import itertools

import xml.sax
from xml.dom import EMPTY_NAMESPACE as NULL_NAMESPACE
from xml.dom import EMPTY_PREFIX as NULL_PREFIX
from xml.dom import XML_NAMESPACE
from xml.dom import Node
from Ft.Xml import Domlette, Sax, CreateInputSource

START_DOCUMENT               = 1
END_DOCUMENT                 = 2
START_ELEMENT                = 3
END_ELEMENT                  = 4
CHARACTER_DATA               = 10
COMMENT                      = 11
PI                           = 12

#
# namespace_mixin is a utility that helps manage namespace prefix mappings
#

class namespace_mixin:
    def __init__(self):
        self._ns_prefix = {XML_NAMESPACE: [u'xml'], NULL_NAMESPACE: [NULL_PREFIX]}
        self._prefix_ns = {u'xml': [XML_NAMESPACE], NULL_PREFIX: [NULL_NAMESPACE]}
        return
    
    def startPrefixMapping(self, prefix, uri):
        self._ns_prefix.setdefault(uri, []).append(prefix)
        self._prefix_ns.setdefault(prefix, []).append(uri)
        return

    def endPrefixMapping(self, prefix):
        uri = self._prefix_ns[prefix].pop()
        prefix = self._ns_prefix[uri].pop()
        #assert prefix == uri
        return

    def name_to_qname(self, name):
        #print self._ns_prefix
        #print self._prefix_ns
        uri, local = name
        prefix = self._ns_prefix[uri][-1]
        qname = ( prefix and ( prefix + u':' ) or '') + local
        return qname


def sniff_namespace(source):
    class report_nss:
        def __init__(self, parser):
            self.nss = {}
            self.parser = parser
            return

        def startPrefixMapping(self, prefix, uri):
            self.nss[prefix] = uri
            return
        
        def startElementNS(self, name, qname, attribs):
            self.parser.setProperty(Sax.PROPERTY_YIELD_RESULT, self.nss)
            return

    parser = Sax.CreateParser()

    parser.setFeature(xml.sax.handler.feature_external_pes, False)
    parser.setFeature(Sax.FEATURE_GENERATOR, True)
    handler = report_nss(parser)
    parser.setContentHandler(handler)
    elem_iter = parser.parse(CreateInputSource(source))
    nss = elem_iter.next()
    #for attribs in attribs_iterator:
    #     for name in attribs.keys(): print name
    return nss


#
# Tenorsax framework: helps linerarize SAX logic
#

from xml import sax  #Python 2.3 or higher required
class tenorsax(namespace_mixin, sax.ContentHandler):
    def __init__(self, consumer):
        namespace_mixin.__init__(self)
        self.consumer = consumer
        self.dispatcher = consumer.top_dispatcher
        self.curr_gen = None
        return
    
    def startElementNS(self, name, qname, attributes):
        (ns, local) = name
        qname = self.name_to_qname(name)
        #print "Start element", (name, qname)
        self.consumer.event = (START_ELEMENT, ns, local)
        self.consumer.params = attributes
        self.curr_gen = tenorsax.event_loop_body(self.dispatcher, self.curr_gen, self.consumer.event)
        return
    
    def endElementNS(self, name, qname):
        (ns, local) = name
        qname = self.name_to_qname(name)
        #print "end element", (name, qname)
        self.consumer.event = (END_ELEMENT, ns, local)
        self.consumer.params = None
        self.curr_gen = tenorsax.event_loop_body(self.dispatcher, self.curr_gen, self.consumer.event)
        return
    
    def characters(self, text):
        #print "characters", text
        self.consumer.event = (CHARACTER_DATA,)
        self.consumer.params = text
        self.curr_gen = tenorsax.event_loop_body(self.dispatcher, self.curr_gen, self.consumer.event)
        return

    def event_loop_body(dispatcher, curr_gen, event):
        if curr_gen:
            curr_gen = tenorsax.execute_delegate(curr_gen)
        else:
            curr_gen = tenorsax.check_for_delegate(dispatcher, event)
        return curr_gen
    event_loop_body = staticmethod(event_loop_body)

    def execute_delegate(curr_gen):
        try:
            curr_gen.next()
        except StopIteration:
            curr_gen = None
        return curr_gen
    execute_delegate = staticmethod(execute_delegate)

    def check_for_delegate(dispatcher, event):
        if event[0] == START_ELEMENT:
            end_condition = (END_ELEMENT,) + event[1:]
        else:
            end_condition = None
        curr_gen = None
        delegate_generator = dispatcher.get(event)
        if delegate_generator:
            #Fire up the generator
            curr_gen = delegate_generator(end_condition)
            try:
                curr_gen.next()
            except StopIteration:
                print "immediate end"
                #Immediate generator termination
                curr_gen = None
        return curr_gen
    check_for_delegate = staticmethod(check_for_delegate)

#
#
#

from xml import sax
from xml.dom import XML_NAMESPACE, XMLNS_NAMESPACE
from xml.dom import EMPTY_NAMESPACE as NULL_NAMESPACE
from xml.dom import EMPTY_PREFIX as NULL_PREFIX

from Ft.Xml.Xslt import parser as XPatternParser

from Ft.Xml.Xslt.XPatterns import Patterns
from Ft.Xml.Xslt.XPatterns import Pattern
from Ft.Xml.Xslt.XPatterns import DocumentNodeTest
from Ft.Xml.XPath.ParsedNodeTest import LocalNameTest
from Ft.Xml.XPath.ParsedNodeTest import NamespaceTest
from Ft.Xml.XPath.ParsedNodeTest import QualifiedNameTest
from Ft.Xml.XPath.ParsedNodeTest import PrincipalTypeTest
from Ft.Xml.XPath import Context, Compile

DUMMY_DOCELEM = u'dummy'
START_STATE = 0
#A special state that means
#"immediately pop off the state stack and transit to the top state"
POP_STATE = 9999
TOP = -1
ANY = '?'
ATTRIB = '@'
#Used to figure out whether a wildcard event is user-specified,
#Or added internally
EXPLICIT, IMPLICIT = (True, False)


class xpattern_state_machine:
    """
    A simple state machine that interprets XPatterns
    A state is "live" when it represents the successful completion
    of an XPattern.
    """
    PARSER = XPatternParser.new()
    
    def __init__(self, repr_xp, xp, nss):
        self._state_table = {START_STATE: {}}
        self._live_states = {}
        self._attrib_tests = {}
        self._ignored_subtree_states = []
        self._push_states = []
        self._substate_depth = 0
        #States that should be hooked into for matching later XPatterns
        self._hook_states = {}
        newest_state = START_STATE
        last_state = START_STATE
        for subpat in xp.patterns:
            steps = subpat.steps[:]
            steps.reverse()
            for (step_count, (axis_type, node_test, ancestor)) in enumerate(steps):
                #Note: XSLT patterns only allow child or attribute axis
                handled = False
                attrib_test = None
                if axis_type == Node.ATTRIBUTE_NODE:
                    if (isinstance(node_test, LocalNameTest)
                        or isinstance(node_test, QualifiedNameTest)
                        or isinstance(node_test, NamespaceTest)
                        or isinstance(node_test, PrincipalTypeTest)
                        ):
                        attrib_test = node_test
                        start_event = (1, ATTRIB, None)
                        end_event = (0, ATTRIB, None)
                        handled = True
                elif isinstance(node_test, DocumentNodeTest):
                    start_event = (1, None, None)
                    end_event = (0, None, None)
                    handled = True
                elif isinstance(node_test, LocalNameTest):
                    if node_test.nodeType == Node.ELEMENT_NODE:
                        start_event = (1, None, node_test._name)
                        end_event = (0, None, node_test._name)
                        handled = True
                elif isinstance(node_test, QualifiedNameTest):
                    if node_test.nodeType == Node.ELEMENT_NODE:
                        ns = nss[node_test._prefix]
                        start_event = (1, ns, node_test._localName)
                        end_event = (0, ns, node_test._localName)
                        handled = True
                elif isinstance(node_test, PrincipalTypeTest):
                    if node_test.nodeType == Node.ELEMENT_NODE:
                        start_event = (1, ANY, EXPLICIT)
                        end_event = (0, ANY, EXPLICIT)
                        handled = True
                elif isinstance(node_test, NamespaceTest):
                    if node_test.nodeType == Node.ELEMENT_NODE:
                        ns = nss[node_test._prefix]
                        start_event = (1, ns, ANY)
                        end_event = (0, ns, ANY)
                        handled = True
                if not(handled):
                    import sys; print >> sys.stderr, "Pattern step not supported:", (axis_type, node_test, ancestor), "Node test class", node_test.__class__
                    continue

                #Say the full input is /u/v/w|/a/b/c and we're currently
                #Working the sub-pattern /a/b/c
                #top_state is the pattern 
                #last_state is the last

                last_state = newest_state
                newest_state += 1
                if not(step_count):
                    if attrib_test:
                    #    #Because we want to treat @X as */@X
                    #    start_event = (1, ANY, EXPLICIT)
                    #    end_event = (0, ANY, EXPLICIT)
                        attribute_test_state = newest_state
                    self._state_table[newest_state] = {end_event: POP_STATE}
                    if isinstance(node_test, DocumentNodeTest):
                        self._state_table[START_STATE][start_event] = newest_state
                    else:
                        for state in self._state_table:
                            self._state_table[state][start_event] = newest_state
                    self._hook_states[newest_state] = start_event
                    self._push_states.append(newest_state)
                    #if attrib_test:
                        #Because we want to treat @X as */@X
                    #    newest_state += 1
                    #    start_event = (1, ATTRIB, None)
                    #    end_event = (0, ATTRIB, None)
                    #    self._state_table[newest_state] = {end_event: newest_state - 1}
                    #    self._state_table[newest_state -1][start_event] = newest_state
                else:
                    if attrib_test:
                        attribute_test_state = newest_state
                    self._state_table[newest_state] = {end_event: parent_start_element_state}
                    self._state_table[parent_start_element_state][start_event] = newest_state
                for state in self._hook_states:
                    self._state_table[newest_state][self._hook_states[state]] = state
                
                start_element_state = newest_state
                #complete_state = top_state #The state representing completion of an XPattern
                if step_count and not ancestor and not isinstance(node_test, PrincipalTypeTest):
                    #Insert a state, which handles any child element
                    #Not explicitly matching some other state (so that
                    #/a/b/c is not a mistaken match for XPattern /a/c)
                    start_event = (1, ANY, IMPLICIT)
                    end_event = (0, ANY, IMPLICIT)
                    newest_state += 1
                    self._state_table[newest_state] = {}
                    self._state_table[parent_start_element_state][start_event] = newest_state
                    self._state_table[newest_state][end_event] = parent_start_element_state
                    self._ignored_subtree_states.append(newest_state)
                    #self._hook_states[newest_state] = start_event
                    for state in self._hook_states:
                        self._state_table[newest_state][self._hook_states[state]] = state
                parent_start_element_state = start_element_state
            self._live_states[start_element_state] = repr_xp
            if attrib_test:
                self._attrib_tests[attribute_test_state] = Compile('@'+repr(node_test))
        self._state = START_STATE
        self.entering_xpatterns = []
        self.leaving_xpatterns = []
        self.current_xpatterns = []
        self.tree_depth = 0
        self.depth_marks = []
        self.state_stack = []
        #print self._state_table; print self._live_states; print self._push_states; print self._attrib_tests
        return

    def event(self, is_start, ns, local):
        """
        Register an event and effect any state transitions
        found in the state table
        """
        #We only have a chunk ready for the handler in
        #the explicit case below
        self.entering_xpatterns = []
        self.leaving_xpatterns = []
        self.attribute_test = None
        self.tree_depth += is_start and 1 or -1
        #print "event", (is_start, ns, local), self._state, self.state_stack, self.tree_depth, self.depth_marks
        #An end event is never significant unless we know we're expecting it
        if not is_start and self.depth_marks and self.tree_depth != self.depth_marks[-1]:
            return self._state
        lookup_from = self._state_table[self._state]
        #FIXME: second part should be an element node test "*", should not match, say, start document
        if not lookup_from.has_key((is_start, ns, local)) and (ns, local) == (None, None):
            return self._state

        def process_transition(new_state):
            if new_state != self._state:
                if is_start and new_state in self._push_states:
                    self.state_stack.append(self._state)
                elif new_state == POP_STATE:
                    new_state = self.state_stack.pop()
                self._state = new_state
                if is_start:
                    self.depth_marks.append(self.tree_depth - 1)
                elif self.depth_marks:
                    self.depth_marks.pop()
            if is_start:
                attrib_state = self._state_table[new_state].get((1, ATTRIB, None))
                if attrib_state is not None:
                    self.attribute_test = self._attrib_tests[attrib_state]

        if ((is_start, ns, local) in lookup_from) or ((is_start, ns, ANY) in lookup_from):
            try:
                new_state = lookup_from[(is_start, ns, local)]
            except KeyError:
                new_state = lookup_from[(is_start, ns, ANY)]
            if (new_state in self._live_states):
                #Entering a defined XPattern chunk
                self.entering_xpatterns.append(self._live_states[new_state])
                self.current_xpatterns.append(self._live_states[new_state])
            elif (self._state in self._live_states):
                #Leaving a defined XPattern chunk
                self.leaving_xpatterns.append(self.current_xpatterns.pop())
            process_transition(new_state)
        elif (is_start, ANY, EXPLICIT) in lookup_from:
            new_state = lookup_from[(is_start, ANY, EXPLICIT)]
            if (new_state in self._live_states):
                #Entering a defined XPattern chunk
                self.entering_xpatterns.append(self._live_states[new_state])
                self.current_xpatterns.append(self._live_states[new_state])
            elif (self._state in self._live_states):
                #Leaving a defined XPattern chunk
                self.leaving_xpatterns.append(self.current_xpatterns.pop())
            process_transition(new_state)
        elif (is_start, ANY, IMPLICIT) in lookup_from:
            new_state = lookup_from[(is_start, ANY, IMPLICIT)]
            process_transition(new_state)
        else:
            #Identity transition: from a state back to itself
            process_transition(self._state)
        #print self.entering_xpatterns,self.leaving_xpatterns,self.current_xpatterns
        return self._state

    def status(self):
        """
        1 if currently within an XPattern, 0 if not
        Calling code might also want to just check
        self.current_xpatterns directly
        """
        return not not self.current_xpatterns


class xpattern_state_manager:
    """
    And aggregation of multiple state machines, one for each registered pattern
    """
    PARSER = XPatternParser.new()
    
    def __init__(self, xpatterns, nss):
        if not hasattr(xpatterns[0], "match"):
            self._xpatterns = [ (p, self.PARSER.parse(p)) for p in xpatterns ]
        else:
            self._xpatterns = [ (repr(xp), self.PARSER.parse(p)) for p in xpatterns ]
        self._machines = [ xpattern_state_machine(repr_xp, xp, nss) for repr_xp, xp in self._xpatterns ]
        return

    def event(self, is_start, ns, local):
        for machine in self._machines:
            machine.event(is_start, ns, local)
        #FIXME: Slow and clumsy
        self.entering_xpatterns = list(itertools.chain(*[m.entering_xpatterns for m in self._machines]))
        self.leaving_xpatterns = list(itertools.chain(*[m.leaving_xpatterns for m in self._machines]))
        self.current_xpatterns = list(itertools.chain(*[m.current_xpatterns for m in self._machines]))
        self.attribute_tests = [m.attribute_test for m in self._machines if m.attribute_test]
        #print "manager event", (self.entering_xpatterns, self.leaving_xpatterns, self.current_xpatterns)
        return

    def status(self):
        """
        1 if currently within an XPattern, 0 if not
        Calling code might also want to just check
        self.current_xpatterns directly
        """
        return not not self.current_xpatterns


class sax2dom_chunker(sax.ContentHandler):
    """
    Note: Ignores nodes prior to the document element, such as PIs and
    text nodes.  Collapses CDATA sections into plain text
    Only designed to work if you set the feature
      sax.handler.feature_namespaces
    to 1 on the parser you use.

    xpatterns - list of XPatterns.  Only portions of the
        tree within these patterns will be instantiated as DOM (as
        chunks fed to chunk_consumer in sequence)
        If None (the default, a DOM node will be created representing
        the entire tree.

    nss - a dictionary of prefix -> namespace name mappings used to
        interpret XPatterns
    
    chunk_consumer - a callable object taking a DOM node.  It will be
        invoked as each DOM chunk is prepared.
    
    domimpl - DOM implemention to build, e.g. mindom (the default)
        or cDomlette or pxdom (if you have the right third-party
        packages installed).
    
    owner_doc - for advanced uses, if you want to use an existing
        DOM document object as the owner of all created nodes.
        
    """
    def __init__(self,
                 xpatterns=None,
                 nss=None,
                 chunk_consumer=None,
                 domimpl=Domlette.implementation,
                 owner_doc=None,
                 ):
        nss = nss or {}
        #HINT: To use minidom
        #domimpl = xml.dom.minidom.getDOMImplementation()
        self._impl = domimpl
        if isinstance(xpatterns, str) or isinstance(xpatterns, unicode) :
            xpatterns = [xpatterns]
        #print xpatterns
        if owner_doc:
            self._owner_doc = owner_doc
        else:
            try:
                dt = self._impl.createDocumentType(DUMMY_DOCELEM, None, u'')
            except AttributeError:
                #Domlette doesn't need createDocumentType
                dt = None
            self._owner_doc = self._impl.createDocument(
                DUMMY_DOCELEM, DUMMY_DOCELEM, dt)
        #Create a docfrag to hold all the generated nodes.
        root_node = self._owner_doc.createDocumentFragment()
        self._nodeStack = [ root_node ]
        self.state_machine = xpattern_state_manager(xpatterns, nss)
        self._chunk_consumer = chunk_consumer
        return

    def get_root_node(self):
        """
        Only useful if the user does not register trim paths
        If so, then after SAX processing the user can call this
        method to retrieve resulting DOM representing the entire
        document
        """
        return self._nodeStack[0]

    #Overridden ContentHandler methods
    def startDocument(self):
        self.state_machine.event(1, None, None)
        return

    def endDocument(self):
        self.state_machine.event(0, None, None)
        return

    def startElementNS(self, (ns, local), qname, attribs):
        self.state_machine.event(1, ns, local)
        if not self.state_machine.status():
            for attrtest in self.state_machine.attribute_tests:
                e = self._owner_doc.createElementNS(ns, qname or local)
                for (ns, local), value in attribs.items():
                    e.setAttributeNS(ns, attribs.getQNameByName((ns, local)), value)
                #e = dummy_element(qname, attribs, self._nodeStack[0])
                matched = attrtest.evaluate(Context.Context(e))
                if matched:
                    #Feed the consumer
                    #self._nodeStack[TOP].appendChild(matched[0]) #"xml.dom.HierarchyRequestErr: Ft.Xml.cDomlette.Attr nodes cannot be a child of Ft.Xml.cDomlette.DocumentFragment nodes"
                    self._chunk_consumer(matched[0])
                    #Start over with new doc frag so old memory can be reclaimed
                    root_node = self._owner_doc.createDocumentFragment()
                    self._nodeStack = [ root_node ]
            return
            
        new_element = self._owner_doc.createElementNS(ns, qname or local)

        for (attr_ns, lname) in attribs:
            value = attribs[(attr_ns, lname)]
            if attr_ns is not None:
                attr_qname = attribs.getQNameByName((attr_ns, lname))
            else:
                attr_qname = lname
            attr = self._owner_doc.createAttributeNS(
                attr_ns, attr_qname)
            attr_qname = attribs.getQNameByName((attr_ns, lname))
            attr.value = value
            new_element.setAttributeNodeNS(attr)

        self._nodeStack.append(new_element)
        return

    def endElementNS(self, (ns, local), qname):
        self.state_machine.event(0, ns, local)
        if not self.state_machine.status():
            if (self._chunk_consumer and
                self.state_machine.leaving_xpatterns):
                #Complete the element being closed because it
                #Is the last bit of a DOM to be fed to the consumer
                new_element = self._nodeStack[TOP]
                del self._nodeStack[TOP]
                self._nodeStack[TOP].appendChild(new_element)
                #Feed the consumer
                self._chunk_consumer(self._nodeStack[0])
                #Start all over with a new doc frag so the old
                #One's memory can be reclaimed
                root_node = self._owner_doc.createDocumentFragment()
                self._nodeStack = [ root_node ]
            return
        new_element = self._nodeStack[TOP]
        del self._nodeStack[TOP]
        self._nodeStack[TOP].appendChild(new_element)
        return

    def processingInstruction(self, target, data):
        if self.state_machine.status():
            pi = self._owner_doc.createProcessingInstruction(
                target, data)
            self._nodeStack[TOP].appendChild(pi)
        return

    def characters(self, chars):
        if self.state_machine.status():
            new_text = self._owner_doc.createTextNode(chars)
            self._nodeStack[TOP].appendChild(new_text)
        return

    #Overridden LexicalHandler methods
    def comment(self, text):
        if self.state_machine.status():
            new_comment = self._owner_doc.createComment(text)
            self._nodeStack[TOP].appendChild(new_comment)
        return


from xml.sax.saxutils import XMLFilterBase
#FIXME: Set up to use actual PyXML if available
from amara.pyxml_standins import *

class normalize_text_filter(XMLFilterBase, LexicalHandler):
    """
    SAX filter to ensure that contiguous white space nodes are
    delivered merged into a single node
    """
    def __init__(self, *args):
        XMLFilterBase.__init__(self, *args)
        self._accumulator = []
        return

    def _complete_text_node(self):
        if self._accumulator:
            XMLFilterBase.characters(self, ''.join(self._accumulator))
            self._accumulator = []
        return

    def startDocument(self):
        XMLFilterBase.startDocument(self)
        return

    def endDocument(self):
        XMLFilterBase.endDocument(self)
        return

    def startElement(self, name, attrs):
        self._complete_text_node()
        XMLFilterBase.startElement(self, name, attrs)
        return

    def startElementNS(self, name, qname, attrs):
        self._complete_text_node()
        #A bug in Python 2.3 means that we can't just defer to parent, which is broken
        #XMLFilterBase.startElementNS(self, name, qname, attrs)
        self._cont_handler.startElementNS(name, qname, attrs)
        return

    def endElement(self, name):
        self._complete_text_node()
        XMLFilterBase.endElement(self, name)
        return

    def endElementNS(self, name, qname):
        self._complete_text_node()
        XMLFilterBase.endElementNS(self, name, qname)
        return

    def processingInstruction(self, target, body):
        self._complete_text_node()
        XMLFilterBase.processingInstruction(self, target, body)
        return

    def comment(self, body):
        self._complete_text_node()
        #No such thing as an XMLFilterBase.comment :-(
        #XMLFilterBase.comment(self, body)
        self._cont_handler.comment(body)
        return

    def characters(self, text):
        self._accumulator.append(text)
        return

    def ignorableWhitespace(self, ws):
        self._accumulator.append(text)
        return

    #Must be overridden because of a bug in Python 2.0 through 2.4
    #And even still in PyXML 0.8.4.  Missing "return"
    def resolveEntity(self, publicId, systemId):
        return self._ent_handler.resolveEntity(publicId, systemId)

    # Enhancement suggested by James Kew:
    # Override XMLFilterBase.parse to connect the LexicalHandler
    # Can only do this by setting the relevant property
    # May throw SAXNotSupportedException
    def parse(self, source):
        #import inspect; import pprint; pprint.pprint(inspect.stack())
        self._parent.setProperty(property_lexical_handler, self)
        # Delegate to XMLFilterBase for the rest
        XMLFilterBase.parse(self, source)
        return


#
# From xml.dom
#

#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