]> git.saurik.com Git - wxWidgets.git/blobdiff - wxPython/samples/ide/activegrid/util/xmlmarshaller.py
patch 1444066
[wxWidgets.git] / wxPython / samples / ide / activegrid / util / xmlmarshaller.py
index f305f0ada82f6b3c515f37f74783385f03c1d501..6a74d2c486bd2805438516918170f6f2d33f6900 100644 (file)
@@ -2,27 +2,31 @@
 # Name:         xmlmarshaller.py
 # Purpose:
 #
-# Author:       John Spurling
+# Authors:       John Spurling, Joel Hare, Alan Mullendore
 #
 # Created:      7/28/04
 # CVS-ID:       $Id$
 # Copyright:    (c) 2004-2005 ActiveGrid, Inc.
 # License:      wxWindows License
 #----------------------------------------------------------------------------
-from activegrid import util
-import inspect
+import sys
 from types import *
+from activegrid.util.lang import *
+import logging
+ifDefPy()
 import xml.sax
 import xml.sax.handler
-import __builtin__
-from xml.sax import saxutils
+endIfDef()
+import xml.sax.saxutils as saxutils
+import activegrid.util.objutils as objutils
+import activegrid.util.aglogging as aglogging
 
-### ToDO remove maxOccurs "unbounded" resolves to -1 hacks after bug 177 is fixed
+MODULE_PATH = "__main__"
 
-"""
+## ToDO remove maxOccurs "unbounded" resolves to -1 hacks after bug 177 is fixed
 
-More documentation later, but here are some special Python attributes
-that McLane recognizes:
+"""
+Special attributes that we recognize:
 
 name: __xmlname__
 type: string
@@ -30,26 +34,26 @@ description: the name of the xml element for the marshalled object
 
 name: __xmlattributes__
 type: tuple or list
-description: the name(s) of the Python string attribute(s) to be
+description: the name(s) of the Lang string attribute(s) to be
 marshalled as xml attributes instead of nested xml elements. currently
-these can only be strings since there's not a way to get the type
+these can only be strings since there"s not a way to get the type
 information back when unmarshalling.
 
 name: __xmlexclude__
 type: tuple or list
-description: the name(s) of the python attribute(s) to skip when
+description: the name(s) of the lang attribute(s) to skip when
 marshalling.
 
 name: __xmlrename__
 type: dict
-description: describes an alternate Python <-> XML name mapping.  
+description: describes an alternate Lang <-> XML name mapping.  
 Normally the name mapping is the identity function.  __xmlrename__
-overrides that.  The keys are the Python names, the values are their
+overrides that.  The keys are the Lang names, the values are their
 associated XML names.
 
 name: __xmlflattensequence__
 type: dict, tuple, or list
-description: the name(s) of the Python sequence attribute(s) whose
+description: the name(s) of the Lang sequence attribute(s) whose
 items are to be marshalled as a series of xml elements (with an
 optional keyword argument that specifies the element name to use) as
 opposed to containing them in a separate sequence element, e.g.:
@@ -57,12 +61,12 @@ opposed to containing them in a separate sequence element, e.g.:
 myseq = (1, 2)
 <!-- normal way of marshalling -->
 <myseq>
-  <item objtype='int'>1</item>
-  <item objtype='int'>2</item>
+  <item objtype="int">1</item>
+  <item objtype="int">2</item>
 </myseq>
-<!-- with __xmlflattensequence__ set to {'myseq': 'squish'} -->
-<squish objtype='int'>1</squish>
-<squish objtype='int'>2</squish>
+<!-- with __xmlflattensequence__ set to {"myseq": "squish"} -->
+<squish objtype="int">1</squish>
+<squish objtype="int">2</squish>
 
 name: __xmlnamespaces__
 type: dict
@@ -79,16 +83,36 @@ should be used as the default namespace for the object.
 
 name: __xmlattrnamespaces__
 type: dict
-description: a dict assigning the Python object's attributes to the namespaces
+description: a dict assigning the Lang object"s attributes to the namespaces
 defined in __xmlnamespaces__.  Each item in the dict should consist of a
 prefix,attributeList combination where the key is the prefix and the value is
-a list of the Python attribute names.  e.g.:
+a list of the Lang attribute names.  e.g.:
 
 __xmlattrnamespaces__ = { "ag":["firstName", "lastName", "addressLine1", "city"] }
 
+name: __xmlattrgroups__
+type: dict
+description: a dict specifying groups of attributes to be wrapped in an enclosing tag.
+The key is the name of the enclosing tag; the value is a list of attributes to include
+within it. e.g.
+
+__xmlattrgroups__ = {"name": ["firstName", "lastName"], "address": ["addressLine1", "city", "state", "zip"]}
+
+name: __xmlcdatacontent__
+type: string
+description: value is the name of a string attribute that should be assigned CDATA content from the
+source document and that should be marshalled as CDATA.
+
+__xmlcdatacontent__ = "messyContent"
 
 """
 
+global xmlMarshallerLogger
+xmlMarshallerLogger = logging.getLogger("activegrid.util.xmlmarshaller.marshal")
+xmlMarshallerLogger.setLevel(aglogging.LEVEL_WARN)
+# INFO  : low-level info
+# DEBUG : debugging info
+
 ################################################################################
 #
 # module exceptions
@@ -109,7 +133,7 @@ class UnhandledTypeException(Error):
         return "%s is not supported for marshalling." % str(self.typename)
 
 class XMLAttributeIsNotStringType(Error):
-    """Exception raised when an object's attribute is specified to be
+    """Exception raised when an object"s attribute is specified to be
     marshalled as an XML attribute of the enclosing object instead of
     a nested element.
     """
@@ -118,37 +142,40 @@ class XMLAttributeIsNotStringType(Error):
         self.typename = typename
     def __str__(self):
         return """%s was set to be marshalled as an XML attribute
-        instead of a nested element, but the object's type is %s, not
+        instead of a nested element, but the object"s type is %s, not
         string.""" % (self.attrname, self.typename)
 
+class MarshallerException(Exception):
+    pass
+
+class UnmarshallerException(Exception):
+    pass 
+
 ################################################################################
 #
 # constants and such
 #
 ################################################################################
 
-XMLNS = 'xmlns'
-XMLNS_PREFIX = XMLNS + ':'
+XMLNS = "xmlns"
+XMLNS_PREFIX = XMLNS + ":"
 XMLNS_PREFIX_LENGTH = len(XMLNS_PREFIX)
+DEFAULT_NAMESPACE_KEY = "__DEFAULTNS__"
+TYPE_QNAME = "QName"
+XMLSCHEMA_XSD_URL = "http://www.w3.org/2001/XMLSchema"
+AG_URL = "http://www.activegrid.com/ag.xsd"
 
-BASETYPE_ELEMENT_NAME = 'item'
-MEMBERS_TO_SKIP = ('__module__', '__doc__', '__xmlname__', '__xmlattributes__',
-                   '__xmlexclude__', '__xmlflattensequence__', '__xmlnamespaces__',
-                   '__xmldefaultnamespace__', '__xmlattrnamespaces__')
-
-WELL_KNOWN_OBJECTS = { "xs:element"     : "activegrid.model.schema.XsdElement",
-                       "xs:complexType" : "activegrid.model.schema.XsdComplexType",
-                       "xs:complexType" : "activegrid.model.schema.XsdComplexType",
-                       "xs:element"     : "activegrid.model.schema.XsdElement",
-                       "xs:key"         : "activegrid.model.schema.XsdKey",
-                       "xs:field"       : "activegrid.model.schema.XsdKeyField",
-                       "xs:keyref"      : "activegrid.model.schema.XsdKeyRef",
-                       "xs:selector"    : "activegrid.model.schema.XsdKeySelector",              
-                       "xs:schema"      : "activegrid.model.schema.Schema",
-                       "ag:schemaOptions":"activegrid.model.schema.SchemaOptions",
-                       "ag:debug"       : "activegrid.model.processmodel.DebugOperation",
-                     }
+BASETYPE_ELEMENT_NAME = "item"
+DICT_ITEM_NAME = "qqDictItem"
+DICT_ITEM_KEY_NAME = "key"
+DICT_ITEM_VALUE_NAME = "value"
 
+# This list doesn"t seem to be used.
+#   Internal documentation or useless? You make the call!
+##MEMBERS_TO_SKIP = ("__module__", "__doc__", "__xmlname__", "__xmlattributes__",
+##                   "__xmlexclude__", "__xmlflattensequence__", "__xmlnamespaces__",
+##                   "__xmldefaultnamespace__", "__xmlattrnamespaces__",
+##                   "__xmlattrgroups__")
 
 ################################################################################
 #
@@ -156,180 +183,351 @@ WELL_KNOWN_OBJECTS = { "xs:element"     : "activegrid.model.schema.XsdElement",
 #
 ################################################################################
 
-def _objectfactory(objname, objargs=None, xsname=None):
-    try:
-        '''dynamically create an object based on the objname and return
-        it. look it up in the BASETYPE_ELEMENT_MAP first.
-        '''
-##        print "_objectfactory creating an object of type %s and value %s, xsname=%s" % (objname, objargs, xsname)
-        # split the objname into the typename and module path,
-        # importing the module if need be.
-        if not isinstance(objargs, list):
-            objargs = [objargs]
-            
-        if (xsname):
-            try:
-                objname = WELL_KNOWN_OBJECTS[xsname]
-            except KeyError:
-                pass
-
-        objtype = objname.split('.')[-1]
-        pathlist = objname.split('.')
-        modulename = '.'.join(pathlist[0:-1])
-
-##        print "[objectfactory] objtype is %s" % objtype
-##        print "[objectfactory] objargs is %s" % `objargs`
-
-        ## since the bool constructor will turn a string of non-zero
-        ## length into True, we call it with no argument (yielding a
-        ## False) if the string contains 'false'
-        if objtype == 'bool' and objargs[0].lower() == 'false':
-            objargs = None
-
-##        if objtype == 'str':
-##            print type(objargs)
-##            print "string we're unescaping: '%s'" % objargs[0]
-##            objargs = saxutils.unescape(objargs[0])
-        if objtype in ('float', 'int', 'str', 'long'):
-            objargs = [x.strip() for x in objargs]
-
-        if objtype == 'str':
-            objargs = [saxutils.unescape(x) for x in objargs]
-
-        if __builtin__.__dict__.has_key(objname):
-            module = __builtin__
-        else:
-            if modulename:
-                module = __import__(modulename)
-            for name in pathlist[1:-1]:
-                module = module.__dict__[name]
-        if objargs:
-            return module.__dict__[objtype](*objargs)
+def setattrignorecase(object, name, value):
+    if (name not in object.__dict__):
+        namelow = name.lower()
+        for attr in object.__dict__:
+            if attr.lower() == namelow:
+                object.__dict__[attr] = value
+                return
+    object.__dict__[name] = value
+
+def getComplexType(obj):
+    if (hasattr(obj, "__xsdcomplextype__")):
+        return obj.__xsdcomplextype__
+    return None
+
+def _objectfactory(objname, objargs=None, objclass=None):
+    "dynamically create an object based on the objname and return it."
+##    print "[objectfactory] objname [%s]" % (objname)
+    if not isinstance(objargs, list):
+        objargs = [objargs]
+    if (objclass != None):
+        if (len(objargs) > 0):
+            if (hasattr(objclass, "__xmlcdatacontent__")):
+                obj = objclass()
+                contentAttr = obj.__xmlcdatacontent__
+                obj.__dict__[contentAttr] = str(objargs[0])
+                return obj
+            return objclass(*objargs)
         else:
-            if objtype == 'None':
-                return None
-            return module.__dict__[objtype]()
-    except KeyError:
-        raise KeyError("Could not find class %s" % objname)
+            return objclass()
+    return objutils.newInstance(objname, objargs)
 
 class Element:
-    def __init__(self, name, attrs=None):
+    def __init__(self, name, attrs=None, xsname=None):
         self.name = name
         self.attrs = attrs
-        self.content = ''
+        self.content = ""
         self.children = []
+        self.objclass = None
+        self.xsname = xsname
+        
     def getobjtype(self):
-        if self.attrs.has_key('objtype'):
-            return self.attrs.getValue('objtype')
+        objtype = self.attrs.get("objtype")
+        if (objtype == None):
+            if (len(self.children) > 0):
+                objtype = "dict"
+            else:
+                objtype = "str"
+        return objtype
+            
+class NsElement(object):
+    def __init__(self):
+        self.nsMap = {}
+        self.targetNS = None
+        self.defaultNS = None
+        self.prefix = None
+        
+    def isEmpty(self):
+        return ((self.nsMap == {}) and (self.targetNS == None) and (self.defaultNS == None))
+
+    def setKnownTypes(self, masterKnownTypes, masterKnownNamespaces, parentNSE):
+        # if we're a nested element, extend our parent element's mapping
+        if parentNSE != None:
+            self.knownTypes = parentNSE.knownTypes.copy()
+            # but if we have a different default namespace, replace the parent's default mappings
+            if parentNSE.defaultNS != self.defaultNS:
+                newKT = self.knownTypes.copy()
+                for tag in newKT:
+                    if tag.find(':') < 0:
+                        del self.knownTypes[tag]
+            newMap = parentNSE.nsMap.copy()
+            if self.nsMap != {}:
+                for k, v in self.nsMap.iteritems():
+                    newMap[k] = v
+            self.nsMap = newMap
         else:
-            return 'str'
+            self.knownTypes = {}
+        reversedKNS = {}
+        # TODO: instead of starting with the knownNamespaces, start with the "xmlms" mappings
+        # for this element. Then we'd only process the namespaces and tags we need to.
+        # But for now, this works.
+        for long, short in masterKnownNamespaces.iteritems():
+            reversedKNS[short] = long
+        mapLongs = self.nsMap.values()
+        for tag, mapClass in masterKnownTypes.iteritems():
+            i = tag.rfind(':')
+            if i >= 0:                                      # e.g. "wsdl:description"
+                knownTagShort = tag[:i]                     # "wsdl"
+                knownTagName = tag[i+1:]                    # "description"
+                knownTagLong  = reversedKNS[knownTagShort]  # e.g. "http://schemas.xmlsoap.org/wsdl"
+                if (knownTagLong in mapLongs):
+                    for mShort, mLong in self.nsMap.iteritems():
+                        if mLong == knownTagLong:
+                            actualShort = mShort  # e.g. "ws"
+                            actualTag = '%s:%s' % (actualShort, knownTagName)
+                            self.knownTypes[actualTag] = mapClass
+                            break
+                if self.defaultNS == knownTagLong:
+                    self.knownTypes[knownTagName] = mapClass
+            else:                                           # e.g. "ItemSearchRequest"
+                self.knownTypes[tag] = mapClass
+##                print 'mapping <%s> to class "%s"' % (tag, mapClass.__name__)
 
+    def expandQName(self, eName, attrName, attrValue):
+        bigValue = attrValue
+        i = attrValue.rfind(':')
+        if (i < 0):
+            if self.defaultNS != None:
+                bigValue = '%s:%s' % (self.defaultNS, attrValue)
+        else:
+            attrNS = attrValue[:i]
+            attrNCName = attrValue[i+1:]
+            for shortNs, longNs in self.nsMap.iteritems():
+                if shortNs == attrNS:
+                    bigValue = '%s:%s' % (longNs, attrNCName)
+                    break
+##        print '[expandQName] input attrName = "%s" and attrValue "%s"; output = "%s"' % (attrName, attrValue, bigValue)
+        return bigValue
 
 class XMLObjectFactory(xml.sax.ContentHandler):
-    def __init__(self):
+    def __init__(self, knownTypes=None, knownNamespaces=None):
         self.rootelement = None
+        if (knownTypes == None):
+            self.knownTypes = {}
+        else:
+            self.knownTypes = knownTypes
+        if (knownNamespaces == None):
+            self.knownNamespaces = {}
+        else:
+            self.knownNamespaces = knownNamespaces
+        self.skipper = False
         self.elementstack = []
+        self.nsstack = []
+        self.collectContent = None
         xml.sax.handler.ContentHandler.__init__(self)
 
+    def appendElementStack(self, newElement, newNS):
+        self.elementstack.append(newElement)
+        if (newNS.isEmpty()):
+            if (len(self.nsstack) > 0):
+                newNS = self.nsstack[-1]
+            else:
+                newNS.knownTypes = self.knownTypes.copy()
+        else:
+            if (len(self.nsstack) > 0):
+                newNS.setKnownTypes(self.knownTypes, self.knownNamespaces, self.nsstack[-1])
+            else:
+                newNS.setKnownTypes(self.knownTypes, self.knownNamespaces, None)
+        self.nsstack.append(newNS)
+        return newNS
+    
+    def popElementStack(self):
+        element = self.elementstack.pop()
+        nse = self.nsstack.pop()
+        return element, nse
+        
     ## ContentHandler methods
     def startElement(self, name, attrs):
+##        print '[startElement] <%s>' % (name)
+        if name == 'xs:annotation' or name == 'xsd:annotation': # should use namespace mapping here
+            self.skipper = True
+            self.appendElementStack(Element(name, attrs.copy()), NsElement())
+        if self.skipper:
+            return
+        if self.collectContent != None:
+            strVal = '<%s' % (name)
+            for aKey, aVal in attrs.items():
+                strVal += (' %s="%s"' % (aKey, aVal))
+            strVal += '>'
+            self.collectContent.content += strVal
+        xsname = name
         if name.find(':') > -1:  # Strip namespace prefixes for now until actually looking them up in xsd
-            name = name[name.index(':') + 1:]
-##        for attrname in attrs.getNames():
-##            print "%s: %s" % (attrname, attrs.getValue(attrname))
-        element = Element(name, attrs.copy())
-        self.elementstack.append(element)
-##        print self.elementstack
+            name = name[name.rfind(":") + 1:]
+        element = Element(name, attrs.copy(), xsname=xsname)
+        # if the element has namespace attributes, process them and add them to our stack
+        nse = NsElement()
+        for k in attrs.getNames():
+            if k.startswith('xmlns'):
+                longNs = attrs[k]
+                eLongNs = longNs + '/'
+                if str(eLongNs) in asDict(self.knownNamespaces):
+                    longNs = eLongNs
+                if k == 'xmlns':
+                    nse.defaultNS = longNs
+                else:
+                    shortNs = k[6:]
+                    nse.nsMap[shortNs] = longNs
+            elif k == 'targetNamespace':
+                nse.targetNS = attrs.getValue(k)
+        nse = self.appendElementStack(element, nse)
+        element.objclass = nse.knownTypes.get(xsname)
+        if (hasattr(element.objclass, "__xmlcontent__")):
+            self.collectContent = element
 
     def characters(self, content):
-##        print "got content: %s" % content
-        if content:
-            self.elementstack[-1].content += content
+##        print '[characters] "%s" (%s)' % (content, type(content))
+        if (content != None):
+            if self.collectContent != None:
+                self.collectContent.content += content
+            else:
+                self.elementstack[-1].content += content
 
     def endElement(self, name):
-##        print "[endElement] name of element we're at the end of: %s" % name
+##        print "[endElement] </%s>" % name
         xsname = name
-        if name.find(':') > -1:  # Strip namespace prefixes for now until actually looking them up in xsd
-            name = name[name.index(':') + 1:]
-        element = self.elementstack.pop()
+        if name.find(":") > -1:  # Strip namespace prefixes for now until actually looking them up in xsd
+            name = name[name.find(":") + 1:]
+        if self.skipper:
+            if xsname == "xs:annotation" or xsname == "xsd:annotation":     # here too
+                self.skipper = False
+                self.popElementStack()
+            return
+        if self.collectContent != None:
+            if xsname != self.collectContent.xsname:
+                self.collectContent.content += ('</%s>' % (xsname))
+                self.popElementStack()
+                return
+            else:
+                self.collectContent = None
+        oldChildren = self.elementstack[-1].children
+        element, nse = self.popElementStack()
+        if ((len(self.elementstack) > 1) and (self.elementstack[-1].getobjtype() == "None")):
+            parentElement = self.elementstack[-2]
+##            print "[endElement] %s: found parent with objtype==None: using its grandparent" % name
+        elif (len(self.elementstack) > 0):
+            parentElement = self.elementstack[-1]
         objtype = element.getobjtype()
+##        print "element objtype is: ", objtype
+        if (objtype == "None"):
+##            print "[endElement] %s: skipping a (objtype==None) end tag" % name
+            return
         constructorarglist = []
-        if element.content:
+        if (len(element.content) > 0):
             strippedElementContent = element.content.strip()
-            if strippedElementContent:
+            if (len(strippedElementContent) > 0):
                 constructorarglist.append(element.content)
-        obj = _objectfactory(objtype, constructorarglist, xsname)
-        complexType = None
-        if hasattr(obj, '__xsdcomplextype__'):
-            complexType = getattr(obj, '__xsdcomplextype__')
-        if len(self.elementstack) > 0:
-            self.elementstack[-1].children.append((name, obj))
-        else:
-            self.rootelement = obj
-        if element.attrs and not isinstance(obj, list):
+        obj = _objectfactory(objtype, constructorarglist, element.objclass)
+        complexType = getComplexType(obj)
+        if (obj != None):
+            if (hasattr(obj, "__xmlname__") and getattr(obj, "__xmlname__") == "sequence"):
+                self.elementstack[-1].children = oldChildren
+                return
+        if (len(element.attrs) > 0) and not isinstance(obj, list):
+##            print "[endElement] %s: element has attrs and the obj is not a list" % name
             for attrname, attr in element.attrs.items():
                 if attrname == XMLNS or attrname.startswith(XMLNS_PREFIX):
                     if attrname.startswith(XMLNS_PREFIX):
                         ns = attrname[XMLNS_PREFIX_LENGTH:]
                     else:
                         ns = ""
-                    if not hasattr(obj, '__xmlnamespaces__'):
-                        obj.__xmlnamespaces__ = {ns:attr}
-                    elif ns not in obj.__xmlnamespaces__:
-                        if (hasattr(obj.__class__, '__xmlnamespaces__') 
-                           and obj.__xmlnamespaces__ is obj.__class__.__xmlnamespaces__):
+                    if complexType != None:
+                        if not hasattr(obj, "__xmlnamespaces__"):
+                            obj.__xmlnamespaces__ = {ns:attr}
+                        elif ns not in obj.__xmlnamespaces__:
+                            if (hasattr(obj.__class__, "__xmlnamespaces__") 
+                                and (obj.__xmlnamespaces__ is obj.__class__.__xmlnamespaces__)):
                                 obj.__xmlnamespaces__ = dict(obj.__xmlnamespaces__)
-                        obj.__xmlnamespaces__[ns] = attr
-                elif not attrname == 'objtype':
-                    if attrname.find(':') > -1:  # Strip namespace prefixes for now until actually looking them up in xsd
-                        attrname = attrname[attrname.index(':') + 1:]
-                    if complexType:
+                            obj.__xmlnamespaces__[ns] = attr
+                elif not attrname == "objtype":
+                    if attrname.find(":") > -1:  # Strip namespace prefixes for now until actually looking them up in xsd
+                        attrname = attrname[attrname.find(":") + 1:]
+                    if (complexType != None):
                         xsdElement = complexType.findElement(attrname)
-                        if xsdElement:
+                        if (xsdElement != None):
                             type = xsdElement.type
-                            if type:
-                                type = xsdToPythonType(type)
+##                            print 'Unmarshalling element "%s", attribute "%s" with type "%s"' % (name, xsdElement.name, type)
+                            if (type != None):
+                                if (type == TYPE_QNAME):
+                                    attr = nse.expandQName(name, attrname, attr)
+                                type = xsdToLangType(type)
                                 ### ToDO remove maxOccurs hack after bug 177 is fixed
                                 if attrname == "maxOccurs" and attr == "unbounded":
                                     attr = "-1"
                                 attr = _objectfactory(type, attr)
-                    util.setattrignorecase(obj, _toAttrName(obj, attrname), attr)
+                    try:
+                        setattrignorecase(obj, _toAttrName(obj, attrname), attr)
+                    except AttributeError:
+                        errorString = 'Error unmarshalling XML document at line %i, column %i: The object type of attribute "%s" of XML element "%s": not specified or known' % (self._locator.getLineNumber(), self._locator.getColumnNumber(), attrname, name)
+                        raise UnmarshallerException(errorString)
 ##                    obj.__dict__[_toAttrName(obj, attrname)] = attr
         # stuff any child attributes meant to be in a sequence via the __xmlflattensequence__
         flattenDict = {}
-        if hasattr(obj, '__xmlflattensequence__'):
-            for sequencename, xmlnametuple in obj.__xmlflattensequence__.items():
-                for xmlname in xmlnametuple:
-                    flattenDict[xmlname] = sequencename
+        if hasattr(obj, "__xmlflattensequence__"):
+            flatten = obj.__xmlflattensequence__
+            if (isinstance(flatten, dict)):
+                for sequencename, xmlnametuple in flatten.items():
+                    if (xmlnametuple == None):
+                        flattenDict[sequencename] = sequencename
+                    elif (not isinstance(xmlnametuple, (tuple, list))):
+                        flattenDict[str(xmlnametuple)] = sequencename
+                    else:
+                        for xmlname in xmlnametuple:
+##                            print "[endElement]: adding flattenDict[%s] = %s" % (xmlname, sequencename)
+                            flattenDict[xmlname] = sequencename
+            else:
+                raise Exception("Invalid type for __xmlflattensequence___ : it must be a dict")
 
-        # reattach an object's attributes to it
+        # reattach an object"s attributes to it
         for childname, child in element.children:
-            if flattenDict.has_key(childname):
+##            print "[endElement] childname is: ", childname, "; child is: ", child
+            if (childname in flattenDict):
                 sequencename = _toAttrName(obj, flattenDict[childname])
-                try:
-                    sequencevalue = obj.__dict__[sequencename]
-                except AttributeError:
-                    sequencevalue = None
-                if sequencevalue == None:
-                    sequencevalue = []
-                    obj.__dict__[sequencename] = sequencevalue
+                if (not hasattr(obj, sequencename)):
+                    obj.__dict__[sequencename] = []
+                sequencevalue = getattr(obj, sequencename)
+                if (sequencevalue == None):
+                    obj.__dict__[sequencename] = []
+                    sequencevalue = getattr(obj, sequencename)
                 sequencevalue.append(child)
-            elif isinstance(obj, list):
+            elif (objtype == "list"):
                 obj.append(child)
+            elif isinstance(obj, dict):
+                if (childname == DICT_ITEM_NAME):
+                    obj[child[DICT_ITEM_KEY_NAME]] = child[DICT_ITEM_VALUE_NAME]
+                else:
+                    obj[childname] = child
             else:
 ##                print "childname = %s, obj = %s, child = %s" % (childname, repr(obj), repr(child))
-                util.setattrignorecase(obj, _toAttrName(obj, childname), child)
+                try:
+                    setattrignorecase(obj, _toAttrName(obj, childname), child)
+                except AttributeError:
+                    raise MarshallerException("Error unmarshalling child element \"%s\" of XML element \"%s\": object type not specified or known" % (childname, name))
 ##                obj.__dict__[_toAttrName(obj, childname)] = child
 
-        if complexType:
+        if (complexType != None):
             for element in complexType.elements:
                 if element.default:
                     elementName = _toAttrName(obj, element.name)
                     if ((elementName not in obj.__dict__) or (obj.__dict__[elementName] == None)):
-                        pythonType = xsdToPythonType(element.type)
-                        defaultValue = _objectfactory(pythonType, element.default)
+                        langType = xsdToLangType(element.type)
+                        defaultValue = _objectfactory(langType, element.default)
                         obj.__dict__[elementName] = defaultValue
 
+        ifDefPy()
+        if (isinstance(obj, list)):
+            if ((element.attrs.has_key("mutable")) and (element.attrs.getValue("mutable") == "false")):
+                obj = tuple(obj)
+        endIfDef()
+            
+        if (len(self.elementstack) > 0):
+##            print "[endElement] appending child with name: ", name, "; objtype: ", objtype
+            parentElement.children.append((name, obj))
+##            print "parentElement now has ", len(parentElement.children), " children"
+        else:
+            self.rootelement = obj
+            
     def getRootObject(self):
         return self.rootelement
 
@@ -343,307 +541,528 @@ def _toAttrName(obj, name):
 ##        name = "_%s%s" % (obj.__class__.__name__, name)
     return name
 
-__typeMappingXsdToPython = {
+__typeMappingXsdToLang = {
     "string": "str",
     "char": "str",
     "varchar": "str",
-    "date": "str", # ToDO Need to work out how to create python date types
+    "date": "str", # ToDO Need to work out how to create lang date types
     "boolean": "bool",
     "decimal": "float", # ToDO Does python have a better fixed point type?
     "int": "int",
+    "integer":"int",
     "long": "long",
     "float": "float",
     "bool": "bool",
     "str": "str",
     "unicode": "unicode",
+    "short": "int",
+    "duration": "str", # see above (date)
+    "datetime": "str", # see above (date)
+    "time": "str", # see above (date)
+    "double": "float",
+    "QName" : "str",
+    "blob" : "str",         # ag:blob
+    "currency" : "str",     # ag:currency
     }    
 
-def xsdToPythonType(xsdType):
-    try:
-        return __typeMappingXsdToPython[xsdType]
-    except KeyError:
+def xsdToLangType(xsdType):
+    if xsdType.startswith(XMLSCHEMA_XSD_URL):
+        xsdType = xsdType[len(XMLSCHEMA_XSD_URL)+1:]
+    elif xsdType.startswith(AG_URL):
+        xsdType = xsdType[len(AG_URL)+1:]        
+    langType = __typeMappingXsdToLang.get(xsdType)
+    if (langType == None):
         raise Exception("Unknown xsd type %s" % xsdType)
+    return langType
+    
+def langToXsdType(langType):
+    if langType in asDict(__typeMappingXsdToLang):
+        return '%s:%s' % (XMLSCHEMA_XSD_URL, langType)
+    return langType
 
-def _getXmlValue(pythonValue):
-    if (isinstance(pythonValue, bool)):
-        return str(pythonValue).lower()
+def _getXmlValue(langValue):
+    if (isinstance(langValue, bool)):
+        return str(langValue).lower()
+    elif (isinstance(langValue, unicode)):
+        return langValue.encode()
     else:
-        return str(pythonValue)
+        return str(langValue)
 
-def unmarshal(xmlstr):
-    objectfactory = XMLObjectFactory()
-    xml.sax.parseString(xmlstr, objectfactory)
+def unmarshal(xmlstr, knownTypes=None, knownNamespaces=None, xmlSource=None):
+    objectfactory = XMLObjectFactory(knownTypes, knownNamespaces)
+    try:
+        xml.sax.parseString(xmlstr, objectfactory)
+    except xml.sax.SAXParseException, errorData:
+        if xmlSource == None:
+            xmlSource = 'unknown'
+        errorString = 'SAXParseException ("%s") detected at line %d, column %d in XML document from source "%s" ' % (errorData.getMessage(), errorData.getLineNumber(), errorData.getColumnNumber(), xmlSource)
+        raise UnmarshallerException(errorString)
     return objectfactory.getRootObject()
 
-
-def marshal(obj, elementName=None, nameSpacePrefix='', nameSpaces=None, prettyPrint=False, indent=0):
-    if prettyPrint or indent:
-        prefix = ' '*indent
-        newline = '\n'
-        increment = 4
+def marshal(obj, elementName=None, prettyPrint=False, marshalType=True, indent=0, knownTypes=None, knownNamespaces=None, encoding=-1):
+##    print '[marshal] entered with elementName = "%s"' % (elementName)
+    worker = XMLMarshalWorker(prettyPrint=prettyPrint, marshalType=marshalType, knownTypes=knownTypes, knownNamespaces=knownNamespaces)    
+    if obj != None and hasattr(obj, '__xmldeepexclude__'):
+        worker.xmldeepexclude = obj.__xmldeepexclude__
+    xmlstr = "".join(worker._marshal(obj, elementName, indent=indent))
+    if (isinstance(encoding, basestring)):
+        return '<?xml version="1.0" encoding="%s"?>\n%s' % (encoding, xmlstr.encode(encoding))
+    elif (encoding == None):
+        return xmlstr
     else:
-        prefix = ''
-        newline = ''
-        increment = 0
-
-    ## Determine the XML element name. If it isn't specified in the
-    ## parameter list, look for it in the __xmlname__ Python
-    ## attribute, else use the default generic BASETYPE_ELEMENT_NAME.
-    if not nameSpaces: nameSpaces = {}  # Need to do this since if the {} is a default parameter it gets shared by all calls into the function
-    nameSpaceAttrs = ''
-    if hasattr(obj, '__xmlnamespaces__'):
-        for nameSpaceKey, nameSpaceUrl in getattr(obj, '__xmlnamespaces__').items():
-            if nameSpaceUrl in nameSpaces:
-                nameSpaceKey = nameSpaces[nameSpaceUrl]
-            else:
-##                # TODO: Wait to do this until there is shared state for use when going through the object graph
-##                origNameSpaceKey = nameSpaceKey  # Make sure there is no key collision, ie: same key referencing two different URL's
-##                i = 1
-##                while nameSpaceKey in nameSpaces.values():
-##                    nameSpaceKey = origNameSpaceKey + str(i)
-##                    i += 1
-                nameSpaces[nameSpaceUrl] = nameSpaceKey
-                if nameSpaceKey == '':
-                    nameSpaceAttrs += ' xmlns="%s" ' % (nameSpaceUrl)
-                else:
-                    nameSpaceAttrs += ' xmlns:%s="%s" ' % (nameSpaceKey, nameSpaceUrl)
-        nameSpaceAttrs = nameSpaceAttrs.rstrip()
-    if hasattr(obj, '__xmldefaultnamespace__'):
-        nameSpacePrefix = getattr(obj, '__xmldefaultnamespace__') + ':'
-    if not elementName:
-        if hasattr(obj, '__xmlname__'):
-            elementName = nameSpacePrefix + obj.__xmlname__
-        else:
-            elementName = nameSpacePrefix + BASETYPE_ELEMENT_NAME
-    else:
-        elementName = nameSpacePrefix + elementName
-        
-    members_to_skip = []
-    ## Add more members_to_skip based on ones the user has selected
-    ## via the __xmlexclude__ attribute.
-    if hasattr(obj, '__xmlexclude__'):
-        members_to_skip += list(obj.__xmlexclude__)
-    # Marshal the attributes that are selected to be XML attributes.
-    objattrs = ''
-    className = obj.__class__.__name__
-    classNamePrefix = "_" + className
-    if hasattr(obj, '__xmlattributes__'):
-        xmlattributes = obj.__xmlattributes__
-        members_to_skip += xmlattributes
-        for attr in xmlattributes:
-            internalAttrName = attr
-            if (attr.startswith("__") and not attr.endswith("__")): 
-                internalAttrName = classNamePrefix + attr
-            # Fail silently if a python attribute is specified to be
-            # an XML attribute but is missing.
-            try:
-                value = obj.__dict__[internalAttrName]
-            except KeyError:
-                continue
-##                # But, check and see if it is a property first:
-##                if (hasPropertyValue(obj, attr)):
-##                    value = getattr(obj, attr)
-##                else:
-##                    continue
-            xsdElement = None
-            if hasattr(obj, '__xsdcomplextype__'):
-                complexType = getattr(obj, '__xsdcomplextype__')
-                xsdElement = complexType.findElement(attr)
-            if xsdElement:
-                default = xsdElement.default
-                if default == value or default == _getXmlValue(value):
-                    continue
-            elif value == None:
-                continue
-
-            # ToDO remove maxOccurs hack after bug 177 is fixed
-            if attr == "maxOccurs" and value == -1:
-                value = "unbounded"
-
-            if isinstance(value, bool):
-               if value == True:
-                   value = "true"
-               else:
-                   value = "false"
-
-            attrNameSpacePrefix = ''
-            if hasattr(obj, '__xmlattrnamespaces__'):
-                for nameSpaceKey, nameSpaceAttributes in getattr(obj, '__xmlattrnamespaces__').items():
-                    if nameSpaceKey == nameSpacePrefix[:-1]: # Don't need to specify attribute namespace if it is the same as it selement
-                        continue
-                    if attr in nameSpaceAttributes:
-                        attrNameSpacePrefix = nameSpaceKey + ':'
-                        break
-##            if attr.startswith('_'):
-##                attr = attr[1:]
-            if (hasattr(obj, "__xmlrename__") and attr in obj.__xmlrename__):
-                attr = obj.__xmlrename__[attr]
+        return '<?xml version="1.0" encoding="%s"?>\n%s' % (sys.getdefaultencoding(), xmlstr)
 
-            objattrs += ' %s%s="%s"' % (attrNameSpacePrefix, attr, value)
+class XMLMarshalWorker(object):
+    def __init__(self, marshalType=True, prettyPrint=False, knownTypes=None, knownNamespaces=None):
+        if knownTypes == None:
+            self.knownTypes = {}
+        else:
+            self.knownTypes = knownTypes
+        if knownNamespaces == None:
+            self.knownNamespaces = {}
+        else:
+            self.knownNamespaces = knownNamespaces
+        self.prettyPrint = prettyPrint
+        self.marshalType = marshalType
+        self.xmldeepexclude = []
+        self.nsstack = []
 
-    objtype = type(obj)
-    if isinstance(obj, NoneType):
+    def getNSPrefix(self):
+        if len(self.nsstack) > 0:
+            return self.nsstack[-1].prefix
         return ''
-#        return '%s<%s objtype="None"/>%s' % (prefix, elementName, newline)
-    elif isinstance(obj, bool):
-        return '%s<%s objtype="bool">%s</%s>%s' % (prefix, elementName, obj, elementName, newline)
-    elif isinstance(obj, int):
-        return '''%s<%s objtype="int">%s</%s>%s''' % (prefix, elementName, str(obj), elementName, newline)
-    elif isinstance(obj, long):
-        return '%s<%s objtype="long">%s</%s>%s' % (prefix, elementName, str(obj), elementName, newline)
-    elif isinstance(obj, float):
-        return '%s<%s objtype="float">%s</%s>%s' % (prefix, elementName, str(obj), elementName, newline)
-    elif isinstance(obj, basestring):
-        return '''%s<%s>%s</%s>%s''' % (prefix, elementName, saxutils.escape(obj), elementName, newline)
-##    elif isinstance(obj, unicode):
-##        return '''%s<%s>%s</%s>%s''' % (prefix, elementName, obj, elementName, newline)
-    elif isinstance(obj, list):
-        if len(obj) < 1:
-            return ''
-        xmlString = '%s<%s objtype="list">%s' % (prefix, elementName, newline)
-        for item in obj:
-            xmlString += marshal(item, nameSpaces=nameSpaces, indent=indent+increment)
-        xmlString += '%s</%s>%s' % (prefix, elementName, newline)
-        return xmlString
-    elif isinstance(obj, tuple):
-        if len(obj) < 1:
-            return ''
-        xmlString = '%s<%s objtype="list" mutable="false">%s' % (prefix, elementName, newline)
-        for item in obj:
-            xmlString += marshal(item, nameSpaces=nameSpaces, indent=indent+increment)
-        xmlString += '%s</%s>%s' % (prefix, elementName, newline)
-        return xmlString
-    elif isinstance(obj, dict):
-        xmlString = '%s<%s objtype="dict">%s' % (prefix, elementName, newline)
-        subprefix = prefix + ' '*increment
-        subindent = indent + 2*increment
-        for key, val in obj.iteritems():
-            xmlString += "%s<key>%s%s%s</key>%s%s<value>%s%s%s</value>%s" \
-                         % (subprefix, newline, marshal(key, indent=subindent), subprefix, newline, subprefix, newline, marshal(val, nameSpaces=nameSpaces, indent=subindent), subprefix, newline)
-        xmlString += '%s</%s>%s' % (prefix, elementName, newline)
-        return xmlString
-    else:
-        moduleName = obj.__class__.__module__
-        if (moduleName == "activegrid.model.schema"):
-            xmlString = '%s<%s%s%s' % (prefix, elementName, nameSpaceAttrs, objattrs)
+
+    def isKnownType(self, elementName):
+        tagLongNs = None
+        nse = self.nsstack[-1]
+        i = elementName.rfind(':')
+        if i > 0:
+            prefix = elementName[:i]
+            name = elementName[i+1:]
         else:
-            xmlString = '%s<%s%s%s objtype="%s.%s"' % (prefix, elementName, nameSpaceAttrs, objattrs, moduleName, className)
-        # get the member, value pairs for the object, filtering out
-        # the types we don't support.
-        xmlMemberString = ''
-        if hasattr(obj, '__xmlbody__'):
-            xmlMemberString = getattr(obj, obj.__xmlbody__)
+            prefix = DEFAULT_NAMESPACE_KEY
+            name = elementName
+        for shortNs, longNs in nse.nameSpaces.iteritems():
+            if shortNs == prefix:
+                tagLongNs = longNs
+                break
+        if tagLongNs == None:
+            knownTagName = elementName
         else:
-            entryList = obj.__dict__.items()
-##            # Add in properties
-##            for key in obj.__class__.__dict__.iterkeys():
-##                if (key not in members_to_skip and key not in obj.__dict__
-##                    and hasPropertyValue(obj, key)):
-##                    value = getattr(obj, key)
-##                    entryList.append((key, value))
-            entryList.sort()
-            for name, value in entryList:
-##                # special name handling for private "__*" attributes:
-##                # remove the _<class-name> added by Python
-##                if name.startswith(classNamePrefix): name = name[len(classNamePrefix):]
-                if name in members_to_skip: continue
-                if name.startswith('__') and name.endswith('__'): continue
-##                idx = name.find('__')
-##                if idx > 0:
-##                    newName = name[idx+2:]
-##                    if newName:
-##                        name = newName
-                subElementNameSpacePrefix = nameSpacePrefix
-                if hasattr(obj, '__xmlattrnamespaces__'):
-                    for nameSpaceKey, nameSpaceValues in getattr(obj, '__xmlattrnamespaces__').items():
-                        if name in nameSpaceValues:
-                            subElementNameSpacePrefix = nameSpaceKey + ':'
+            knownShortNs = self.knownNamespaces[tagLongNs]
+            knownTagName = knownShortNs + ':' + name
+        if (knownTagName in asDict(self.knownTypes)):
+            knownClass = self.knownTypes[knownTagName]
+            return True
+        return False
+                      
+    def popNSStack(self):
+        self.nsstack.pop()
+
+    def appendNSStack(self, obj):
+        nameSpaces = {}
+        defaultLongNS = None
+        for nse in self.nsstack:
+            for k, v in nse.nsMap.iteritems():
+                nameSpaces[k] = v
+                if k == DEFAULT_NAMESPACE_KEY:
+                    defaultLongNS = v
+        newNS = NsElement()
+        nameSpaceAttrs = ""
+        if hasattr(obj, "__xmlnamespaces__"):
+            ns = getattr(obj, "__xmlnamespaces__")
+            keys = ns.keys()
+            keys.sort()
+            for nameSpaceKey in keys:
+                nameSpaceUrl = ns[nameSpaceKey]
+                if nameSpaceUrl in nameSpaces.values():
+                    for k, v in nameSpaces.iteritems():
+                        if v == nameSpaceUrl:
+                            nameSpaceKey = k
                             break
-                # handle sequences listed in __xmlflattensequence__
-                # specially: instead of listing the contained items inside
-                # of a separate list, as god intended, list them inside
-                # the object containing the sequence.
-                if hasattr(obj, '__xmlflattensequence__') and name in obj.__xmlflattensequence__ and value:
-                    try:
-                        xmlnametuple = obj.__xmlflattensequence__[name]
-                        xmlname = None
-                        if len(xmlnametuple) == 1:
-                            xmlname = xmlnametuple[0]
-                    except:
-                        xmlname = name
-##                        xmlname = name.lower()
-                    for seqitem in value:
-                        xmlMemberString += marshal(seqitem, xmlname, subElementNameSpacePrefix, nameSpaces=nameSpaces, indent=indent+increment)
                 else:
-                    if (hasattr(obj, "__xmlrename__") and name in obj.__xmlrename__):
-                        xmlname = obj.__xmlrename__[name]
+                    if nameSpaceKey == "":
+                        defaultLongNS = nameSpaceUrl
+                        nameSpaces[DEFAULT_NAMESPACE_KEY] = nameSpaceUrl
+                        newNS.nsMap[DEFAULT_NAMESPACE_KEY] = nameSpaceUrl
+                        nameSpaceAttrs += ' xmlns="%s" ' % (nameSpaceUrl)
+                    else:
+                        nameSpaces[nameSpaceKey] = nameSpaceUrl
+                        newNS.nsMap[nameSpaceKey] = nameSpaceUrl
+                        nameSpaceAttrs += ' xmlns:%s="%s" ' % (nameSpaceKey, nameSpaceUrl)
+            nameSpaceAttrs = nameSpaceAttrs.rstrip()
+        if len(self.nsstack) > 0:
+            newNS.prefix = self.nsstack[-1].prefix
+        else:
+            newNS.prefix = ''
+        if hasattr(obj, "__xmldefaultnamespace__"):
+            longPrefixNS = getattr(obj, "__xmldefaultnamespace__")
+            if longPrefixNS == defaultLongNS:
+                newNS.prefix = ''
+            else:
+                try:
+                    for k, v in nameSpaces.iteritems():
+                        if v == longPrefixNS:
+                            newNS.prefix = k + ':'
+                            break;
+##                    print '[appendNSStack] found longPrefixNS in nameSpaces = "%s"' % (newNS.prefix)
+                except:
+                    if (longPrefixNS in asDict(self.knownNamespaces)):
+                        newNS.prefix = self.knownNamespaces[longPrefixNS] + ':'
+                    else:
+                        raise MarshallerException('Error marshalling __xmldefaultnamespace__ ("%s") not defined in namespace stack' % (longPrefixNS))
+        if hasattr(obj, "targetNamespace"):
+            newNS.targetNS = obj.targetNamespace
+        elif len(self.nsstack) > 0:
+            newNS.targetNS = self.nsstack[-1].targetNS
+        newNS.nameSpaces = nameSpaces
+        self.nsstack.append(newNS)
+        return nameSpaceAttrs       
+
+    def contractQName(self, value, obj, attr):
+        value = langToXsdType(value)
+        i = value.rfind(':')
+        if i >= 0:
+            longNS = value[:i]
+        else:
+            # the value doesn't have a namespace and we couldn't map it to an XSD type...what to do?
+            # (a) just write it, as is, and hope it's in the default namespace (for now)
+            # (b) throw an exception so we can track down the bad code (later)
+            return value
+        if (longNS in self.nsstack[-1].nameSpaces.values()):
+            for kShort, vLong in self.nsstack[-1].nameSpaces.iteritems():
+                if vLong == longNS:
+                    shortNS = kShort
+                    break
+        else:
+            shortNS = longNS    # if we can't find the long->short mappping, just use longNS
+        if shortNS == DEFAULT_NAMESPACE_KEY:
+            value = value[i+1:]
+        else:
+            value = shortNS + ':' + value[i+1:]
+        return value
+
+    def _genObjTypeStr(self, typeString):
+        if self.marshalType:
+            return ' objtype="%s"' % typeString
+        return ""
+            
+    def _marshal(self, obj, elementName=None, nameSpacePrefix="", indent=0):
+        if (obj != None):
+            xmlMarshallerLogger.debug("--> _marshal: elementName=%s%s, type=%s, obj=%s, indent=%d" % (nameSpacePrefix, elementName, type(obj), str(obj), indent))
+        else:
+            xmlMarshallerLogger.debug("--> _marshal: elementName=%s%s, obj is None, indent=%d" % (nameSpacePrefix, elementName, indent))
+        excludeAttrs = []
+        excludeAttrs.extend(self.xmldeepexclude)
+        if hasattr(obj, "__xmlexclude__"):
+            excludeAttrs.extend(obj.__xmlexclude__)
+        prettyPrint = self.prettyPrint
+        knownTypes = self.knownTypes
+        xmlString = None
+        if self.prettyPrint or indent:
+            prefix = " "*indent
+            newline = "\n"
+            increment = 2
+        else:
+            prefix = ""
+            newline = ""
+            increment = 0
+        ## Determine the XML element name. If it isn"t specified in the
+        ## parameter list, look for it in the __xmlname__ Lang
+        ## attribute, else use the default generic BASETYPE_ELEMENT_NAME.
+        nameSpaceAttrs = self.appendNSStack(obj)
+        nameSpacePrefix = self.getNSPrefix()       
+        if not elementName:
+            if hasattr(obj, "__xmlname__"):
+                elementName = nameSpacePrefix + obj.__xmlname__
+            else:
+                elementName = nameSpacePrefix + BASETYPE_ELEMENT_NAME
+        else:
+            elementName = nameSpacePrefix + elementName
+##        print '[XMLMarshalWorker._marshal] elementName "%s"; nameSpaceAttrs is "%s"' % (elementName, nameSpaceAttrs)
+    
+        if (hasattr(obj, "__xmlsequencer__")) and (obj.__xmlsequencer__ != None):
+            if (XMLSCHEMA_XSD_URL in self.nsstack[-1].nameSpaces.values()):
+                for kShort, vLong in self.nsstack[-1].nameSpaces.iteritems():
+                    if vLong == XMLSCHEMA_XSD_URL:
+                        xsdPrefix = kShort + ':'
+                        break
+            else:
+                xsdPrefix = 'xs:'
+            elementAdd = xsdPrefix + obj.__xmlsequencer__
+        else:
+            elementAdd = None
+                   
+    ##    print "marshal: entered with elementName: ", elementName
+        members_to_skip = []
+        ## Add more members_to_skip based on ones the user has selected
+        ## via the __xmlexclude__ and __xmldeepexclude__ attributes.
+        members_to_skip.extend(excludeAttrs)
+        # Marshal the attributes that are selected to be XML attributes.
+        objattrs = ""
+        className = ag_className(obj)
+        classNamePrefix = "_" + className
+        if hasattr(obj, "__xmlattributes__"):
+            xmlattributes = obj.__xmlattributes__
+            members_to_skip.extend(xmlattributes)
+            for attr in xmlattributes:
+##                print 'Processing element "%s"; attribute "%s"' % (elementName, attr)
+                internalAttrName = attr
+                ifDefPy()
+                if (attr.startswith("__") and not attr.endswith("__")): 
+                    internalAttrName = classNamePrefix + attr
+                endIfDef()
+                # Fail silently if a python attribute is specified to be
+                # an XML attribute but is missing.
+##            print "marshal:   processing attribute ", internalAttrName
+                attrNameSpacePrefix = ""
+                if hasattr(obj, "__xmlattrnamespaces__"):
+                    for nameSpaceKey, nameSpaceAttributes in getattr(obj, "__xmlattrnamespaces__").iteritems():
+                        if nameSpaceKey == nameSpacePrefix[:-1]: # Don't need to specify attribute namespace if it is the same as its element
+                            continue
+                        if attr in nameSpaceAttributes:
+                            attrNameSpacePrefix = nameSpaceKey + ":"
+                            break
+                attrs = obj.__dict__
+                value = attrs.get(internalAttrName)
+                if (hasattr(obj, "__xmlrename__") and attr in asDict(obj.__xmlrename__)):
+                    attr = obj.__xmlrename__[attr]
+                xsdElement = None
+                complexType = getComplexType(obj)
+                if (complexType != None):
+                    xsdElement = complexType.findElement(attr)
+                if (xsdElement != None):
+                    default = xsdElement.default
+                    if (default != None):
+                        if ((default == value) or (default == _getXmlValue(value))):
+                            continue
                     else:
-                        xmlname = name
-##                    xmlname = name.lower()
-##                    # skip
-##                    if xmlname.startswith('_') and not xmlname.startswith('__'):
-##                        xmlname = xmlname[1:]
-##                    if (indent > 30):
-##                        print "getting pretty deep, xmlname = ", xmlname
-                    xmlMemberString += marshal(value, xmlname, subElementNameSpacePrefix, nameSpaces=nameSpaces, indent=indent+increment)
-        # if we have nested elements, add them here, otherwise close the element tag immediately.
-        if xmlMemberString:
-            xmlString += '>'
-            if hasattr(obj, '__xmlbody__'):
-                xmlString += xmlMemberString
-                xmlString += '</%s>%s' % (elementName, newline)
+                        if (value == None):
+                            continue
+                        elif xsdElement.type == TYPE_QNAME:
+                            value = self.contractQName(value, obj, attr)
+                elif value == None:
+                    continue
+    
+                # ToDO remove maxOccurs hack after bug 177 is fixed
+                if attr == "maxOccurs" and value == -1:
+                    value = "unbounded"
+    
+                if isinstance(value, bool):
+                   if value == True:
+                       value = "true"
+                   else:
+                       value = "false"
+                else:
+                    value = objutils.toDiffableRepr(value)
+    
+                objattrs += ' %s%s="%s"' % (attrNameSpacePrefix, attr, saxutils.escape(value))
+    ##            print "marshal:   new objattrs is: ", objattrs
+        if (obj == None):
+            xmlString = [""]
+        elif isinstance(obj, bool):
+            objTypeStr = self._genObjTypeStr("bool")
+            xmlString = ['%s<%s%s>%s</%s>%s' % (prefix, elementName, objTypeStr, obj, elementName, newline)]
+        elif isinstance(obj, int):
+            objTypeStr = self._genObjTypeStr("int")
+            xmlString = ['%s<%s%s>%s</%s>%s' % (prefix, elementName, objTypeStr, str(obj), elementName, newline)]
+        elif isinstance(obj, long):
+            objTypeStr = self._genObjTypeStr("long")
+            xmlString = ['%s<%s%s>%s</%s>%s' % (prefix, elementName, objTypeStr, str(obj), elementName, newline)]
+        elif isinstance(obj, float):
+            objTypeStr = self._genObjTypeStr("float")
+            xmlString = ['%s<%s%s>%s</%s>%s' % (prefix, elementName, objTypeStr, str(obj), elementName, newline)]
+        elif isinstance(obj, unicode): # have to check before basestring - unicode is instance of base string
+            xmlString = ['%s<%s>%s</%s>%s' % (prefix, elementName, saxutils.escape(obj.encode()), elementName, newline)]
+        elif isinstance(obj, basestring):
+            xmlString = ['%s<%s>%s</%s>%s' % (prefix, elementName, saxutils.escape(obj), elementName, newline)]
+        elif isinstance(obj, list):
+            if len(obj) < 1:
+                xmlString = ""
+            else:
+                objTypeStr = self._genObjTypeStr("list")
+                xmlString = ['%s<%s%s>%s' % (prefix, elementName, objTypeStr, newline)]
+                for item in obj:
+                    xmlString.extend(self._marshal(item, indent=indent+increment))
+                xmlString.append("%s</%s>%s" % (prefix, elementName, newline))
+        elif isinstance(obj, tuple):
+            if len(obj) < 1:
+                xmlString = ""
             else:
-                xmlString += newline
-                xmlString += xmlMemberString
-                xmlString += '%s</%s>%s' % (prefix, elementName, newline)
+                objTypeStr = self._genObjTypeStr("list")
+                xmlString = ['%s<%s%s mutable="false">%s' % (prefix, elementName, objTypeStr, newline)]
+                for item in obj:
+                    xmlString.extend(self._marshal(item, indent=indent+increment))
+                xmlString.append("%s</%s>%s" % (prefix, elementName, newline))
+        elif isinstance(obj, dict):
+            objTypeStr = self._genObjTypeStr("dict")
+            xmlString = ['%s<%s%s>%s' % (prefix, elementName, objTypeStr, newline)]
+            subprefix = prefix + " "*increment
+            subindent = indent + 2*increment
+            keys = obj.keys()
+            keys.sort()
+            for key in keys:
+                xmlString.append("%s<%s>%s" % (subprefix, DICT_ITEM_NAME, newline))
+                xmlString.extend(self._marshal(key, elementName=DICT_ITEM_KEY_NAME, indent=subindent))
+                xmlString.extend(self._marshal(obj[key], elementName=DICT_ITEM_VALUE_NAME, indent=subindent))
+                xmlString.append("%s</%s>%s" % (subprefix, DICT_ITEM_NAME, newline))
+            xmlString.append("%s</%s>%s" % (prefix, elementName, newline))
+        elif hasattr(obj, "__xmlcontent__"):
+            contentValue = getattr(obj, obj.__xmlcontent__)
+            if contentValue == None: 
+                contentValue = ''
+            else:
+                contentValue = saxutils.escape(contentValue)
+            xmlString = ["%s<%s%s%s>%s</%s>%s" % (prefix, elementName, nameSpaceAttrs, objattrs, contentValue, elementName, newline)]        
         else:
-            xmlString = xmlString + '/>%s' % newline
+            # Only add the objtype if the element tag is unknown to us.
+            if (self.isKnownType(elementName) == True): 
+                objTypeStr = ""
+            else:
+                objTypeStr = self._genObjTypeStr("%s.%s" % (obj.__class__.__module__, className))
+            xmlString = ['%s<%s%s%s%s' % (prefix, elementName, nameSpaceAttrs, objattrs, objTypeStr)]
+            # get the member, value pairs for the object, filtering out the types we don"t support
+            if (elementAdd != None):
+                prefix += increment*" "
+                indent += increment
+            xmlMemberString = []
+            if hasattr(obj, "__xmlbody__"):
+                xmlbody = getattr(obj, obj.__xmlbody__)
+                if xmlbody != None:
+                    xmlMemberString.append(xmlbody)           
+            else:
+                if hasattr(obj, "__xmlattrgroups__"):
+                    attrGroups = obj.__xmlattrgroups__.copy()
+                    if (not isinstance(attrGroups, dict)):
+                        raise "__xmlattrgroups__ is not a dict, but must be"
+                    for n in attrGroups.iterkeys():
+                        members_to_skip.extend(attrGroups[n])
+                else:
+                    attrGroups = {}
+                # add the list of all attributes to attrGroups
+                eList = obj.__dict__.keys()    
+                eList.sort()
+                attrGroups["__nogroup__"] = eList
+                
+                for eName, eList in attrGroups.iteritems():
+                    if (eName != "__nogroup__"):
+                        prefix += increment*" "
+                        indent += increment
+                        objTypeStr = self._genObjTypeStr("None")
+                        xmlMemberString.append('%s<%s%s>%s' % (prefix, eName, objTypeStr, newline))
+                    for name in eList:
+                        value = obj.__dict__[name]
+                        if eName == "__nogroup__" and name in members_to_skip: continue
+                        if name.startswith("__") and name.endswith("__"): continue
+                        if (hasattr(obj, "__xmlcdatacontent__") and (obj.__xmlcdatacontent__ == name)):
+                            continue
+                        subElementNameSpacePrefix = nameSpacePrefix
+                        if hasattr(obj, "__xmlattrnamespaces__"):
+                            for nameSpaceKey, nameSpaceValues in getattr(obj, "__xmlattrnamespaces__").iteritems():
+                                if name in nameSpaceValues:
+                                    subElementNameSpacePrefix = nameSpaceKey + ":"
+                                    break
+                        # handle sequences listed in __xmlflattensequence__
+                        # specially: instead of listing the contained items inside
+                        # of a separate list, as God intended, list them inside
+                        # the object containing the sequence.
+                        if (hasattr(obj, "__xmlflattensequence__") and (value != None) and (name in asDict(obj.__xmlflattensequence__))):
+                            xmlnametuple = obj.__xmlflattensequence__[name]
+                            if (xmlnametuple == None):
+                                xmlnametuple = [name]
+                            elif (not isinstance(xmlnametuple, (tuple,list))):
+                                xmlnametuple = [str(xmlnametuple)]
+                            xmlname = None
+                            if (len(xmlnametuple) == 1):
+                                xmlname = xmlnametuple[0]
+    ##                        ix = 0
+                            if not isinstance(value, (list, tuple)):
+                              value = [value]
+                            for seqitem in value:
+    ##                            xmlname = xmlnametuple[ix]
+    ##                            ix += 1
+    ##                            if (ix >= len(xmlnametuple)):
+    ##                                ix = 0
+                                xmlMemberString.extend(self._marshal(seqitem, xmlname, subElementNameSpacePrefix, indent=indent+increment))
+                        else:
+                            if (hasattr(obj, "__xmlrename__") and name in asDict(obj.__xmlrename__)):
+                                xmlname = obj.__xmlrename__[name]
+                            else:
+                                xmlname = name
+                            xmlMemberString.extend(self._marshal(value, xmlname, subElementNameSpacePrefix, indent=indent+increment))
+                    if (eName != "__nogroup__"):
+                        xmlMemberString.append("%s</%s>%s" % (prefix, eName, newline))
+                        prefix = prefix[:-increment]
+                        indent -= increment
+    
+            # if we have nested elements, add them here, otherwise close the element tag immediately.
+            newList = []
+            for s in xmlMemberString:
+                if (len(s) > 0): newList.append(s)
+            xmlMemberString = newList
+            if len(xmlMemberString) > 0:
+                xmlString.append(">")
+                if hasattr(obj, "__xmlbody__"):
+                    xmlString.extend(xmlMemberString)
+                    xmlString.append("</%s>%s" % (elementName, newline))
+                else:
+                    xmlString.append(newline)
+                    if (elementAdd != None):
+                        xmlString.append("%s<%s>%s" % (prefix, elementAdd, newline))
+                    xmlString.extend(xmlMemberString)
+                    if (elementAdd != None):
+                        xmlString.append("%s</%s>%s" % (prefix, elementAdd, newline))
+                        prefix = prefix[:-increment]
+                        indent -= increment
+                    xmlString.append("%s</%s>%s" % (prefix, elementName, newline))
+            else:
+                if hasattr(obj, "__xmlcdatacontent__"):
+                    cdataAttr = obj.__xmlcdatacontent__
+                    cdataContent = obj.__dict__[cdataAttr]
+                    xmlString.append("><![CDATA[%s]]></%s>%s" % (cdataContent, elementName, newline))
+                else:
+                    xmlString.append("/>%s" % newline)
+    ##        return xmlString
+        xmlMarshallerLogger.debug("<-- _marshal: %s" % str(xmlString))
+        #print "<-- _marshal: %s" % str(xmlString)
+        self.popNSStack()
         return xmlString
 
-# We don't use this anymore but in case we want to get properties this is how
-# you do it
-def hasPropertyValue(obj, attr):
-    hasProp = False
-    try:
-        prop = obj.__class__.__dict__[attr]
-        if (isinstance(prop, property)):
-            hasProp = hasattr(obj, attr)
-            if (hasProp):
-                # It's a property and it has a value but sometimes we don't want it.
-                # If there is a _hasattr method execute it and the
-                # result will tell us whether to include this value
-                try:
-                    hasProp = obj._hasattr(attr)
-                except:
-                    pass
-    except KeyError:
-        pass
-    return hasProp
-
-if __name__ == '__main__':
-    from xmlmarshallertests import Person, marshalledint, marshalledlist
-
-    l = [1, 2, 3]
-    d = {'1': 1, '2': 2}
-    outerlist = [l]
-    xmlstr = marshal(d, "d", prettyPrint=True)
-    print xmlstr
-
-    person = Person()
-    person.firstName = "Albert"
-    person.lastName = "Camus"
-    person.addressLine1 = "23 Absurd St."
-    person.city = "Ennui"
-    person.state = "MO"
-    person.zip = "54321"
-    person._phoneNumber = "808-303-2323"
-    person.favoriteWords = ['angst', 'ennui', 'existence']
-    person.weight = 150
-
-    xmlstring = marshal(person, 'person', prettyPrint=True)
-    print xmlstring
-
-    obj = unmarshal(marshalledlist)
-    print "obj has type %s and value %s" % (type(obj), str(obj))
-    for item in obj:
-        print "item: %s" % str(item)
+# A simple test, to be executed when the xmlmarshaller is run standalone
+class MarshallerPerson:
+    __xmlname__ = "person"
+    __xmlexclude__ = ["fabulousness",]
+    __xmlattributes__ = ("nonSmoker",)
+    __xmlrename__ = {"_phoneNumber": "telephone"}
+    __xmlflattensequence__ = {"favoriteWords": ("vocabulary",)}
+    __xmlattrgroups__ = {"name": ["firstName", "lastName"], "address": ["addressLine1", "city", "state", "zip"]}
+
+    def setPerson(self):
+        self.firstName = "Albert"
+        self.lastName = "Camus"
+        self.addressLine1 = "23 Absurd St."
+        self.city = "Ennui"
+        self.state = "MO"
+        self.zip = "54321"
+        self._phoneNumber = "808-303-2323"
+        self.favoriteWords = ["angst", "ennui", "existence"]
+        self.phobias = ["war", "tuberculosis", "cars"]
+        self.weight = 150
+        self.fabulousness = "tres tres"
+        self.nonSmoker = False
+
+if isMain(__name__):
+    p1 = MarshallerPerson()
+    p1.setPerson() 
+    xmlP1 = marshal(p1, prettyPrint=True, encoding="utf-8")        
+    print "\n########################"
+    print   "# testPerson test case #"
+    print   "########################"
+    print xmlP1
+    p2 = unmarshal(xmlP1)
+    xmlP2 = marshal(p2, prettyPrint=True, encoding="utf-8")
+    if xmlP1 == xmlP2:
+        print "Success: repeated marshalling yields identical results"
+    else:
+        print "Failure: repeated marshalling yields different results"
+        print xmlP2