ruby-kafka 1.1.0 → 1.2.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 22b958f3d452f35c3c6084a5b8b790f370b96069e87ef5974838e5ea4b6945e1
-  data.tar.gz: eb6d704dbaace9c13a99bc93924d8a6b84ee5eb686c9c863a81d085ffeb7e92d
+  metadata.gz: 0f8391cc7b1989cb5f669796bc4ad647b77d882e6506fae42bab18acb8a6bcc6
+  data.tar.gz: 012baaff5d2cc9eb17e3a7b7342f49f7c905a5f91d26078fa0ecf2f0fa81a2ad
 SHA512:
-  metadata.gz: 22c1d59bcdd42849849122b559f0d161653a5cfa492ffb36f28a875b109444d670a76bc62ba0856ae2ebce95fd10abce9e16834928fc7fe6eb2ee006827d307c
-  data.tar.gz: 3a51df6b1d40e1edbd96c06f3473319f0ac1b072b040b676a8e1d980fbaad114ad0248a7670cb7ae0a51a2c79ac390f2f4cfe8f65eedb5dd55be020c6011bc18
+  metadata.gz: 7f4e9302ca0ab41a6fded75f95ed866d959a2a027b90051ed7f3d7fba573aa63e57be7692d004dbeb3bacd99fe44b24188f81b0f3bed50f68e1da5189262271f
+  data.tar.gz: 7266bdd50e66a7ab9b3c71025468af0a4bea778fd312c41d4c699e2609420c11c28accb3246407ce98330bf521c109691bd1d4943dd532a44ee6732f1a410922

data/.ruby-version

@@ -1 +1 @@
-2.5.1
+2.7.1

data/CHANGELOG.md

@@ -4,6 +4,11 @@ Changes and additions to the library will be listed here.
 
 ## Unreleased
 
+## 1.2.0
+
+- Add producer consumer interceptors (#837).
+- Add support for configuring the client partitioner (#848).
+
 ## 1.1.0
 
 - Extra sanity checking when marking offsets as processed (#824).

data/README.md

@@ -349,6 +349,26 @@ partition = PartitioningScheme.assign(partitions, event)
 producer.produce(event, topic: "events", partition: partition)
 ```
 
+Another option is to configure a custom client partitioner that implements `call(partition_count, message)` and uses the same schema as the other client. For example:
+
+```ruby
+class CustomPartitioner
+  def call(partition_count, message)
+    ...
+  end
+end
+  
+partitioner = CustomPartitioner.new
+Kafka.new(partitioner: partitioner, ...)
+```
+
+Or, simply create a Proc handling the partitioning logic instead of having to add a new class. For example:
+
+```ruby
+partitioner = -> (partition_count, message) { ... }
+Kafka.new(partitioner: partitioner, ...)
+```
+
 #### Buffering and Error Handling
 
 The producer is designed for resilience in the face of temporary network errors, Kafka broker failovers, and other issues that prevent the client from writing messages to the destination topics. It does this by employing local, in-memory buffers. Only when messages are acknowledged by a Kafka broker will they be removed from the buffer.

data/lib/kafka/async_producer.rb

@@ -246,10 +246,10 @@ module Kafka
 
       private
 
-      def produce(*args)
+      def produce(value, **kwargs)
         retries = 0
         begin
-          @producer.produce(*args)
+          @producer.produce(value, **kwargs)
         rescue BufferOverflow => e
           deliver_messages
           if @max_retries == -1

data/lib/kafka/client.rb

@@ -62,10 +62,15 @@ module Kafka
     #
     # @param sasl_over_ssl [Boolean] whether to enforce SSL with SASL
     #
+    # @param ssl_ca_certs_from_system [Boolean] whether to use the CA certs from the
+    #   system's default certificate store.
+    #
+    # @param partitioner [Partitioner, nil] the partitioner that should be used by the client.
+    #
     # @param sasl_oauth_token_provider [Object, nil] OAuthBearer Token Provider instance that
     #   implements method token. See {Sasl::OAuth#initialize}
     #
-    # @param verify_hostname [Boolean, true] whether to verify that the host serving
+    # @param ssl_verify_hostname [Boolean, true] whether to verify that the host serving
     #   the SSL certificate and the signing chain of the certificate have the correct domains
     #   based on the CA certificate
     #
@@ -75,7 +80,7 @@ module Kafka
                    ssl_client_cert_key_password: nil, ssl_client_cert_chain: nil, sasl_gssapi_principal: nil,
                    sasl_gssapi_keytab: nil, sasl_plain_authzid: '', sasl_plain_username: nil, sasl_plain_password: nil,
                    sasl_scram_username: nil, sasl_scram_password: nil, sasl_scram_mechanism: nil,
-                   sasl_over_ssl: true, ssl_ca_certs_from_system: false, sasl_oauth_token_provider: nil, ssl_verify_hostname: true)
+                   sasl_over_ssl: true, ssl_ca_certs_from_system: false, partitioner: nil, sasl_oauth_token_provider: nil, ssl_verify_hostname: true)
       @logger = TaggedLogger.new(logger)
       @instrumenter = Instrumenter.new(client_id: client_id)
       @seed_brokers = normalize_seed_brokers(seed_brokers)
@@ -119,6 +124,7 @@ module Kafka
       )
 
       @cluster = initialize_cluster
+      @partitioner = partitioner || Partitioner.new
     end
 
     # Delivers a single message to the Kafka cluster.
@@ -157,7 +163,7 @@ module Kafka
 
       if partition.nil?
         partition_count = @cluster.partitions_for(topic).count
-        partition = Partitioner.partition_for_key(partition_count, message)
+        partition = @partitioner.call(partition_count, message)
       end
 
       buffer = MessageBuffer.new
@@ -248,6 +254,9 @@ module Kafka
     #   be in a message set before it should be compressed. Note that message sets
     #   are per-partition rather than per-topic or per-producer.
     #
+    # @param interceptors [Array<Object>] a list of producer interceptors the implement
+    #   `call(Kafka::PendingMessage)`.
+    #
     # @return [Kafka::Producer] the Kafka producer.
     def producer(
       compression_codec: nil,
@@ -261,7 +270,8 @@ module Kafka
       idempotent: false,
       transactional: false,
       transactional_id: nil,
-      transactional_timeout: 60
+      transactional_timeout: 60,
+      interceptors: []
     )
       cluster = initialize_cluster
       compressor = Compressor.new(
@@ -291,6 +301,8 @@ module Kafka
         retry_backoff: retry_backoff,
         max_buffer_size: max_buffer_size,
         max_buffer_bytesize: max_buffer_bytesize,
+        partitioner: @partitioner,
+        interceptors: interceptors
       )
     end
 
@@ -343,6 +355,8 @@ module Kafka
     # @param refresh_topic_interval [Integer] interval of refreshing the topic list.
     #   If it is 0, the topic list won't be refreshed (default)
     #   If it is n (n > 0), the topic list will be refreshed every n seconds
+    # @param interceptors [Array<Object>] a list of consumer interceptors that implement
+    #   `call(Kafka::FetchedBatch)`.
     # @return [Consumer]
     def consumer(
         group_id:,
@@ -353,7 +367,8 @@ module Kafka
         heartbeat_interval: 10,
         offset_retention_time: nil,
         fetcher_max_queue_size: 100,
-        refresh_topic_interval: 0
+        refresh_topic_interval: 0,
+        interceptors: []
     )
       cluster = initialize_cluster
 
@@ -407,7 +422,8 @@ module Kafka
         fetcher: fetcher,
         session_timeout: session_timeout,
         heartbeat: heartbeat,
-        refresh_topic_interval: refresh_topic_interval
+        refresh_topic_interval: refresh_topic_interval,
+        interceptors: interceptors
       )
     end
 

data/lib/kafka/consumer.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require "kafka/consumer_group"
+require "kafka/interceptors"
 require "kafka/offset_manager"
 require "kafka/fetcher"
 require "kafka/pause"
@@ -44,7 +45,8 @@ module Kafka
   #
   class Consumer
 
-    def initialize(cluster:, logger:, instrumenter:, group:, fetcher:, offset_manager:, session_timeout:, heartbeat:, refresh_topic_interval: 0)
+    def initialize(cluster:, logger:, instrumenter:, group:, fetcher:, offset_manager:,
+                   session_timeout:, heartbeat:, refresh_topic_interval: 0, interceptors: [])
       @cluster = cluster
       @logger = TaggedLogger.new(logger)
       @instrumenter = instrumenter
@@ -54,6 +56,7 @@ module Kafka
       @fetcher = fetcher
       @heartbeat = heartbeat
       @refresh_topic_interval = refresh_topic_interval
+      @interceptors = Interceptors.new(interceptors: interceptors, logger: logger)
 
       @pauses = Hash.new {|h, k|
         h[k] = Hash.new {|h2, k2|
@@ -76,7 +79,7 @@ module Kafka
       @current_offsets = Hash.new { |h, k| h[k] = {} }
 
       # Map storing subscribed topics with their configuration
-      @subscribed_topics = Concurrent::Map.new
+      @subscribed_topics = Hash.new
 
       # Set storing topics that matched topics in @subscribed_topics
       @matched_topics = Set.new
@@ -220,6 +223,7 @@ module Kafka
         batches = fetch_batches
 
         batches.each do |batch|
+          batch = @interceptors.call(batch)
           batch.messages.each do |message|
             notification = {
               topic: message.topic,
@@ -311,12 +315,13 @@ module Kafka
           unless batch.empty?
             raw_messages = batch.messages
             batch.messages = raw_messages.reject(&:is_control_record)
+            batch = @interceptors.call(batch)
 
             notification = {
               topic: batch.topic,
               partition: batch.partition,
               last_offset: batch.last_offset,
-              last_create_time: batch.messages.last.try(:create_time),
+              last_create_time: batch.messages.last && batch.messages.last.create_time,
               offset_lag: batch.offset_lag,
               highwater_mark_offset: batch.highwater_mark_offset,
               message_count: batch.messages.count,

data/lib/kafka/datadog.rb

@@ -96,8 +96,8 @@ module Kafka
       private
 
       %w[increment histogram count timing gauge].each do |type|
-        define_method(type) do |*args|
-          emit(type, *args)
+        define_method(type) do |*args, **kwargs|
+          emit(type, *args, **kwargs)
         end
       end
 

data/lib/kafka/interceptors.rb

@@ -0,0 +1,33 @@
+# frozen_string_literal: true
+
+module Kafka
+  # Holds a list of interceptors that implement `call`
+  # and wraps calls to a chain of custom interceptors.
+  class Interceptors
+    def initialize(interceptors:, logger:)
+      @interceptors = interceptors || []
+      @logger = TaggedLogger.new(logger)
+    end
+
+    # This method is called when the client produces a message or once the batches are fetched.
+    # The message returned from the first call is passed to the second interceptor call, and so on in an
+    # interceptor chain. This method does not throw exceptions.
+    #
+    # @param intercepted [Kafka::PendingMessage || Kafka::FetchedBatch] the produced message or
+    #   fetched batch.
+    #
+    # @return [Kafka::PendingMessage || Kafka::FetchedBatch] the intercepted message or batch
+    #   returned by the last interceptor.
+    def call(intercepted)
+      @interceptors.each do |interceptor|
+        begin
+          intercepted = interceptor.call(intercepted)
+        rescue Exception => e
+          @logger.warn "Error executing interceptor for topic: #{intercepted.topic} partition: #{intercepted.partition}: #{e.message}\n#{e.backtrace.join("\n")}"
+        end
+      end
+
+      intercepted
+    end
+  end
+end

data/lib/kafka/partitioner.rb

@@ -19,7 +19,7 @@ module Kafka
     # @param message [Kafka::PendingMessage] the message that should be assigned
     #   a partition.
     # @return [Integer] the partition number.
-    def self.partition_for_key(partition_count, message)
+    def call(partition_count, message)
       raise ArgumentError if partition_count == 0
 
       # If no explicit partition key is specified we use the message key instead.

data/lib/kafka/producer.rb

@@ -7,6 +7,7 @@ require "kafka/produce_operation"
 require "kafka/pending_message_queue"
 require "kafka/pending_message"
 require "kafka/compressor"
+require "kafka/interceptors"
 
 module Kafka
   # Allows sending messages to a Kafka cluster.
@@ -129,7 +130,9 @@ module Kafka
   class Producer
     class AbortTransaction < StandardError; end
 
-    def initialize(cluster:, transaction_manager:, logger:, instrumenter:, compressor:, ack_timeout:, required_acks:, max_retries:, retry_backoff:, max_buffer_size:, max_buffer_bytesize:)
+    def initialize(cluster:, transaction_manager:, logger:, instrumenter:, compressor:, ack_timeout:,
+                   required_acks:, max_retries:, retry_backoff:, max_buffer_size:,
+                   max_buffer_bytesize:, partitioner:, interceptors: [])
       @cluster = cluster
       @transaction_manager = transaction_manager
       @logger = TaggedLogger.new(logger)
@@ -141,6 +144,8 @@ module Kafka
       @max_buffer_size = max_buffer_size
       @max_buffer_bytesize = max_buffer_bytesize
       @compressor = compressor
+      @partitioner = partitioner
+      @interceptors = Interceptors.new(interceptors: interceptors, logger: logger)
 
       # The set of topics that are produced to.
       @target_topics = Set.new
@@ -191,7 +196,7 @@ module Kafka
       # We want to fail fast if `topic` isn't a String
       topic = topic.to_str
 
-      message = PendingMessage.new(
+      message = @interceptors.call(PendingMessage.new(
         value: value && value.to_s,
         key: key && key.to_s,
         headers: headers,
@@ -199,7 +204,7 @@ module Kafka
         partition: partition && Integer(partition),
         partition_key: partition_key && partition_key.to_s,
         create_time: create_time
-      )
+      ))
 
       if buffer_size >= @max_buffer_size
         buffer_overflow topic,
@@ -455,7 +460,7 @@ module Kafka
 
           if partition.nil?
             partition_count = @cluster.partitions_for(message.topic).count
-            partition = Partitioner.partition_for_key(partition_count, message)
+            partition = @partitioner.call(partition_count, message)
           end
 
           @buffer.write(

data/lib/kafka/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kafka
-  VERSION = "1.1.0"
+  VERSION = "1.2.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-kafka
 version: !ruby/object:Gem::Version
-  version: 1.1.0
+  version: 1.2.0
 platform: ruby
 authors:
 - Daniel Schierbeck
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2020-06-09 00:00:00.000000000 Z
+date: 2020-08-03 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: digest-crc
@@ -386,6 +386,7 @@ files:
 - lib/kafka/gzip_codec.rb
 - lib/kafka/heartbeat.rb
 - lib/kafka/instrumenter.rb
+- lib/kafka/interceptors.rb
 - lib/kafka/lz4_codec.rb
 - lib/kafka/message_buffer.rb
 - lib/kafka/offset_manager.rb