event = self.getEvent()
+        parents = [node]
+        while event:
+            token, cur_node = event
+            if cur_node is node:
+                return
+            if token != END_ELEMENT:
+                parents[-1].appendChild(cur_node)
+            if token == START_ELEMENT:
+                parents.append(cur_node)
+            elif token == END_ELEMENT:
+                del parents[-1]
+            event = self.getEvent()
+
+    def getEvent(self):
+        # use IncrementalParser interface, so we get the desired
+        # pull effect
+        if not self.pulldom.firstEvent[1]:
+            self.pulldom.lastEvent = self.pulldom.firstEvent
+        while not self.pulldom.firstEvent[1]:
+            buf = self.stream.read(self.bufsize)
+            if not buf:
+                self.parser.close()
+                return None
+            self.parser.feed(buf)
+        rc = self.pulldom.firstEvent[1][0]
+        self.pulldom.firstEvent[1] = self.pulldom.firstEvent[1][1]
+        return rc
+
+    def _slurp(self):
+        """ Fallback replacement for getEvent() using the
+            standard SAX2 interface, which means we slurp the
+            SAX events into memory (no performance gain, but
+            we are compatible to all SAX parsers).
+        """
+        self.parser.parse(self.stream)
+        self.getEvent = self._emit
+        return self._emit()
+
+    def _emit(self):
+        """ Fallback replacement for getEvent() that emits
+            the events that _slurp() read previously.
+        """
+        rc = self.pulldom.firstEvent[1][0]
+        self.pulldom.firstEvent[1] = self.pulldom.firstEvent[1][1]
+        return rc
+
+    def clear(self):
+        """clear(): Explicitly release parsing objects"""
+        self.pulldom.clear()
+        del self.pulldom
+        self.parser = None
+        self.stream = None
+
+class SAX2DOM(PullDOM):
+
+    def startElementNS(self, name, tagName , attrs):
+        PullDOM.startElementNS(self, name, tagName, attrs)
+        curNode = self.elementStack[-1]
+        parentNode = self.elementStack[-2]
+        parentNode.appendChild(curNode)
+
+    def startElement(self, name, attrs):
+        PullDOM.startElement(self, name, attrs)
+        curNode = self.elementStack[-1]
+        parentNode = self.elementStack[-2]
+        parentNode.appendChild(curNode)
+
+    def processingInstruction(self, target, data):
+        PullDOM.processingInstruction(self, target, data)
+        node = self.lastEvent[0][1]
+        parentNode = self.elementStack[-1]
+        parentNode.appendChild(node)
+
+    def ignorableWhitespace(self, chars):
+        PullDOM.ignorableWhitespace(self, chars)
+        node = self.lastEvent[0][1]
+        parentNode = self.elementStack[-1]
+        parentNode.appendChild(node)
+
+    def characters(self, chars):
+        PullDOM.characters(self, chars)
+        node = self.lastEvent[0][1]
+        parentNode = self.elementStack[-1]
+        parentNode.appendChild(node)
+
+
+default_bufsize = (2 ** 14) - 20
+
+def parse(stream_or_string, parser=None, bufsize=None):
+    if bufsize is None:
+        bufsize = default_bufsize
+    if isinstance(stream_or_string, str):
+        stream = open(stream_or_string, 'rb')
+    else:
+        stream = stream_or_string
+    if not parser:
+        parser = xml.sax.make_parser()
+    return DOMEventStream(stream, parser, bufsize)
+
+def parseString(string, parser=None):
+    from io import StringIO
+
+    bufsize = len(string)
+    buf = StringIO(string)
+    if not parser:
+        parser = xml.sax.make_parser()
+    return DOMEventStream(buf, parser, bufsize)

third_party/python/Lib/xml/dom/xmlbuilder.py

@@ -0,0 +1,410 @@
+"""Implementation of the DOM Level 3 'LS-Load' feature."""
+
+import copy
+import warnings
+import xml.dom
+
+from xml.dom.NodeFilter import NodeFilter
+
+
+__all__ = ["DOMBuilder", "DOMEntityResolver", "DOMInputSource"]
+
+
+class Options:
+    """Features object that has variables set for each DOMBuilder feature.
+
+    The DOMBuilder class uses an instance of this class to pass settings to
+    the ExpatBuilder class.
+    """
+
+    # Note that the DOMBuilder class in LoadSave constrains which of these
+    # values can be set using the DOM Level 3 LoadSave feature.
+
+    namespaces = 1
+    namespace_declarations = True
+    validation = False
+    external_parameter_entities = True
+    external_general_entities = True
+    external_dtd_subset = True
+    validate_if_schema = False
+    validate = False
+    datatype_normalization = False
+    create_entity_ref_nodes = True
+    entities = True
+    whitespace_in_element_content = True
+    cdata_sections = True
+    comments = True
+    charset_overrides_xml_encoding = True
+    infoset = False
+    supported_mediatypes_only = False
+
+    errorHandler = None
+    filter = None
+
+
+class DOMBuilder:
+    entityResolver = None
+    errorHandler = None
+    filter = None
+
+    ACTION_REPLACE = 1
+    ACTION_APPEND_AS_CHILDREN = 2
+    ACTION_INSERT_AFTER = 3
+    ACTION_INSERT_BEFORE = 4
+
+    _legal_actions = (ACTION_REPLACE, ACTION_APPEND_AS_CHILDREN,
+                      ACTION_INSERT_AFTER, ACTION_INSERT_BEFORE)
+
+    def __init__(self):
+        self._options = Options()
+
+    def _get_entityResolver(self):
+        return self.entityResolver
+    def _set_entityResolver(self, entityResolver):
+        self.entityResolver = entityResolver
+
+    def _get_errorHandler(self):
+        return self.errorHandler
+    def _set_errorHandler(self, errorHandler):
+        self.errorHandler = errorHandler
+
+    def _get_filter(self):
+        return self.filter
+    def _set_filter(self, filter):
+        self.filter = filter
+
+    def setFeature(self, name, state):
+        if self.supportsFeature(name):
+            state = state and 1 or 0
+            try:
+                settings = self._settings[(_name_xform(name), state)]
+            except KeyError:
+                raise xml.dom.NotSupportedErr(
+                    "unsupported feature: %r" % (name,))
+            else:
+                for name, value in settings:
+                    setattr(self._options, name, value)
+        else:
+            raise xml.dom.NotFoundErr("unknown feature: " + repr(name))
+
+    def supportsFeature(self, name):
+        return hasattr(self._options, _name_xform(name))
+
+    def canSetFeature(self, name, state):
+        key = (_name_xform(name), state and 1 or 0)
+        return key in self._settings
+
+    # This dictionary maps from (feature,value) to a list of
+    # (option,value) pairs that should be set on the Options object.
+    # If a (feature,value) setting is not in this dictionary, it is
+    # not supported by the DOMBuilder.
+    #
+    _settings = {
+        ("namespace_declarations", 0): [
+            ("namespace_declarations", 0)],
+        ("namespace_declarations", 1): [
+            ("namespace_declarations", 1)],
+        ("validation", 0): [
+            ("validation", 0)],
+        ("external_general_entities", 0): [
+            ("external_general_entities", 0)],
+        ("external_general_entities", 1): [
+            ("external_general_entities", 1)],
+        ("external_parameter_entities", 0): [
+            ("external_parameter_entities", 0)],
+        ("external_parameter_entities", 1): [
+            ("external_parameter_entities", 1)],
+        ("validate_if_schema", 0): [
+            ("validate_if_schema", 0)],
+        ("create_entity_ref_nodes", 0): [
+            ("create_entity_ref_nodes", 0)],
+        ("create_entity_ref_nodes", 1): [
+            ("create_entity_ref_nodes", 1)],
+        ("entities", 0): [
+            ("create_entity_ref_nodes", 0),
+            ("entities", 0)],
+        ("entities", 1): [
+            ("entities", 1)],
+        ("whitespace_in_element_content", 0): [
+            ("whitespace_in_element_content", 0)],
+        ("whitespace_in_element_content", 1): [
+            ("whitespace_in_element_content", 1)],
+        ("cdata_sections", 0): [
+            ("cdata_sections", 0)],
+        ("cdata_sections", 1): [
+            ("cdata_sections", 1)],
+        ("comments", 0): [
+            ("comments", 0)],
+        ("comments", 1): [
+            ("comments", 1)],
+        ("charset_overrides_xml_encoding", 0): [
+            ("charset_overrides_xml_encoding", 0)],
+        ("charset_overrides_xml_encoding", 1): [
+            ("charset_overrides_xml_encoding", 1)],
+        ("infoset", 0): [],
+        ("infoset", 1): [
+            ("namespace_declarations", 0),
+            ("validate_if_schema", 0),
+            ("create_entity_ref_nodes", 0),
+            ("entities", 0),
+            ("cdata_sections", 0),
+            ("datatype_normalization", 1),
+            ("whitespace_in_element_content", 1),
+            ("comments", 1),
+            ("charset_overrides_xml_encoding", 1)],
+        ("supported_mediatypes_only", 0): [
+            ("supported_mediatypes_only", 0)],
+        ("namespaces", 0): [
+            ("namespaces", 0)],
+        ("namespaces", 1): [
+            ("namespaces", 1)],
+    }
+
+    def getFeature(self, name):
+        xname = _name_xform(name)
+        try:
+            return getattr(self._options, xname)
+        except AttributeError:
+            if name == "infoset":
+                options = self._options
+                return (options.datatype_normalization
+                        and options.whitespace_in_element_content
+                        and options.comments
+                        and options.charset_overrides_xml_encoding
+                        and not (options.namespace_declarations
+                                 or options.validate_if_schema
+                                 or options.create_entity_ref_nodes
+                                 or options.entities
+                                 or options.cdata_sections))
+            raise xml.dom.NotFoundErr("feature %s not known" % repr(name))
+
+    def parseURI(self, uri):
+        if self.entityResolver:
+            input = self.entityResolver.resolveEntity(None, uri)
+        else:
+            input = DOMEntityResolver().resolveEntity(None, uri)
+        return self.parse(input)
+
+    def parse(self, input):
+        options = copy.copy(self._options)
+        options.filter = self.filter
+        options.errorHandler = self.errorHandler
+        fp = input.byteStream
+        if fp is None and options.systemId:
+            import urllib.request
+            fp = urllib.request.urlopen(input.systemId)
+        return self._parse_bytestream(fp, options)
+
+    def parseWithContext(self, input, cnode, action):
+        if action not in self._legal_actions:
+            raise ValueError("not a legal action")
+        raise NotImplementedError("Haven't written this yet...")
+
+    def _parse_bytestream(self, stream, options):
+        import xml.dom.expatbuilder
+        builder = xml.dom.expatbuilder.makeBuilder(options)
+        return builder.parseFile(stream)
+
+
+def _name_xform(name):
+    return name.lower().replace('-', '_')
+
+
+class DOMEntityResolver(object):
+    __slots__ = '_opener',
+
+    def resolveEntity(self, publicId, systemId):
+        assert systemId is not None
+        source = DOMInputSource()
+        source.publicId = publicId
+        source.systemId = systemId
+        source.byteStream = self._get_opener().open(systemId)
+
+        # determine the encoding if the transport provided it
+        source.encoding = self._guess_media_encoding(source)
+
+        # determine the base URI is we can
+        import posixpath, urllib.parse
+        parts = urllib.parse.urlparse(systemId)
+        scheme, netloc, path, params, query, fragment = parts
+        # XXX should we check the scheme here as well?
+        if path and not path.endswith("/"):
+            path = posixpath.dirname(path) + "/"
+            parts = scheme, netloc, path, params, query, fragment
+            source.baseURI = urllib.parse.urlunparse(parts)
+
+        return source
+
+    def _get_opener(self):
+        try:
+            return self._opener
+        except AttributeError:
+            self._opener = self._create_opener()
+            return self._opener
+
+    def _create_opener(self):
+        import urllib.request
+        return urllib.request.build_opener()
+
+    def _guess_media_encoding(self, source):
+        info = source.byteStream.info()
+        if "Content-Type" in info:
+            for param in info.getplist():
+                if param.startswith("charset="):
+                    return param.split("=", 1)[1].lower()
+
+
+class DOMInputSource(object):
+    __slots__ = ('byteStream', 'characterStream', 'stringData',
+                 'encoding', 'publicId', 'systemId', 'baseURI')
+
+    def __init__(self):
+        self.byteStream = None
+        self.characterStream = None
+        self.stringData = None
+        self.encoding = None
+        self.publicId = None
+        self.systemId = None
+        self.baseURI = None
+
+    def _get_byteStream(self):
+        return self.byteStream
+    def _set_byteStream(self, byteStream):
+        self.byteStream = byteStream
+
+    def _get_characterStream(self):
+        return self.characterStream
+    def _set_characterStream(self, characterStream):
+        self.characterStream = characterStream
+
+    def _get_stringData(self):
+        return self.stringData
+    def _set_stringData(self, data):
+        self.stringData = data
+
+    def _get_encoding(self):
+        return self.encoding
+    def _set_encoding(self, encoding):
+        self.encoding = encoding
+
+    def _get_publicId(self):
+        return self.publicId
+    def _set_publicId(self, publicId):
+        self.publicId = publicId
+
+    def _get_systemId(self):
+        return self.systemId
+    def _set_systemId(self, systemId):
+        self.systemId = systemId
+
+    def _get_baseURI(self):
+        return self.baseURI
+    def _set_baseURI(self, uri):
+        self.baseURI = uri
+
+
+class DOMBuilderFilter:
+    """Element filter which can be used to tailor construction of
+    a DOM instance.
+    """
+
+    # There's really no need for this class; concrete implementations
+    # should just implement the endElement() and startElement()
+    # methods as appropriate.  Using this makes it easy to only
+    # implement one of them.
+
+    FILTER_ACCEPT = 1
+    FILTER_REJECT = 2
+    FILTER_SKIP = 3
+    FILTER_INTERRUPT = 4
+
+    whatToShow = NodeFilter.SHOW_ALL
+
+    def _get_whatToShow(self):
+        return self.whatToShow
+
+    def acceptNode(self, element):
+        return self.FILTER_ACCEPT
+
+    def startContainer(self, element):
+        return self.FILTER_ACCEPT
+
+del NodeFilter
+
+
+class _AsyncDeprecatedProperty:
+    def warn(self, cls):
+        clsname = cls.__name__
+        warnings.warn(
+            "{cls}.async is deprecated; use {cls}.async_".format(cls=clsname),
+            DeprecationWarning)
+
+    def __get__(self, instance, cls):
+        self.warn(cls)
+        if instance is not None:
+            return instance.async_
+        return False
+
+    def __set__(self, instance, value):
+        self.warn(type(instance))
+        setattr(instance, 'async_', value)
+
+
+class DocumentLS:
+    """Mixin to create documents that conform to the load/save spec."""
+
+    async_ = False
+    locals()['async'] = _AsyncDeprecatedProperty()  # Avoid DeprecationWarning
+
+    def _get_async(self):
+        return False
+
+    def _set_async(self, flag):
+        if flag:
+            raise xml.dom.NotSupportedErr(
+                "asynchronous document loading is not supported")
+
+    def abort(self):
+        # What does it mean to "clear" a document?  Does the
+        # documentElement disappear?
+        raise NotImplementedError(
+            "haven't figured out what this means yet")
+
+    def load(self, uri):
+        raise NotImplementedError("haven't written this yet")
+
+    def loadXML(self, source):
+        raise NotImplementedError("haven't written this yet")
+
+    def saveXML(self, snode):
+        if snode is None:
+            snode = self
+        elif snode.ownerDocument is not self:
+            raise xml.dom.WrongDocumentErr()
+        return snode.toxml()
+
+
+del _AsyncDeprecatedProperty
+
+
+class DOMImplementationLS:
+    MODE_SYNCHRONOUS = 1
+    MODE_ASYNCHRONOUS = 2
+
+    def createDOMBuilder(self, mode, schemaType):
+        if schemaType is not None:
+            raise xml.dom.NotSupportedErr(
+                "schemaType not yet supported")
+        if mode == self.MODE_SYNCHRONOUS:
+            return DOMBuilder()
+        if mode == self.MODE_ASYNCHRONOUS:
+            raise xml.dom.NotSupportedErr(
+                "asynchronous builders are not supported")
+        raise ValueError("unknown value for mode")
+
+    def createDOMWriter(self):
+        raise NotImplementedError(
+            "the writer interface hasn't been written yet!")
+
+    def createDOMInputSource(self):
+        return DOMInputSource()