# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
# 
# http://www.apache.org/licenses/LICENSE-2.0
# 
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
import calendar
import datetime
import json
import uuid
import math
from collections import OrderedDict
from decimal import *
import logging
from datetime import timedelta

import six
from aenum import Enum
from isodate import parse_duration, duration_isoformat

from gremlin_python import statics
from gremlin_python.statics import FloatType, FunctionType, IntType, LongType, TypeType, DictType, ListType, SetType, SingleByte, ByteBufferType, SingleChar
from gremlin_python.process.traversal import Binding, Bytecode, Direction, P, TextP, Traversal, Traverser, TraversalStrategy, T
from gremlin_python.structure.graph import Edge, Property, Vertex, VertexProperty, Path

log = logging.getLogger(__name__)

# When we fall back to a superclass's serializer, we iterate over this map.
# We want that iteration order to be consistent, so we use an OrderedDict,
# not a dict.
_serializers = OrderedDict()
_deserializers = {}


class GraphSONTypeType(type):
    def __new__(mcs, name, bases, dct):
        cls = super(GraphSONTypeType, mcs).__new__(mcs, name, bases, dct)
        if not name.startswith('_'):
            if cls.python_type:
                _serializers[cls.python_type] = cls
            if cls.graphson_type:
                _deserializers[cls.graphson_type] = cls
        return cls


class GraphSONUtil(object):
    TYPE_KEY = "@type"
    VALUE_KEY = "@value"

    @classmethod
    def typedValue(cls, type_name, value, prefix="g"):
        out = {cls.TYPE_KEY: cls.formatType(prefix, type_name)}
        if value is not None:
            out[cls.VALUE_KEY] = value
        return out

    @classmethod
    def formatType(cls, prefix, type_name):
        return "%s:%s" % (prefix, type_name)


# Read/Write classes split to follow precedence of the Java API
class GraphSONWriter(object):
    def __init__(self, serializer_map=None):
        """
        :param serializer_map: map from Python type to serializer instance implementing `dictify`
        """
        self.serializers = _serializers.copy()
        if serializer_map:
            self.serializers.update(serializer_map)

    def writeObject(self, objectData):
        # to JSON
        return json.dumps(self.toDict(objectData), separators=(',', ':'))

    def toDict(self, obj):
        """
        Encodes python objects in GraphSON type-tagged dict values
        """
        try:
            return self.serializers[type(obj)].dictify(obj, self)
        except KeyError:
            for key, serializer in self.serializers.items():
                if isinstance(obj, key):
                    return serializer.dictify(obj, self)

        if isinstance(obj, dict):
            return dict((self.toDict(k), self.toDict(v)) for k, v in obj.items())
        elif isinstance(obj, set):
            return set([self.toDict(o) for o in obj])
        elif isinstance(obj, list):
            return [self.toDict(o) for o in obj]
        else:
            return obj


class GraphSONReader(object):
    def __init__(self, deserializer_map=None):
        """
        :param deserializer_map: map from GraphSON type tag to deserializer instance implementing `objectify`
        """
        self.deserializers = _deserializers.copy()
        if deserializer_map:
            self.deserializers.update(deserializer_map)

    def readObject(self, jsonData):
        # from JSON
        return self.toObject(json.loads(jsonData))

    def toObject(self, obj):
        """
        Unpacks GraphSON type-tagged dict values into objects mapped in self.deserializers
        """
        if isinstance(obj, dict):
            try:
                return self.deserializers[obj[GraphSONUtil.TYPE_KEY]].objectify(obj[GraphSONUtil.VALUE_KEY], self)
            except KeyError:
                pass
            return dict((self.toObject(k), self.toObject(v)) for k, v in obj.items())
        elif isinstance(obj, set):
            return set([self.toObject(o) for o in obj])
        elif isinstance(obj, list):
            return [self.toObject(o) for o in obj]
        else:
            return obj


@six.add_metaclass(GraphSONTypeType)
class _GraphSONTypeIO(object):
    python_type = None
    graphson_type = None

    symbolMap = {"global_": "global", "as_": "as", "in_": "in", "and_": "and",
                 "or_": "or", "is_": "is", "not_": "not", "from_": "from",
                 "set_": "set", "list_": "list", "all_": "all", "with_": "with",
                 "filter_": "filter", "id_": "id", "max_": "max", "min_": "min", "sum_": "sum"}

    @classmethod
    def unmangleKeyword(cls, symbol):
        return cls.symbolMap.get(symbol, symbol)

    def dictify(self, obj, writer):
        raise NotImplementedError()

    def objectify(self, d, reader):
        raise NotImplementedError()


class _BytecodeSerializer(_GraphSONTypeIO):
    @classmethod
    def _dictify_instructions(cls, instructions, writer):
        out = []
        for instruction in instructions:
            inst = [instruction[0]]
            inst.extend(writer.toDict(arg) for arg in instruction[1:])
            out.append(inst)
        return out

    @classmethod
    def dictify(cls, bytecode, writer):
        if isinstance(bytecode, Traversal):
            bytecode = bytecode.bytecode
        out = {}
        if bytecode.source_instructions:
            out["source"] = cls._dictify_instructions(bytecode.source_instructions, writer)
        if bytecode.step_instructions:
            out["step"] = cls._dictify_instructions(bytecode.step_instructions, writer)
        return GraphSONUtil.typedValue("Bytecode", out)


class TraversalSerializer(_BytecodeSerializer):
    python_type = Traversal


class BytecodeSerializer(_BytecodeSerializer):
    python_type = Bytecode


class VertexSerializer(_GraphSONTypeIO):
    python_type = Vertex
    graphson_type = "g:Vertex"

    @classmethod
    def dictify(cls, vertex, writer):
        return GraphSONUtil.typedValue("Vertex", {"id": writer.toDict(vertex.id),
                                                  "label": writer.toDict(vertex.label)})


class EdgeSerializer(_GraphSONTypeIO):
    python_type = Edge
    graphson_type = "g:Edge"

    @classmethod
    def dictify(cls, edge, writer):
        return GraphSONUtil.typedValue("Edge", {"id": writer.toDict(edge.id),
                                                "outV": writer.toDict(edge.outV.id),
                                                "outVLabel": writer.toDict(edge.outV.label),
                                                "label": writer.toDict(edge.label),
                                                "inV": writer.toDict(edge.inV.id),
                                                "inVLabel": writer.toDict(edge.inV.label)})


class VertexPropertySerializer(_GraphSONTypeIO):
    python_type = VertexProperty
    graphson_type = "g:VertexProperty"

    @classmethod
    def dictify(cls, vertex_property, writer):
        return GraphSONUtil.typedValue("VertexProperty", {"id": writer.toDict(vertex_property.id),
                                                          "label": writer.toDict(vertex_property.label),
                                                          "value": writer.toDict(vertex_property.value),
                                                          "vertex": writer.toDict(vertex_property.vertex.id)})


class PropertySerializer(_GraphSONTypeIO):
    python_type = Property
    graphson_type = "g:Property"

    @classmethod
    def dictify(cls, property, writer):
        elementDict = writer.toDict(property.element)
        if elementDict is not None:
            valueDict = elementDict["@value"]
            if "outVLabel" in valueDict:
                del valueDict["outVLabel"]
            if "inVLabel" in valueDict:
                del valueDict["inVLabel"]
            if "properties" in valueDict:
                del valueDict["properties"]
            if "value" in valueDict:
                del valueDict["value"]
        return GraphSONUtil.typedValue("Property", {"key": writer.toDict(property.key),
                                                    "value": writer.toDict(property.value),
                                                    "element": elementDict})


class TraversalStrategySerializer(_GraphSONTypeIO):
    python_type = TraversalStrategy

    @classmethod
    def dictify(cls, strategy, writer):
        configuration = {}
        for key in strategy.configuration:
            configuration[key] = writer.toDict(strategy.configuration[key])
        return GraphSONUtil.typedValue(strategy.strategy_name, configuration)


class TraverserIO(_GraphSONTypeIO):
    python_type = Traverser
    graphson_type = "g:Traverser"

    @classmethod
    def dictify(cls, traverser, writer):
        return GraphSONUtil.typedValue("Traverser", {"value": writer.toDict(traverser.object),
                                                     "bulk": writer.toDict(traverser.bulk)})

    @classmethod
    def objectify(cls, d, reader):
        return Traverser(reader.toObject(d["value"]),
                         reader.toObject(d["bulk"]))


class EnumSerializer(_GraphSONTypeIO):
    python_type = Enum

    @classmethod
    def dictify(cls, enum, _):
        return GraphSONUtil.typedValue(cls.unmangleKeyword(type(enum).__name__),
                                       cls.unmangleKeyword(str(enum.name)))


class PSerializer(_GraphSONTypeIO):
    python_type = P

    @classmethod
    def dictify(cls, p, writer):
        out = {"predicate": p.operator,
               "value": [writer.toDict(p.value), writer.toDict(p.other)] if p.other is not None else
               writer.toDict(p.value)}
        return GraphSONUtil.typedValue("P", out)


class TextPSerializer(_GraphSONTypeIO):
    python_type = TextP

    @classmethod
    def dictify(cls, p, writer):
        out = {"predicate": p.operator,
               "value": [writer.toDict(p.value), writer.toDict(p.other)] if p.other is not None else
               writer.toDict(p.value)}
        return GraphSONUtil.typedValue("TextP", out)


class BindingSerializer(_GraphSONTypeIO):
    python_type = Binding

    @classmethod
    def dictify(cls, binding, writer):
        out = {"key": binding.key,
               "value": writer.toDict(binding.value)}
        return GraphSONUtil.typedValue("Binding", out)


class LambdaSerializer(_GraphSONTypeIO):
    python_type = FunctionType

    @classmethod
    def dictify(cls, lambda_object, writer):
        lambda_result = lambda_object()
        script = lambda_result if isinstance(lambda_result, str) else lambda_result[0]
        language = statics.default_lambda_language if isinstance(lambda_result, str) else lambda_result[1]
        out = {"script": script,
               "language": language}
        if language == "gremlin-jython" or language == "gremlin-python":
            if not script.strip().startswith("lambda"):
                script = "lambda " + script
                out["script"] = script
            out["arguments"] = six.get_function_code(eval(out["script"])).co_argcount
        elif language == "gremlin-groovy" and "->" in script:
            # if the user has explicitly added parameters to the groovy closure then we can easily detect one or two
            # arg lambdas - if we can't detect 1 or 2 then we just go with "unknown"
            args = script[0:script.find("->")]
            out["arguments"] = 2 if "," in args else 1
        else:
            out["arguments"] = -1

        return GraphSONUtil.typedValue("Lambda", out)


class TypeSerializer(_GraphSONTypeIO):
    python_type = TypeType

    @classmethod
    def dictify(cls, typ, writer):
        return writer.toDict(typ())


class UUIDIO(_GraphSONTypeIO):
    python_type = uuid.UUID
    graphson_type = "g:UUID"
    graphson_base_type = "UUID"

    @classmethod
    def dictify(cls, obj, writer):
        return GraphSONUtil.typedValue(cls.graphson_base_type, str(obj))

    @classmethod
    def objectify(cls, d, reader):
        return cls.python_type(d)


class DateIO(_GraphSONTypeIO):
    python_type = datetime.datetime
    graphson_type = "g:Date"
    graphson_base_type = "Date"

    @classmethod
    def dictify(cls, obj, writer):
        try:
            timestamp_seconds = calendar.timegm(obj.utctimetuple())
            pts = timestamp_seconds * 1e3 + getattr(obj, 'microsecond', 0) / 1e3
        except AttributeError:
            pts = calendar.timegm(obj.timetuple()) * 1e3

        ts = int(round(pts))
        return GraphSONUtil.typedValue(cls.graphson_base_type, ts)

    @classmethod
    def objectify(cls, ts, reader):
        # Python timestamp expects seconds
        return datetime.datetime.utcfromtimestamp(ts / 1000.0)


# Based on current implementation, this class must always be declared before FloatIO.
# Seems pretty fragile for future maintainers. Maybe look into this.
class TimestampIO(_GraphSONTypeIO):
    """A timestamp in Python is type float"""
    python_type = statics.timestamp
    graphson_type = "g:Timestamp"
    graphson_base_type = "Timestamp"

    @classmethod
    def dictify(cls, obj, writer):
        # Java timestamp expects milliseconds integer
        # Have to use int because of legacy Python
        ts = int(round(obj * 1000))
        return GraphSONUtil.typedValue(cls.graphson_base_type, ts)

    @classmethod
    def objectify(cls, ts, reader):
        # Python timestamp expects seconds
        return cls.python_type(ts / 1000.0)


class _NumberIO(_GraphSONTypeIO):
    @classmethod
    def dictify(cls, n, writer):
        if isinstance(n, bool):  # because isinstance(False, int) and isinstance(True, int)
            return n
        return GraphSONUtil.typedValue(cls.graphson_base_type, n)

    @classmethod
    def objectify(cls, v, _):
        return cls.python_type(v)


class ListIO(_GraphSONTypeIO):
    python_type = ListType
    graphson_type = "g:List"

    @classmethod
    def dictify(cls, l, writer):
        new_list = []
        for obj in l:
            new_list.append(writer.toDict(obj))
        return GraphSONUtil.typedValue("List", new_list)

    @classmethod
    def objectify(cls, l, reader):
        new_list = []
        for obj in l:
            new_list.append(reader.toObject(obj))
        return new_list


class SetIO(_GraphSONTypeIO):
    python_type = SetType
    graphson_type = "g:Set"

    @classmethod
    def dictify(cls, s, writer):
        new_list = []
        for obj in s:
            new_list.append(writer.toDict(obj))
        return GraphSONUtil.typedValue("Set", new_list)

    @classmethod
    def objectify(cls, s, reader):
        """
        By default, returns a python set

        In case Java returns numeric values of different types which
        python don't recognize, coerce and return a list.
        See comments of TINKERPOP-1844 for more details
        """
        new_list = [reader.toObject(obj) for obj in s]
        new_set = set(new_list)
        if len(new_list) != len(new_set):
            log.warning("Coercing g:Set to list due to java numeric values. "
                        "See TINKERPOP-1844 for more details.")
            return new_list

        return new_set


class MapType(_GraphSONTypeIO):
    python_type = DictType
    graphson_type = "g:Map"

    @classmethod
    def dictify(cls, d, writer):
        l = []
        for key in d:
            l.append(writer.toDict(key))
            l.append(writer.toDict(d[key]))
        return GraphSONUtil.typedValue("Map", l)

    @classmethod
    def objectify(cls, l, reader):
        new_dict = {}
        if len(l) > 0:
            x = 0
            while x < len(l):
                new_dict[reader.toObject(l[x])] = reader.toObject(l[x + 1])
                x = x + 2
        return new_dict


class BulkSetIO(_GraphSONTypeIO):
    graphson_type = "g:BulkSet"

    @classmethod
    def objectify(cls, l, reader):
        new_list = []

        # this approach basically mimics what currently existed in 3.3.4 and prior versions where BulkSet is
        # basically just coerced to list. the limitation here is that if the value of a bulk exceeds the size of
        # a list (into the long space) then stuff won't work nice.
        if len(l) > 0:
            x = 0
            while x < len(l):
                obj = reader.toObject(l[x])
                bulk = reader.toObject(l[x + 1])
                for y in range(bulk):
                    new_list.append(obj)
                x = x + 2
        return new_list


class FloatIO(_NumberIO):
    python_type = FloatType
    graphson_type = "g:Float"
    graphson_base_type = "Float"

    @classmethod
    def dictify(cls, n, writer):
        if isinstance(n, bool):  # because isinstance(False, int) and isinstance(True, int)
            return n
        elif math.isnan(n):
            return GraphSONUtil.typedValue(cls.graphson_base_type, "NaN")
        elif math.isinf(n) and n > 0:
            return GraphSONUtil.typedValue(cls.graphson_base_type, "Infinity")
        elif math.isinf(n) and n < 0:
            return GraphSONUtil.typedValue(cls.graphson_base_type, "-Infinity")
        else:
            return GraphSONUtil.typedValue(cls.graphson_base_type, n)

    @classmethod
    def objectify(cls, v, _):
        if isinstance(v, str):
            if v == 'NaN':
                return float('nan')
            elif v == "Infinity":
                return float('inf')
            elif v == "-Infinity":
                return float('-inf')

        return cls.python_type(v)


class BigDecimalIO(_NumberIO):
    python_type = Decimal
    graphson_type = "gx:BigDecimal"
    graphson_base_type = "BigDecimal"

    @classmethod
    def dictify(cls, n, writer):
        if isinstance(n, bool):  # because isinstance(False, int) and isinstance(True, int)
            return n
        elif math.isnan(n):
            return GraphSONUtil.typedValue(cls.graphson_base_type, "NaN", "gx")
        elif math.isinf(n) and n > 0:
            return GraphSONUtil.typedValue(cls.graphson_base_type, "Infinity", "gx")
        elif math.isinf(n) and n < 0:
            return GraphSONUtil.typedValue(cls.graphson_base_type, "-Infinity", "gx")
        else:
            return GraphSONUtil.typedValue(cls.graphson_base_type, str(n), "gx")

    @classmethod
    def objectify(cls, v, _):
        if isinstance(v, str):
            if v == 'NaN':
                return Decimal('nan')
            elif v == "Infinity":
                return Decimal('inf')
            elif v == "-Infinity":
                return Decimal('-inf')

        return Decimal(v)


class DoubleIO(FloatIO):
    graphson_type = "g:Double"
    graphson_base_type = "Double"


class Int64IO(_NumberIO):
    python_type = LongType
    graphson_type = "g:Int64"
    graphson_base_type = "Int64"

    @classmethod
    def dictify(cls, n, writer):
        # if we exceed Java long range then we need a BigInteger
        if isinstance(n, bool):
            return n
        elif n < -9223372036854775808 or n > 9223372036854775807:
            return GraphSONUtil.typedValue("BigInteger", str(n), "gx")
        else:
            return GraphSONUtil.typedValue(cls.graphson_base_type, n)


class BigIntegerIO(Int64IO):
    graphson_type = "gx:BigInteger"


class Int32IO(Int64IO):
    python_type = IntType
    graphson_type = "g:Int32"
    graphson_base_type = "Int32"

    @classmethod
    def dictify(cls, n, writer):
        # if we exceed Java int range then we need a long
        if isinstance(n, bool):
            return n
        elif n < -9223372036854775808 or n > 9223372036854775807:
            return GraphSONUtil.typedValue("BigInteger", str(n), "gx")
        elif n < -2147483648 or n > 2147483647:
            return GraphSONUtil.typedValue("Int64", n)
        else:
            return GraphSONUtil.typedValue(cls.graphson_base_type, n)

class ByteIO(_NumberIO):
    python_type = SingleByte
    graphson_type = "gx:Byte"
    graphson_base_type = "Byte"

    @classmethod
    def dictify(cls, n, writer):
        if isinstance(n, bool):  # because isinstance(False, int) and isinstance(True, int)
            return n
        return GraphSONUtil.typedValue(cls.graphson_base_type, n, "gx")

    @classmethod
    def objectify(cls, v, _):
        return int.__new__(SingleByte, v)


class ByteBufferIO(_GraphSONTypeIO):
    python_type = ByteBufferType
    graphson_type = "gx:ByteBuffer"
    graphson_base_type = "ByteBuffer"

    @classmethod
    def dictify(cls, n, writer):
        return GraphSONUtil.typedValue(cls.graphson_base_type, "".join(chr(x) for x in n), "gx")

    @classmethod
    def objectify(cls, v, _):
        return cls.python_type(v, "utf8")


class CharIO(_GraphSONTypeIO):
    python_type = SingleChar
    graphson_type = "gx:Char"
    graphson_base_type = "Char"

    @classmethod
    def dictify(cls, n, writer):
        return GraphSONUtil.typedValue(cls.graphson_base_type, n, "gx")

    @classmethod
    def objectify(cls, v, _):
        return str.__new__(SingleChar, v)


class DurationIO(_GraphSONTypeIO):
    python_type = timedelta
    graphson_type = "gx:Duration"
    graphson_base_type = "Duration"

    @classmethod
    def dictify(cls, n, writer):
        return GraphSONUtil.typedValue(cls.graphson_base_type, duration_isoformat(n), "gx")

    @classmethod
    def objectify(cls, v, _):
        return parse_duration(v)


class VertexDeserializer(_GraphSONTypeIO):
    graphson_type = "g:Vertex"

    @classmethod
    def objectify(cls, d, reader):
        return Vertex(reader.toObject(d["id"]), d.get("label", "vertex"))


class EdgeDeserializer(_GraphSONTypeIO):
    graphson_type = "g:Edge"

    @classmethod
    def objectify(cls, d, reader):
        return Edge(reader.toObject(d["id"]),
                    Vertex(reader.toObject(d["outV"]), d.get("outVLabel", "vertex")),
                    d.get("label", "edge"),
                    Vertex(reader.toObject(d["inV"]), d.get("inVLabel", "vertex")))


class VertexPropertyDeserializer(_GraphSONTypeIO):
    graphson_type = "g:VertexProperty"

    @classmethod
    def objectify(cls, d, reader):
        vertex = Vertex(reader.toObject(d.get("vertex"))) if "vertex" in d else None
        return VertexProperty(reader.toObject(d["id"]),
                              d["label"],
                              reader.toObject(d["value"]),
                              vertex)


class PropertyDeserializer(_GraphSONTypeIO):
    graphson_type = "g:Property"

    @classmethod
    def objectify(cls, d, reader):
        element = reader.toObject(d["element"]) if "element" in d else None
        return Property(d["key"], reader.toObject(d["value"]), element)


class PathDeserializer(_GraphSONTypeIO):
    graphson_type = "g:Path"

    @classmethod
    def objectify(cls, d, reader):
        return Path(reader.toObject(d["labels"]), reader.toObject(d["objects"]))


class TDeserializer(_GraphSONTypeIO):
    graphson_type = "g:T"

    @classmethod
    def objectify(cls, d, reader):
        return T[d]


class DirectionIO(_GraphSONTypeIO):
    graphson_type = "g:Direction"
    graphson_base_type = "Direction"
    python_type = Direction

    @classmethod
    def dictify(cls, d, writer):
        return GraphSONUtil.typedValue(cls.graphson_base_type, d.name, "g")

    @classmethod
    def objectify(cls, d, reader):
        return Direction[d]


class TraversalMetricsDeserializer(_GraphSONTypeIO):
    graphson_type = "g:TraversalMetrics"

    @classmethod
    def objectify(cls, d, reader):
        return reader.toObject(d)


class MetricsDeserializer(_GraphSONTypeIO):
    graphson_type = "g:Metrics"

    @classmethod
    def objectify(cls, d, reader):
        return reader.toObject(d)