fs2.kafka.consumer.KafkaAssignment.scala Maven / Gradle / Ivy
The newest version!
/*
* Copyright 2018-2024 OVO Energy Limited
*
* SPDX-License-Identifier: Apache-2.0
*/
package fs2.kafka.consumer
import scala.collection.immutable.SortedSet
import cats.data.NonEmptySet
import fs2.*
import fs2.kafka.instances.*
import org.apache.kafka.common.TopicPartition
trait KafkaAssignment[F[_]] {
/**
* Returns the set of partitions currently assigned to this consumer.
*/
def assignment: F[SortedSet[TopicPartition]]
/**
* `Stream` where the elements are the set of `TopicPartition`s currently assigned to this
* consumer. The stream emits whenever a rebalance changes partition assignments.
*/
def assignmentStream: Stream[F, SortedSet[TopicPartition]]
/**
* Manually assigns the specified list of topic partitions to the consumer. This function does
* not allow for incremental assignment and will replace the previous assignment (if there is
* one).
*
* Manual topic assignment through this method does not use the consumer's group management
* functionality. As such, there will be no rebalance operation triggered when group membership
* or cluster and topic metadata change. Note that it is not possible to use both manual
* partition assignment with `assign` and group assignment with `subscribe`.
*
* If auto-commit is enabled, an async commit (based on the old assignment) will be triggered
* before the new assignment replaces the old one.
*
* To unassign all partitions, use [[KafkaConsumer#unsubscribe]].
*
* @see
* org.apache.kafka.clients.consumer.KafkaConsumer#assign
*/
def assign(partitions: NonEmptySet[TopicPartition]): F[Unit]
/**
* Manually assigns the specified list of partitions for the specified topic to the consumer.
* This function does not allow for incremental assignment and will replace the previous
* assignment (if there is one).
*
* Manual topic assignment through this method does not use the consumer's group management
* functionality. As such, there will be no rebalance operation triggered when group membership
* or cluster and topic metadata change. Note that it is not possible to use both manual
* partition assignment with `assign` and group assignment with `subscribe`.
*
* If auto-commit is enabled, an async commit (based on the old assignment) will be triggered
* before the new assignment replaces the old one.
*
* To unassign all partitions, use [[KafkaConsumer#unsubscribe]].
*
* @see
* org.apache.kafka.clients.consumer.KafkaConsumer#assign
*/
def assign(topic: String, partitions: NonEmptySet[Int]): F[Unit] =
assign(partitions.map(new TopicPartition(topic, _)))
/**
* Manually assigns all partitions for the specified topic to the consumer.
*/
def assign(topic: String): F[Unit]
}