"""
Turtle RDF graph serializer for RDFLib.
See <http://www.w3.org/TeamSubmission/turtle/> for syntax specification.
"""

from collections import defaultdict

from rdflib.compat import cmp_to_key
from rdflib.term import BNode, Literal, URIRef
from rdflib.exceptions import Error
from rdflib.serializer import Serializer
from rdflib.namespace import RDF, RDFS

__all__ = ['RecursiveSerializer', 'TurtleSerializer']

def _object_comparator(a,b):
    """
    for nice clean output we sort the objects of triples,
    some of them are literals,
    these are sorted according to the sort order of the underlying python objects
    in py3 not all things are comparable.
    This falls back on comparing string representations when not.
    """

    try:
        if a>b: return 1
        if a<b: return -1
        return 0

    except TypeError:
        a = str(a)
        b = str(b)
        return (a > b) - (a < b)


class RecursiveSerializer(Serializer):

    topClasses = [RDFS.Class]
    predicateOrder = [RDF.type, RDFS.label]
    maxDepth = 10
    indentString = "  "

    def __init__(self, store):

        super(RecursiveSerializer, self).__init__(store)
        self.stream = None
        self.reset()

    def addNamespace(self, prefix, uri):
        if prefix in self.namespaces and self.namespaces[prefix]!=uri:
            raise Exception("Trying to override namespace prefix %s => %s, but it's already bound to %s"%(prefix, uri, self.namespaces[prefix]))
        self.namespaces[prefix] = uri

    def checkSubject(self, subject):
        """Check to see if the subject should be serialized yet"""
        if ((self.isDone(subject))
            or (subject not in self._subjects)
            or ((subject in self._topLevels) and (self.depth > 1))
            or (isinstance(subject, URIRef)
                and (self.depth >= self.maxDepth))):
            return False
        return True

    def isDone(self, subject):
        """Return true if subject is serialized"""
        return subject in self._serialized

    def orderSubjects(self):
        seen = {}
        subjects = []

        for classURI in self.topClasses:
            members = list(self.store.subjects(RDF.type, classURI))
            members.sort()

            for member in members:
                subjects.append(member)
                self._topLevels[member] = True
                seen[member] = True

        recursable = [
            (isinstance(subject, BNode),
             self._references[subject], subject)
            for subject in self._subjects if subject not in seen]

        recursable.sort()
        subjects.extend([subject for (isbnode, refs, subject) in recursable])

        return subjects

    def preprocess(self):
        for triple in self.store.triples((None, None, None)):
            self.preprocessTriple(triple)

    def preprocessTriple(self, xxx_todo_changeme):
        (s, p, o) = xxx_todo_changeme
        self._references[o]+=1
        self._subjects[s] = True

    def reset(self):
        self.depth = 0
        self.lists = {}
        self.namespaces = {}
        self._references = defaultdict(int)
        self._serialized = {}
        self._subjects = {}
        self._topLevels = {}

        for prefix, ns in self.store.namespaces():
            self.addNamespace(prefix, ns)

    def buildPredicateHash(self, subject):
        """
        Build a hash key by predicate to a list of objects for the given
        subject
        """
        properties = {}
        for s, p, o in self.store.triples((subject, None, None)):
            oList = properties.get(p, [])
            oList.append(o)
            properties[p] = oList
        return properties

    def sortProperties(self, properties):
        """Take a hash from predicate uris to lists of values.
           Sort the lists of values.  Return a sorted list of properties."""
        # Sort object lists
        for prop, objects in list(properties.items()):
            objects.sort(key=cmp_to_key(_object_comparator))

        # Make sorted list of properties
        propList = []
        seen = {}
        for prop in self.predicateOrder:
            if (prop in properties) and (prop not in seen):
                propList.append(prop)
                seen[prop] = True
        props = list(properties.keys())
        props.sort()
        for prop in props:
            if prop not in seen:
                propList.append(prop)
                seen[prop] = True
        return propList

    def subjectDone(self, subject):
        """Mark a subject as done."""
        self._serialized[subject] = True

    def indent(self, modifier=0):
        """Returns indent string multiplied by the depth"""
        return (self.depth + modifier) * self.indentString

    def write(self, text):
        """Write text in given encoding."""
        self.stream.write(text.encode(self.encoding, 'replace'))


SUBJECT = 0
VERB = 1
OBJECT = 2

_GEN_QNAME_FOR_DT = False
_SPACIOUS_OUTPUT = False


class TurtleSerializer(RecursiveSerializer):

    short_name = "turtle"
    indentString = '    '

    def __init__(self, store):
        self._ns_rewrite = {}
        super(TurtleSerializer, self).__init__(store)
        self.keywords = {
            RDF.type: 'a'
        }
        self.reset()
        self.stream = None
        self._spacious = _SPACIOUS_OUTPUT

    def addNamespace(self, prefix, namespace):
        # Turtle does not support prefix that start with _
        # if they occur in the graph, rewrite to p_blah
        # this is more complicated since we need to make sure p_blah
        # does not already exist. And we register namespaces as we go, i.e.
        # we may first see a triple with prefix _9 - rewrite it to p_9
        # and then later find a triple with a "real" p_9 prefix

        # so we need to keep track of ns rewrites we made so far.

        if (prefix > '' and prefix[0] == '_') \
                or self.namespaces.get(prefix, namespace) != namespace:

            if prefix not in self._ns_rewrite:
                p = "p" + prefix
                while p in self.namespaces:
                    p = "p" + p
                self._ns_rewrite[prefix] = p

            prefix = self._ns_rewrite.get(prefix, prefix)

        super(TurtleSerializer, self).addNamespace(prefix, namespace)
        return prefix

    def reset(self):
        super(TurtleSerializer, self).reset()
        self._shortNames = {}
        self._started = False
        self._ns_rewrite = {}

    def serialize(self, stream, base=None, encoding=None,
                  spacious=None, **args):
        self.reset()
        self.stream = stream
        self.base = base

        if spacious is not None:
            self._spacious = spacious

        self.preprocess()
        subjects_list = self.orderSubjects()

        self.startDocument()

        firstTime = True
        for subject in subjects_list:
            if self.isDone(subject):
                continue
            if firstTime:
                firstTime = False
            if self.statement(subject) and not firstTime:
                self.write('\n')

        self.endDocument()
        stream.write("\n".encode('ascii'))

    def preprocessTriple(self, triple):
        super(TurtleSerializer, self).preprocessTriple(triple)
        for i, node in enumerate(triple):
            if node in self.keywords:
                continue
            # Don't use generated prefixes for subjects and objects
            self.getQName(node, gen_prefix=(i == VERB))
            if isinstance(node, Literal) and node.datatype:
                self.getQName(node.datatype, gen_prefix=_GEN_QNAME_FOR_DT)
        p = triple[1]
        if isinstance(p, BNode): # hmm - when is P ever a bnode?
            self._references[p]+=1

    def getQName(self, uri, gen_prefix=True):
        if not isinstance(uri, URIRef):
            return None

        parts = None

        try:
            parts = self.store.compute_qname(uri, generate=gen_prefix)
        except:

            # is the uri a namespace in itself?
            pfx = self.store.store.prefix(uri)

            if pfx is not None:
                parts = (pfx, uri, '')
            else:
                # nothing worked
                return None

        prefix, namespace, local = parts

        # QName cannot end with .
        if local.endswith("."): return None

        prefix = self.addNamespace(prefix, namespace)

        return '%s:%s' % (prefix, local)

    def startDocument(self):
        self._started = True
        ns_list = sorted(self.namespaces.items())
        for prefix, uri in ns_list:
            self.write(self.indent() + '@prefix %s: <%s> .\n' % (prefix, uri))
        if ns_list and self._spacious:
            self.write('\n')

    def endDocument(self):
        if self._spacious:
            self.write('\n')

    def statement(self, subject):
        self.subjectDone(subject)
        return self.s_squared(subject) or self.s_default(subject)

    def s_default(self, subject):
        self.write('\n' + self.indent())
        self.path(subject, SUBJECT)
        self.predicateList(subject)
        self.write(' .')
        return True

    def s_squared(self, subject):
        if (self._references[subject] > 0) or not isinstance(subject, BNode):
            return False
        self.write('\n' + self.indent() + '[]')
        self.predicateList(subject)
        self.write(' .')
        return True

    def path(self, node, position, newline=False):
        if not (self.p_squared(node, position, newline)
                or self.p_default(node, position, newline)):
            raise Error("Cannot serialize node '%s'" % (node, ))

    def p_default(self, node, position, newline=False):
        if position != SUBJECT and not newline:
            self.write(' ')
        self.write(self.label(node, position))
        return True

    def label(self, node, position):
        if node == RDF.nil:
            return '()'
        if position is VERB and node in self.keywords:
            return self.keywords[node]
        if isinstance(node, Literal):
            return node._literal_n3(
                use_plain=True,
                qname_callback=lambda dt: self.getQName(
                    dt, _GEN_QNAME_FOR_DT))
        else:
            node = self.relativize(node)

            return self.getQName(node, position == VERB) or node.n3()

    def p_squared(self, node, position, newline=False):
        if (not isinstance(node, BNode)
                or node in self._serialized
                or self._references[node] > 1
                or position == SUBJECT):
            return False

        if not newline:
            self.write(' ')

        if self.isValidList(node):
            # this is a list
            self.write('(')
            self.depth += 1  # 2
            self.doList(node)
            self.depth -= 1  # 2
            self.write(' )')
        else:
            self.subjectDone(node)
            self.depth += 2
            # self.write('[\n' + self.indent())
            self.write('[')
            self.depth -= 1
            # self.predicateList(node, newline=True)
            self.predicateList(node, newline=False)
            # self.write('\n' + self.indent() + ']')
            self.write(' ]')
            self.depth -= 1

        return True

    def isValidList(self, l):
        """
        Checks if l is a valid RDF list, i.e. no nodes have other properties.
        """
        try:
            if self.store.value(l, RDF.first) is None:
                return False
        except:
            return False
        while l:
            if l != RDF.nil and len(
                    list(self.store.predicate_objects(l))) != 2:
                return False
            l = self.store.value(l, RDF.rest)
        return True

    def doList(self, l):
        while l:
            item = self.store.value(l, RDF.first)
            if item is not None:
                self.path(item, OBJECT)
                self.subjectDone(l)
            l = self.store.value(l, RDF.rest)

    def predicateList(self, subject, newline=False):
        properties = self.buildPredicateHash(subject)
        propList = self.sortProperties(properties)
        if len(propList) == 0:
            return
        self.verb(propList[0], newline=newline)
        self.objectList(properties[propList[0]])
        for predicate in propList[1:]:
            self.write(' ;\n' + self.indent(1))
            self.verb(predicate, newline=True)
            self.objectList(properties[predicate])

    def verb(self, node, newline=False):
        self.path(node, VERB, newline)

    def objectList(self, objects):
        count = len(objects)
        if count == 0:
            return
        depthmod = (count == 1) and 0 or 1
        self.depth += depthmod
        self.path(objects[0], OBJECT)
        for obj in objects[1:]:
            self.write(',\n' + self.indent(1))
            self.path(obj, OBJECT, newline=True)
        self.depth -= depthmod