You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@flink.apache.org by he...@apache.org on 2020/08/15 08:52:26 UTC
[flink] branch master updated: [FLINK-18879][python] Support Row
Serialization and Deserialization schemas for Python DataStream API.
(#13150)
This is an automated email from the ASF dual-hosted git repository.
hequn pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/flink.git
The following commit(s) were added to refs/heads/master by this push:
new 1fc400d [FLINK-18879][python] Support Row Serialization and Deserialization schemas for Python DataStream API. (#13150)
1fc400d is described below
commit 1fc400d1f94a10d2f5b227cc8eaa1ac9754ec237
Author: Shuiqiang Chen <ac...@alibaba-inc.com>
AuthorDate: Sat Aug 15 16:51:07 2020 +0800
[FLINK-18879][python] Support Row Serialization and Deserialization schemas for Python DataStream API. (#13150)
---
.../pyflink/common/serialization_schemas.py | 367 +++++++++++++++++++++
.../common/tests/test_serialization_schemas.py | 103 ++++++
2 files changed, 470 insertions(+)
diff --git a/flink-python/pyflink/common/serialization_schemas.py b/flink-python/pyflink/common/serialization_schemas.py
new file mode 100644
index 0000000..3748e5a
--- /dev/null
+++ b/flink-python/pyflink/common/serialization_schemas.py
@@ -0,0 +1,367 @@
+################################################################################
+# 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.
+################################################################################
+from py4j.java_gateway import java_import
+from pyflink.common.typeinfo import TypeInformation, WrapperTypeInfo
+
+from pyflink.util.utils import load_java_class
+
+from pyflink.java_gateway import get_gateway
+
+
+class SerializationSchema(object):
+ """
+ Base class for SerializationSchema. The serialization schema describes how to turn a data object
+ into a different serialized representation. Most data sinks (for example Apache Kafka) require
+ the data to be handed to them in a specific format (for example as byte strings).
+ """
+ def __init__(self, j_serialization_schema=None):
+ self._j_serialization_schema = j_serialization_schema
+
+
+class DeserializationSchema(object):
+ """
+ Base class for DeserializationSchema. The deserialization schema describes how to turn the byte
+ messages delivered by certain data sources (for example Apache Kafka) into data types (Java/
+ Scala objects) that are processed by Flink.
+
+ In addition, the DeserializationSchema describes the produced type which lets Flink create
+ internal serializers and structures to handle the type.
+ """
+ def __init__(self, j_deserialization_schema=None):
+ self._j_deserialization_schema = j_deserialization_schema
+
+
+class SimpleStringSchema(SerializationSchema, DeserializationSchema):
+ """
+ Very simple serialization/deserialization schema for strings. By default, the serializer uses
+ 'UTF-8' for string/byte conversion.
+ """
+
+ def __init__(self, charset: str = 'UTF-8'):
+ gate_way = get_gateway()
+ j_char_set = gate_way.jvm.java.nio.charset.Charset.forName(charset)
+ j_simple_string_serialization_schema = gate_way \
+ .jvm.org.apache.flink.api.common.serialization.SimpleStringSchema(j_char_set)
+ SerializationSchema.__init__(self,
+ j_serialization_schema=j_simple_string_serialization_schema)
+ DeserializationSchema.__init__(
+ self, j_deserialization_schema=j_simple_string_serialization_schema)
+
+
+class JsonRowDeserializationSchema(DeserializationSchema):
+ """
+ Deserialization schema from JSON to Flink types.
+
+ Deserializes a byte[] message as a JSON object and reads the specified fields.
+
+ Failures during deserialization are forwarded as wrapped IOExceptions.
+ """
+ def __init__(self, j_deserialization_schema):
+ super(JsonRowDeserializationSchema, self).__init__(j_deserialization_schema)
+
+ @staticmethod
+ def builder():
+ """
+ A static method to get a Builder for JsonRowDeserializationSchema.
+ """
+ return JsonRowDeserializationSchema.Builder()
+
+ class Builder(object):
+ """
+ Builder for JsonRowDeserializationSchema.
+ """
+
+ def __init__(self):
+ self._type_info = None
+ self._fail_on_missing_field = False
+ self._ignore_parse_errors = False
+
+ def type_info(self, type_info: TypeInformation):
+ """
+ Creates a JSON deserialization schema for the given type information.
+
+ :param type_info: Type information describing the result type. The field names of Row
+ are used to parse the JSON properties.
+ """
+ self._type_info = type_info
+ return self
+
+ def json_schema(self, json_schema: str):
+ """
+ Creates a JSON deserialization schema for the given JSON schema.
+
+ :param json_schema: JSON schema describing the result type.
+ """
+ if json_schema is None:
+ raise TypeError("The json_schema must not be None.")
+ j_type_info = get_gateway().jvm \
+ .org.apache.flink.formats.json.JsonRowSchemaConverter.convert(json_schema)
+ self._type_info = WrapperTypeInfo(j_type_info)
+ return self
+
+ def fail_on_missing_field(self):
+ """
+ Configures schema to fail if a JSON field is missing. A missing field is ignored and the
+ field is set to null by default.
+ """
+ self._fail_on_missing_field = True
+ return self
+
+ def ignore_parse_errors(self):
+ """
+ Configures schema to fail when parsing json failed. An exception will be thrown when
+ parsing json fails.
+ """
+ self._ignore_parse_errors = True
+ return self
+
+ def build(self):
+ JBuilder = get_gateway().jvm.org.apache.flink.formats.json.JsonRowDeserializationSchema\
+ .Builder
+ j_builder = JBuilder(self._type_info.get_java_type_info())
+
+ if self._fail_on_missing_field:
+ j_builder = j_builder.fialOnMissingField()
+
+ if self._ignore_parse_errors:
+ j_builder = j_builder.ignoreParseErrors()
+
+ j_deserialization_schema = j_builder.build()
+ return JsonRowDeserializationSchema(j_deserialization_schema=j_deserialization_schema)
+
+
+class JsonRowSerializationSchema(SerializationSchema):
+ """
+ Serialization schema that serializes an object of Flink types into a JSON bytes. Serializes the
+ input Flink object into a JSON string and converts it into byte[].
+
+ Result byte[] message can be deserialized using JsonRowDeserializationSchema.
+ """
+
+ def __init__(self, j_serialization_schema):
+ super(JsonRowSerializationSchema, self).__init__(j_serialization_schema)
+
+ @staticmethod
+ def builder():
+ return JsonRowSerializationSchema.Builder()
+
+ class Builder(object):
+ """
+ Builder for JsonRowSerializationSchema.
+ """
+ def __init__(self):
+ self._type_info = None
+
+ def with_type_info(self, type_info: TypeInformation):
+ """
+ Creates a JSON serialization schema for the given type information.
+
+ :param type_info: Type information describing the result type. The field names of Row
+ are used to parse the JSON properties.
+ """
+ self._type_info = type_info
+ return self
+
+ def build(self):
+ if self._type_info is None:
+ raise TypeError("Typeinfo should be set.")
+
+ j_builder = get_gateway().jvm \
+ .org.apache.flink.formats.json.JsonRowSerializationSchema.builder()
+
+ j_schema = j_builder.withTypeInfo(self._type_info.get_java_type_info()).build()
+ return JsonRowSerializationSchema(j_serialization_schema=j_schema)
+
+
+class CsvRowDeserializationSchema(DeserializationSchema):
+ """
+ Deserialization schema from CSV to Flink types. Deserializes a byte[] message as a JsonNode and
+ converts it to Row.
+
+ Failure during deserialization are forwarded as wrapped IOException.
+ """
+
+ def __init__(self, j_deserialization_schema):
+ super(CsvRowDeserializationSchema, self).__init__(
+ j_deserialization_schema=j_deserialization_schema)
+
+ class Builder(object):
+ """
+ A builder for creating a CsvRowDeserializationSchema.
+ """
+ def __init__(self, type_info: TypeInformation):
+ if type_info is None:
+ raise TypeError("Type information must not be None")
+ self._j_builder = get_gateway().jvm\
+ .org.apache.flink.formats.csv.CsvRowDeserializationSchema.Builder(
+ type_info.get_java_type_info())
+
+ def set_field_delimiter(self, delimiter: str):
+ self._j_builder = self._j_builder.setFieldDelimiter(delimiter)
+ return self
+
+ def set_allow_comments(self, allow_comments: bool):
+ self._j_builder = self._j_builder.setAllowComments(allow_comments)
+ return self
+
+ def set_array_element_delimiter(self, delimiter: str):
+ self._j_builder = self._j_builder.setArrayElementDelimiter(delimiter)
+ return self
+
+ def set_quote_character(self, c: str):
+ self._j_builder = self._j_builder.setQuoteCharacter(c)
+ return self
+
+ def set_escape_character(self, c: str):
+ self._j_builder = self._j_builder.setEscapeCharacter(c)
+ return self
+
+ def set_null_literal(self, null_literal: str):
+ self._j_builder = self._j_builder.setNullLiteral(null_literal)
+ return self
+
+ def set_ignore_parse_errors(self, ignore_parse_errors: bool):
+ self._j_builder = self._j_builder.setIgnoreParseErrors(ignore_parse_errors)
+ return self
+
+ def build(self):
+ j_csv_row_deserialization_schema = self._j_builder.build()
+ return CsvRowDeserializationSchema(
+ j_deserialization_schema=j_csv_row_deserialization_schema)
+
+
+class CsvRowSerializationSchema(SerializationSchema):
+ """
+ Serialization schema that serializes an object of Flink types into a CSV bytes. Serializes the
+ input row into an ObjectNode and converts it into byte[].
+
+ Result byte[] messages can be deserialized using CsvRowDeserializationSchema.
+ """
+ def __init__(self, j_csv_row_serialization_schema):
+ super(CsvRowSerializationSchema, self).__init__(j_csv_row_serialization_schema)
+
+ class Builder(object):
+ """
+ A builder for creating a CsvRowSerializationSchema.
+ """
+ def __init__(self, type_info: TypeInformation):
+ if type_info is None:
+ raise TypeError("Type information must not be None")
+ self._j_builder = get_gateway().jvm\
+ .org.apache.flink.formats.csv.CsvRowSerializationSchema.Builder(
+ type_info.get_java_type_info())
+
+ def set_field_delimiter(self, c: str):
+ self._j_builder = self._j_builder.setFieldDelimiter(c)
+ return self
+
+ def set_line_delimiter(self, delimiter: str):
+ self._j_builder = self._j_builder.setLineDelimiter(delimiter)
+ return self
+
+ def set_array_element_delimiter(self, delimiter: str):
+ self._j_builder = self._j_builder.setArrayElementDelimiter(delimiter)
+ return self
+
+ def disable_quote_character(self):
+ self._j_builder = self._j_builder.disableQuoteCharacter()
+ return self
+
+ def set_quote_character(self, c: str):
+ self._j_builder = self._j_builder.setQuoteCharacter(c)
+ return self
+
+ def set_escape_character(self, c: str):
+ self._j_builder = self._j_builder.setEscapeCharacter(c)
+ return self
+
+ def set_null_literal(self, s: str):
+ self._j_builder = self._j_builder.setNullLiteral(s)
+ return self
+
+ def build(self):
+ j_serialization_schema = self._j_builder.build()
+ return CsvRowSerializationSchema(j_csv_row_serialization_schema=j_serialization_schema)
+
+
+class AvroRowDeserializationSchema(DeserializationSchema):
+ """
+ Deserialization schema from Avro bytes to Row. Deserializes the byte[] messages into (nested)
+ Flink rows. It converts Avro types into types that are compatible with Flink's Table & SQL API.
+
+ Projects with Avro records containing logical date/time types need to add a JodaTime dependency.
+ """
+ def __init__(self, record_class: str = None, avro_schema_string: str = None):
+ """
+ Creates an Avro deserialization schema for the given specific record class or Avro schema
+ string. Having the concrete Avro record class might improve performance.
+
+ :param record_class: Avro record class used to deserialize Avro's record to Flink's row.
+ :param avro_schema_string: Avro schema string to deserialize Avro's record to Flink's row.
+ """
+
+ if avro_schema_string is None and record_class is None:
+ raise TypeError("record_class or avro_schema_string should be specified.")
+ j_deserialization_schema = None
+ if record_class is not None:
+ gateway = get_gateway()
+ java_import(gateway.jvm, record_class)
+ j_record_class = load_java_class(record_class)
+ JAvroRowDeserializationSchema = get_gateway().jvm \
+ .org.apache.flink.formats.avro.AvroRowDeserializationSchema
+ j_deserialization_schema = JAvroRowDeserializationSchema(j_record_class)
+
+ elif avro_schema_string is not None:
+ JAvroRowDeserializationSchema = get_gateway().jvm \
+ .org.apache.flink.formats.avro.AvroRowDeserializationSchema
+ j_deserialization_schema = JAvroRowDeserializationSchema(avro_schema_string)
+
+ super(AvroRowDeserializationSchema, self).__init__(j_deserialization_schema)
+
+
+class AvroRowSerializationSchema(SerializationSchema):
+ """
+ Serialization schema that serializes to Avro binary format.
+ """
+
+ def __init__(self, record_class: str = None, avro_schema_string: str = None):
+ """
+ Creates AvroSerializationSchema that serializes SpecificRecord using provided schema or
+ record class.
+
+ :param record_class: Avro record class used to serialize Flink's row to Avro's record.
+ :param avro_schema_string: Avro schema string to serialize Flink's row to Avro's record.
+ """
+ if avro_schema_string is None and record_class is None:
+ raise TypeError("record_class or avro_schema_string should be specified.")
+
+ j_serialization_schema = None
+ if record_class is not None:
+ gateway = get_gateway()
+ java_import(gateway.jvm, record_class)
+ j_record_class = load_java_class(record_class)
+ JAvroRowSerializationSchema = get_gateway().jvm \
+ .org.apache.flink.formats.avro.AvroRowSerializationSchema
+ j_serialization_schema = JAvroRowSerializationSchema(j_record_class)
+
+ elif avro_schema_string is not None:
+ JAvroRowSerializationSchema = get_gateway().jvm \
+ .org.apache.flink.formats.avro.AvroRowSerializationSchema
+ j_serialization_schema = JAvroRowSerializationSchema(avro_schema_string)
+
+ super(AvroRowSerializationSchema, self).__init__(j_serialization_schema)
diff --git a/flink-python/pyflink/common/tests/test_serialization_schemas.py b/flink-python/pyflink/common/tests/test_serialization_schemas.py
new file mode 100644
index 0000000..02e51e5
--- /dev/null
+++ b/flink-python/pyflink/common/tests/test_serialization_schemas.py
@@ -0,0 +1,103 @@
+################################################################################
+# 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.
+################################################################################
+from pyflink.common.serialization_schemas import JsonRowSerializationSchema, \
+ JsonRowDeserializationSchema, CsvRowSerializationSchema, CsvRowDeserializationSchema, \
+ SimpleStringSchema
+from pyflink.common.typeinfo import Types
+from pyflink.java_gateway import get_gateway
+from pyflink.testing.test_case_utils import PyFlinkTestCase
+
+
+class TestRowSerializationSchemas(PyFlinkTestCase):
+
+ def test_simple_string_schema(self):
+ expected_string = 'test string'
+ simple_string_schema = SimpleStringSchema()
+ self.assertEqual(expected_string.encode(encoding='utf-8'),
+ simple_string_schema._j_serialization_schema.serialize(expected_string))
+
+ self.assertEqual(expected_string, simple_string_schema._j_deserialization_schema
+ .deserialize(expected_string.encode(encoding='utf-8')))
+
+ def test_json_row_serialization_deserialization_schema(self):
+ jsons = ["{\"svt\":\"2020-02-24T12:58:09.209+0800\"}",
+ "{\"svt\":\"2020-02-24T12:58:09.209+0800\", "
+ "\"ops\":{\"id\":\"281708d0-4092-4c21-9233-931950b6eccf\"},\"ids\":[1, 2, 3]}",
+ "{\"svt\":\"2020-02-24T12:58:09.209+0800\"}"]
+ expected_jsons = ["{\"svt\":\"2020-02-24T12:58:09.209+0800\",\"ops\":null,\"ids\":null}",
+ "{\"svt\":\"2020-02-24T12:58:09.209+0800\","
+ "\"ops\":{\"id\":\"281708d0-4092-4c21-9233-931950b6eccf\"},"
+ "\"ids\":[1,2,3]}",
+ "{\"svt\":\"2020-02-24T12:58:09.209+0800\",\"ops\":null,\"ids\":null}"]
+
+ row_schema = Types.ROW_NAMED(["svt", "ops", "ids"],
+ [Types.STRING(),
+ Types.ROW_NAMED(['id'], [Types.STRING()]),
+ Types.PRIMITIVE_ARRAY(Types.INT())])
+
+ json_row_serialization_schema = JsonRowSerializationSchema.builder() \
+ .with_type_info(row_schema).build()
+ json_row_deserialization_schema = JsonRowDeserializationSchema.builder() \
+ .type_info(row_schema).build()
+
+ for i in range(len(jsons)):
+ j_row = json_row_deserialization_schema._j_deserialization_schema\
+ .deserialize(bytes(jsons[i], encoding='utf-8'))
+ result = str(json_row_serialization_schema._j_serialization_schema.serialize(j_row),
+ encoding='utf-8')
+ self.assertEqual(expected_jsons[i], result)
+
+ def test_csv_row_serialization_schema(self):
+ JRow = get_gateway().jvm.org.apache.flink.types.Row
+
+ j_row = JRow(3)
+ j_row.setField(0, "BEGIN")
+ j_row.setField(2, "END")
+
+ def field_assertion(field_info, csv_value, value, field_delimiter):
+ row_info = Types.ROW([Types.STRING(), field_info, Types.STRING()])
+ expected_csv = "BEGIN" + field_delimiter + csv_value + field_delimiter + "END\n"
+ j_row.setField(1, value)
+
+ csv_row_serialization_schema = CsvRowSerializationSchema.Builder(row_info)\
+ .set_escape_character('*').set_quote_character('\'')\
+ .set_array_element_delimiter(':').set_field_delimiter(';').build()
+ csv_row_deserialization_schema = CsvRowDeserializationSchema.Builder(row_info)\
+ .set_escape_character('*').set_quote_character('\'')\
+ .set_array_element_delimiter(':').set_field_delimiter(';').build()
+
+ serialized_bytes = csv_row_serialization_schema._j_serialization_schema.serialize(j_row)
+ self.assertEqual(expected_csv, str(serialized_bytes, encoding='utf-8'))
+
+ j_deserialized_row = csv_row_deserialization_schema._j_deserialization_schema\
+ .deserialize(expected_csv.encode("utf-8"))
+ self.assertTrue(j_row.equals(j_deserialized_row))
+
+ field_assertion(Types.STRING(), "'123''4**'", "123'4*", ";")
+ field_assertion(Types.STRING(), "'a;b''c'", "a;b'c", ";")
+ field_assertion(Types.INT(), "12", 12, ";")
+
+ test_j_row = JRow(2)
+ test_j_row.setField(0, "1")
+ test_j_row.setField(1, "hello")
+
+ field_assertion(Types.ROW([Types.STRING(), Types.STRING()]), "'1:hello'", test_j_row, ";")
+ test_j_row.setField(1, "hello world")
+ field_assertion(Types.ROW([Types.STRING(), Types.STRING()]), "'1:hello world'", test_j_row,
+ ";")
+ field_assertion(Types.STRING(), "null", "null", ";")