Exemple #1
0
def _encode_map(dict_obj):
    kv_list = [(key.encode('utf-8'), val.encode('utf-8'))
               for key, val in dict_obj.items()]
    coder = IterableCoder(
        TupleCoder(
            [LengthPrefixCoder(BytesCoder()),
             LengthPrefixCoder(BytesCoder())]))
    coder_urns = [
        'beam:coder:iterable:v1', 'beam:coder:kv:v1', 'beam:coder:bytes:v1',
        'beam:coder:bytes:v1'
    ]
    return ConfigValue(coder_urn=coder_urns, payload=coder.encode(kv_list))
Exemple #2
0
 def test_implicit_payload_builder_with_bytes(self):
   values = PayloadBase.bytes_values
   builder = ImplicitSchemaPayloadBuilder(values)
   result = builder.build()
   if sys.version_info[0] < 3:
     # in python 2.x bytes coder will be inferred
     args = {
         'integer_example': ConfigValue(
             coder_urn=['beam:coder:varint:v1'],
             payload=VarIntCoder()
             .get_impl().encode_nested(values['integer_example'])),
         'string_example': ConfigValue(
             coder_urn=['beam:coder:bytes:v1'],
             payload=StrUtf8Coder()
             .get_impl().encode_nested(values['string_example'])),
         'list_of_strings': ConfigValue(
             coder_urn=['beam:coder:iterable:v1',
                        'beam:coder:bytes:v1'],
             payload=IterableCoder(StrUtf8Coder())
             .get_impl().encode_nested(values['list_of_strings'])),
         'optional_kv': ConfigValue(
             coder_urn=['beam:coder:kv:v1',
                        'beam:coder:bytes:v1',
                        'beam:coder:double:v1'],
             payload=TupleCoder([StrUtf8Coder(), FloatCoder()])
             .get_impl().encode_nested(values['optional_kv'])),
     }
     expected = get_payload(args)
     self.assertEqual(result, expected)
   else:
     expected = get_payload(PayloadBase.args)
     self.assertEqual(result, expected)
Exemple #3
0
    class SimpleTestStatefulDoFn(DoFn):
      BUFFER_STATE = CombiningValueStateSpec(
          'buffer',
          IterableCoder(VarIntCoder()), ToListCombineFn())
      EXPIRY_TIMER = TimerSpec('expiry1', TimeDomain.WATERMARK)

      def process(self, element, buffer=DoFn.StateParam(BUFFER_STATE),
                  timer1=DoFn.TimerParam(EXPIRY_TIMER)):
        unused_key, value = element
        buffer.add(value)
        timer1.set(20)

      @on_timer(EXPIRY_TIMER)
      def expiry_callback(self, buffer=DoFn.StateParam(BUFFER_STATE),
                          timer=DoFn.TimerParam(EXPIRY_TIMER)):
        yield ''.join(str(x) for x in sorted(buffer.read()))
Exemple #4
0
class PayloadBase(object):
    values = {
        'integer_example': 1,
        'boolean': True,
        'string_example': u'thing',
        'list_of_strings': [u'foo', u'bar'],
        'optional_kv': (u'key', 1.1),
        'optional_integer': None,
    }

    bytes_values = {
        'integer_example': 1,
        'boolean': True,
        'string_example': 'thing',
        'list_of_strings': ['foo', 'bar'],
        'optional_kv': ('key', 1.1),
        'optional_integer': None,
    }

    args = {
        'integer_example':
        ConfigValue(coder_urn=['beam:coder:varint:v1'],
                    payload=VarIntCoder().get_impl().encode_nested(
                        values['integer_example'])),
        'boolean':
        ConfigValue(coder_urn=['beam:coder:bool:v1'],
                    payload=BooleanCoder().get_impl().encode_nested(
                        values['boolean'])),
        'string_example':
        ConfigValue(coder_urn=['beam:coder:string_utf8:v1'],
                    payload=StrUtf8Coder().get_impl().encode_nested(
                        values['string_example'])),
        'list_of_strings':
        ConfigValue(
            coder_urn=['beam:coder:iterable:v1', 'beam:coder:string_utf8:v1'],
            payload=IterableCoder(StrUtf8Coder()).get_impl().encode_nested(
                values['list_of_strings'])),
        'optional_kv':
        ConfigValue(coder_urn=[
            'beam:coder:kv:v1', 'beam:coder:string_utf8:v1',
            'beam:coder:double:v1'
        ],
                    payload=TupleCoder([
                        StrUtf8Coder(), FloatCoder()
                    ]).get_impl().encode_nested(values['optional_kv'])),
    }

    def get_payload_from_typing_hints(self, values):
        """Return ExternalConfigurationPayload based on python typing hints"""
        raise NotImplementedError

    def get_payload_from_beam_typehints(self, values):
        """Return ExternalConfigurationPayload based on beam typehints"""
        raise NotImplementedError

    def test_typing_payload_builder(self):
        result = self.get_payload_from_typing_hints(self.values)
        expected = get_payload(self.args)
        self.assertEqual(result, expected)

    def test_typing_payload_builder_with_bytes(self):
        """
    string_utf8 coder will be used even if values are not unicode in python 2.x
    """
        result = self.get_payload_from_typing_hints(self.bytes_values)
        expected = get_payload(self.args)
        self.assertEqual(result, expected)

    def test_typehints_payload_builder(self):
        result = self.get_payload_from_beam_typehints(self.values)
        expected = get_payload(self.args)
        self.assertEqual(result, expected)

    def test_typehints_payload_builder_with_bytes(self):
        """
    string_utf8 coder will be used even if values are not unicode in python 2.x
    """
        result = self.get_payload_from_beam_typehints(self.bytes_values)
        expected = get_payload(self.args)
        self.assertEqual(result, expected)

    def test_optional_error(self):
        """
    value can only be None if typehint is Optional
    """
        with self.assertRaises(RuntimeError):
            self.get_payload_from_typing_hints({k: None for k in self.values})
Exemple #5
0
def _encode_list(list_obj):
    encoded_list = [val.encode('utf-8') for val in list_obj]
    coder = IterableCoder(LengthPrefixCoder(BytesCoder()))
    coder_urns = ['beam:coder:iterable:v1', 'beam:coder:bytes:v1']
    return ConfigValue(coder_urn=coder_urns,
                       payload=coder.encode(encoded_list))