flink-python/pyflink/proto/flink-fn-execution.proto

/*
 * 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.
 */

// NOTE: File flink_fn_execution_pb2.py is generated from this file. Please re-generate it by calling
// gen_protos.py whenever this file is changed.
syntax = "proto3";

package org.apache.flink.fn_execution.v1;

option java_package = "org.apache.flink.fnexecution.v1";
option java_outer_classname = "FlinkFnApi";

message Input {
  oneof input {
    UserDefinedFunction udf = 1;
    int32 inputOffset = 2;
    bytes inputConstant = 3;
  }
}

// User-defined function definition. It supports chaining functions, that's, the execution
// result of one user-defined function as the input of another user-defined function.
message UserDefinedFunction {
  // The serialized representation of the user-defined function
  bytes payload = 1;

  // The input arguments of the user-defined function, it could be one of the following:
  // 1. A column from the input row
  // 2. The result of another user-defined function
  // 3. The constant value of the column
  repeated Input inputs = 2;

  // The index of the over window used in pandas batch over window aggregation
  int32 window_index = 3;

  // Whether it's pandas UDF
  bool is_pandas_udf = 5;
}

// A list of user-defined functions to be executed in a batch.
message UserDefinedFunctions {
  repeated UserDefinedFunction udfs = 1;
  bool metric_enabled = 2;
  repeated OverWindow windows = 3;
}

// Used to describe the info of over window in pandas batch over window aggregation
message OverWindow {
  enum WindowType {
    RANGE_UNBOUNDED = 0;
    RANGE_UNBOUNDED_PRECEDING = 1;
    RANGE_UNBOUNDED_FOLLOWING = 2;
    RANGE_SLIDING = 3;
    ROW_UNBOUNDED = 4;
    ROW_UNBOUNDED_PRECEDING = 5;
    ROW_UNBOUNDED_FOLLOWING = 6;
    ROW_SLIDING = 7;
  }
  WindowType window_type = 1;
  int64 lower_boundary = 2;
  int64 upper_boundary = 3;
}

// User defined DataStream function definition.
message UserDefinedDataStreamFunction {
  enum FunctionType {
    MAP = 0;
    FLAT_MAP = 1;
    REDUCE = 2;
    CO_MAP = 3;
    CO_FLAT_MAP = 4;
    PROCESS = 5;
    KEYED_PROCESS = 6;
    TIMESTAMP_ASSIGNER = 7;
  }

  message JobParameter {
    string key = 1;
    string value = 2;
  }

  message RuntimeContext {
    string task_name = 1;
    string task_name_with_subtasks = 2;
    int32 number_of_parallel_subtasks = 3;
    int32 max_number_of_parallel_subtasks = 4;
    int32 index_of_this_subtask = 5;
    int32 attempt_number = 6;
    repeated JobParameter job_parameters = 7;
  }

  FunctionType function_type = 1;
  RuntimeContext runtime_context = 2;
  bytes payload = 3;
  bool metric_enabled = 4;
  TypeInfo key_type_info = 5;
}

message UserDefinedAggregateFunction {
  message DataViewSpec {
    message ListView {
      Schema.FieldType element_type = 1;
    }
    message MapView {
      Schema.FieldType key_type = 1;
      Schema.FieldType value_type = 2;
    }
    string name = 1;
    int32 field_index = 2;
    oneof data_view {
      ListView list_view = 3;
      MapView map_view = 4;
    }
  }

  // The serialized representation of the user-defined function
  bytes payload = 1;

  // The input arguments of the user-defined function, it could be one of the following:
  // 1. A column from the input row
  // 2. The result of another user-defined function
  // 3. The constant value of the column
  repeated Input inputs = 2;

  repeated DataViewSpec specs = 3;

  int32 filter_arg = 4;

  bool distinct = 5;
}

// A list of the user-defined aggregate functions to be executed in a group aggregate operation.
message UserDefinedAggregateFunctions {
  repeated UserDefinedAggregateFunction udfs = 1;
  bool metric_enabled = 2;
  // The array of the key indexes.
  repeated int32 grouping = 3;
  // Generate retract messages if true.
  bool generate_update_before = 4;
  // The schema of the grouping keys.
  Schema.FieldType key_type = 5;
  // The index of a count aggregate used to calculate the number of accumulated rows.
  int32 index_of_count_star = 6;
  // Cleanup the expired state if true.
  bool state_cleaning_enabled = 7;
  // The state cache size.
  int32 state_cache_size = 8;
  // The map state read cache size.
  int32 map_state_read_cache_size = 9;
  // The map_state_write_cache_size.
  int32 map_state_write_cache_size = 10;
  // True if the count(*) agg is inserted by the planner.
  bool count_star_inserted = 11;
}

// A representation of the data schema.
message Schema {
  enum TypeName {
    ROW = 0;
    TINYINT = 1;
    SMALLINT = 2;
    INT = 3;
    BIGINT = 4;
    DECIMAL = 5;
    FLOAT = 6;
    DOUBLE = 7;
    DATE = 8;
    TIME = 9;
    TIMESTAMP = 10;
    BOOLEAN = 11;
    BINARY = 12;
    VARBINARY = 13;
    CHAR = 14;
    VARCHAR = 15;
    BASIC_ARRAY = 16;
    MAP = 17;
    MULTISET = 18;
    LOCAL_ZONED_TIMESTAMP = 19;
    ZONED_TIMESTAMP = 20;
  }

  message MapInfo {
    FieldType key_type = 1;
    FieldType value_type = 2;
  }

  message TimeInfo {
    int32 precision = 1;
  }

  message TimestampInfo {
    int32 precision = 1;
  }

  message LocalZonedTimestampInfo {
    int32 precision = 1;
  }

  message ZonedTimestampInfo {
    int32 precision = 1;
  }

  message DecimalInfo {
    int32 precision = 1;
    int32 scale = 2;
  }

  message BinaryInfo {
    int32 length = 1;
  }

  message VarBinaryInfo {
    int32 length = 1;
  }

  message CharInfo {
    int32 length = 1;
  }

  message VarCharInfo {
    int32 length = 1;
  }

  message FieldType {
    TypeName type_name = 1;
    bool nullable = 2;
    oneof type_info {
      FieldType collection_element_type = 3;
      MapInfo map_info = 4;
      Schema row_schema = 5;
      DecimalInfo decimal_info = 6;
      TimeInfo time_info = 7;
      TimestampInfo timestamp_info = 8;
      LocalZonedTimestampInfo local_zoned_timestamp_info = 9;
      ZonedTimestampInfo zoned_timestamp_info = 10;
      BinaryInfo binary_info = 11;
      VarBinaryInfo var_binary_info = 12;
      CharInfo char_info = 13;
      VarCharInfo var_char_info = 14;
    }
  }

  message Field {
    string name = 1;
    string description = 2;
    FieldType type = 3;
  }

  repeated Field fields = 1;
}

// A representation of the data type information in DataStream.
message TypeInfo {
  enum TypeName {
    ROW = 0;
    STRING = 1;
    BYTE = 2;
    BOOLEAN = 3;
    SHORT = 4;
    INT = 5;
    LONG = 6;
    FLOAT = 7;
    DOUBLE = 8;
    CHAR = 9;
    BIG_INT = 10;
    BIG_DEC = 11;
    SQL_DATE = 12;
    SQL_TIME = 13;
    SQL_TIMESTAMP = 14;
    BASIC_ARRAY = 15;
    PICKLED_BYTES = 16;
    TUPLE = 17;
    PRIMITIVE_ARRAY = 18;
  }


  message FieldType {
    TypeName type_name = 1;
    oneof type_info {
      FieldType collection_element_type = 2;
      TypeInfo row_type_info = 3;
      TypeInfo tuple_type_info = 4;
    }
  }

  message Field {
    string name = 1;
    //description is used in Table schema, no need in DataStream.
    string description = 2;
    FieldType type = 3;
  }
  repeated Field field = 1;
}