karafka-rdkafka 0.13.8 → 0.13.10

data/lib/rdkafka/config.rb

@@ -1,11 +1,9 @@
 # frozen_string_literal: true
 
-require "logger"
-
 module Rdkafka
   # Configuration for a Kafka consumer or producer. You can create an instance and use
   # the consumer and producer methods to create a client. Documentation of the available
-  # configuration options is available on https://github.com/edenhill/librdkafka/blob/master/CONFIGURATION.md.
+  # configuration options is available on https://github.com/confluentinc/librdkafka/blob/master/CONFIGURATION.md.
   class Config
     # @private
     @@logger = Logger.new(STDOUT)
@@ -53,13 +51,13 @@ module Rdkafka
 
     # Set a callback that will be called every time the underlying client emits statistics.
     # You can configure if and how often this happens using `statistics.interval.ms`.
-    # The callback is called with a hash that's documented here: https://github.com/edenhill/librdkafka/blob/master/STATISTICS.md
+    # The callback is called with a hash that's documented here: https://github.com/confluentinc/librdkafka/blob/master/STATISTICS.md
     #
     # @param callback [Proc, #call] The callback
     #
     # @return [nil]
     def self.statistics_callback=(callback)
-      raise TypeError.new("Callback has to be callable") unless callback.respond_to?(:call)
+      raise TypeError.new("Callback has to be callable") unless callback.respond_to?(:call) || callback == nil
       @@statistics_callback = callback
     end
 
@@ -114,6 +112,7 @@ module Rdkafka
     def initialize(config_hash = {})
       @config_hash = DEFAULT_CONFIG.merge(config_hash)
       @consumer_rebalance_listener = nil
+      @consumer_poll_set = true
     end
 
     # Set a config option.
@@ -142,12 +141,28 @@ module Rdkafka
       @consumer_rebalance_listener = listener
     end
 
-    # Create a consumer with this configuration.
+    # Should we use a single queue for the underlying consumer and events.
     #
-    # @raise [ConfigError] When the configuration contains invalid options
-    # @raise [ClientCreationError] When the native client cannot be created
+    # This is an advanced API that allows for more granular control of the polling process.
+    # When this value is set to `false` (`true` by defualt), there will be two queues that need to
+    # be polled:
+    #   - main librdkafka queue for events
+    #   - consumer queue with messages and rebalances
+    #
+    # It is recommended to use the defaults and only set it to `false` in advance multi-threaded
+    # and complex cases where granular events handling control is needed.
+    #
+    # @param poll_set [Boolean]
+    def consumer_poll_set=(poll_set)
+      @consumer_poll_set = poll_set
+    end
+
+    # Creates a consumer with this configuration.
     #
     # @return [Consumer] The created consumer
+    #
+    # @raise [ConfigError] When the configuration contains invalid options
+    # @raise [ClientCreationError] When the native client cannot be created
     def consumer
       opaque = Opaque.new
       config = native_config(opaque)
@@ -160,8 +175,8 @@ module Rdkafka
       # Create native client
       kafka = native_kafka(config, :rd_kafka_consumer)
 
-      # Redirect the main queue to the consumer
-      Rdkafka::Bindings.rd_kafka_poll_set_consumer(kafka)
+      # Redirect the main queue to the consumer queue
+      Rdkafka::Bindings.rd_kafka_poll_set_consumer(kafka) if @consumer_poll_set
 
       # Return consumer with Kafka client
       Rdkafka::Consumer.new(
@@ -175,10 +190,10 @@ module Rdkafka
 
     # Create a producer with this configuration.
     #
+    # @return [Producer] The created producer
+    #
     # @raise [ConfigError] When the configuration contains invalid options
     # @raise [ClientCreationError] When the native client cannot be created
-    #
-    # @return [Producer] The created producer
     def producer
       # Create opaque
       opaque = Opaque.new
@@ -200,12 +215,12 @@ module Rdkafka
       end
     end
 
-    # Create an admin instance with this configuration.
+    # Creates an admin instance with this configuration.
+    #
+    # @return [Admin] The created admin instance
     #
     # @raise [ConfigError] When the configuration contains invalid options
     # @raise [ClientCreationError] When the native client cannot be created
-    #
-    # @return [Admin] The created admin instance
     def admin
       opaque = Opaque.new
       config = native_config(opaque)

data/lib/rdkafka/consumer/headers.rb

@@ -18,13 +18,11 @@ module Rdkafka
 
       # Reads a librdkafka native message's headers and returns them as a Ruby Hash
       #
-      # @param [librdkakfa message] native_message
+      # @private
       #
+      # @param [librdkakfa message] native_message
       # @return [Hash<String, String>] headers Hash for the native_message
-      #
       # @raise [Rdkafka::RdkafkaError] when fail to read headers
-      #
-      # @private
       def self.from_native(native_message)
         headers_ptrptr = FFI::MemoryPointer.new(:pointer)
         err = Rdkafka::Bindings.rd_kafka_message_headers(native_message, headers_ptrptr)

data/lib/rdkafka/consumer.rb

@@ -12,6 +12,7 @@ module Rdkafka
   # `each_slice` to consume batches of messages.
   class Consumer
     include Enumerable
+    include Helpers::Time
 
     # @private
     def initialize(native_kafka)
@@ -54,13 +55,11 @@ module Rdkafka
       @native_kafka.closed?
     end
 
-    # Subscribe to one or more topics letting Kafka handle partition assignments.
+    # Subscribes to one or more topics letting Kafka handle partition assignments.
     #
     # @param topics [Array<String>] One or more topic names
-    #
-    # @raise [RdkafkaError] When subscribing fails
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When subscribing fails
     def subscribe(*topics)
       closed_consumer_check(__method__)
 
@@ -83,9 +82,8 @@ module Rdkafka
 
     # Unsubscribe from all subscribed topics.
     #
-    # @raise [RdkafkaError] When unsubscribing fails
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When unsubscribing fails
     def unsubscribe
       closed_consumer_check(__method__)
 
@@ -101,10 +99,8 @@ module Rdkafka
     # Pause producing or consumption for the provided list of partitions
     #
     # @param list [TopicPartitionList] The topic with partitions to pause
-    #
-    # @raise [RdkafkaTopicPartitionListError] When pausing subscription fails.
-    #
     # @return [nil]
+    # @raise [RdkafkaTopicPartitionListError] When pausing subscription fails.
     def pause(list)
       closed_consumer_check(__method__)
 
@@ -128,13 +124,11 @@ module Rdkafka
       end
     end
 
-    # Resume producing consumption for the provided list of partitions
+    # Resumes producing consumption for the provided list of partitions
     #
     # @param list [TopicPartitionList] The topic with partitions to pause
-    #
-    # @raise [RdkafkaError] When resume subscription fails.
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When resume subscription fails.
     def resume(list)
       closed_consumer_check(__method__)
 
@@ -157,11 +151,10 @@ module Rdkafka
       end
     end
 
-    # Return the current subscription to topics and partitions
-    #
-    # @raise [RdkafkaError] When getting the subscription fails.
+    # Returns the current subscription to topics and partitions
     #
     # @return [TopicPartitionList]
+    # @raise [RdkafkaError] When getting the subscription fails.
     def subscription
       closed_consumer_check(__method__)
 
@@ -184,7 +177,6 @@ module Rdkafka
     # Atomic assignment of partitions to consume
     #
     # @param list [TopicPartitionList] The topic with partitions to assign
-    #
     # @raise [RdkafkaError] When assigning fails
     def assign(list)
       closed_consumer_check(__method__)
@@ -208,9 +200,8 @@ module Rdkafka
 
     # Returns the current partition assignment.
     #
-    # @raise [RdkafkaError] When getting the assignment fails.
-    #
     # @return [TopicPartitionList]
+    # @raise [RdkafkaError] When getting the assignment fails.
     def assignment
       closed_consumer_check(__method__)
 
@@ -244,14 +235,14 @@ module Rdkafka
     end
 
     # Return the current committed offset per partition for this consumer group.
-    # The offset field of each requested partition will either be set to stored offset or to -1001 in case there was no stored offset for that partition.
+    # The offset field of each requested partition will either be set to stored offset or to -1001
+    # in case there was no stored offset for that partition.
     #
-    # @param list [TopicPartitionList, nil] The topic with partitions to get the offsets for or nil to use the current subscription.
+    # @param list [TopicPartitionList, nil] The topic with partitions to get the offsets for or nil
+    #   to use the current subscription.
     # @param timeout_ms [Integer] The timeout for fetching this information.
-    #
-    # @raise [RdkafkaError] When getting the committed positions fails.
-    #
     # @return [TopicPartitionList]
+    # @raise [RdkafkaError] When getting the committed positions fails.
     def committed(list=nil, timeout_ms=1200)
       closed_consumer_check(__method__)
 
@@ -307,10 +298,8 @@ module Rdkafka
     # @param topic [String] The topic to query
     # @param partition [Integer] The partition to query
     # @param timeout_ms [Integer] The timeout for querying the broker
-    #
-    # @raise [RdkafkaError] When querying the broker fails.
-    #
     # @return [Integer] The low and high watermark
+    # @raise [RdkafkaError] When querying the broker fails.
     def query_watermark_offsets(topic, partition, timeout_ms=200)
       closed_consumer_check(__method__)
 
@@ -343,10 +332,9 @@ module Rdkafka
     #
     # @param topic_partition_list [TopicPartitionList] The list to calculate lag for.
     # @param watermark_timeout_ms [Integer] The timeout for each query watermark call.
-    #
+    # @return [Hash<String, Hash<Integer, Integer>>] A hash containing all topics with the lag
+    #   per partition
     # @raise [RdkafkaError] When querying the broker fails.
-    #
-    # @return [Hash<String, Hash<Integer, Integer>>] A hash containing all topics with the lag per partition
     def lag(topic_partition_list, watermark_timeout_ms=100)
       out = {}
 
@@ -395,10 +383,8 @@ module Rdkafka
     # When using this `enable.auto.offset.store` should be set to `false` in the config.
     #
     # @param message [Rdkafka::Consumer::Message] The message which offset will be stored
-    #
-    # @raise [RdkafkaError] When storing the offset fails
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When storing the offset fails
     def store_offset(message)
       closed_consumer_check(__method__)
 
@@ -430,10 +416,8 @@ module Rdkafka
     # message at the given offset.
     #
     # @param message [Rdkafka::Consumer::Message] The message to which to seek
-    #
-    # @raise [RdkafkaError] When seeking fails
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When seeking fails
     def seek(message)
       closed_consumer_check(__method__)
 
@@ -503,10 +487,8 @@ module Rdkafka
     #
     # @param list [TopicPartitionList,nil] The topic with partitions to commit
     # @param async [Boolean] Whether to commit async or wait for the commit to finish
-    #
-    # @raise [RdkafkaError] When committing fails
-    #
     # @return [nil]
+    # @raise [RdkafkaError] When committing fails
     def commit(list=nil, async=false)
       closed_consumer_check(__method__)
 
@@ -532,10 +514,8 @@ module Rdkafka
     # Poll for the next message on one of the subscribed topics
     #
     # @param timeout_ms [Integer] Timeout of this poll
-    #
-    # @raise [RdkafkaError] When polling fails
-    #
     # @return [Message, nil] A message or nil if there was no new message within the timeout
+    # @raise [RdkafkaError] When polling fails
     def poll(timeout_ms)
       closed_consumer_check(__method__)
 
@@ -561,17 +541,40 @@ module Rdkafka
       end
     end
 
+    # Polls the main rdkafka queue (not the consumer one). Do **NOT** use it if `consumer_poll_set`
+    #   was set to `true`.
+    #
+    # Events will cause application-provided callbacks to be called.
+    #
+    # Events (in the context of the consumer):
+    #   - error callbacks
+    #   - stats callbacks
+    #   - any other callbacks supported by librdkafka that are not part of the consumer_poll, that
+    #     would have a callback configured and activated.
+    #
+    # This method needs to be called at regular intervals to serve any queued callbacks waiting to
+    # be called. When in use, does **NOT** replace `#poll` but needs to run complementary with it.
+    #
+    # @param timeout_ms [Integer] poll timeout. If set to 0 will run async, when set to -1 will
+    #   block until any events available.
+    #
+    # @note This method technically should be called `#poll` and the current `#poll` should be
+    #   called `#consumer_poll` though we keep the current naming convention to make it backward
+    #   compatible.
+    def events_poll(timeout_ms = 0)
+      @native_kafka.with_inner do |inner|
+        Rdkafka::Bindings.rd_kafka_poll(inner, timeout_ms)
+      end
+    end
+
     # Poll for new messages and yield for each received one. Iteration
     # will end when the consumer is closed.
     #
-    # If `enable.partition.eof` is turned on in the config this will raise an
-    # error when an eof is reached, so you probably want to disable that when
-    # using this method of iteration.
+    # If `enable.partition.eof` is turned on in the config this will raise an error when an eof is
+    # reached, so you probably want to disable that when using this method of iteration.
     #
     # @raise [RdkafkaError] When polling fails
-    #
     # @yieldparam message [Message] Received message
-    #
     # @return [nil]
     def each
       loop do
@@ -624,9 +627,7 @@ module Rdkafka
     # that you may or may not see again.
     #
     # @param max_items [Integer] Maximum size of the yielded array of messages
-    #
     # @param bytes_threshold [Integer] Threshold number of total message bytes in the yielded array of messages
-    #
     # @param timeout_ms [Integer] max time to wait for up to max_items
     #
     # @raise [RdkafkaError] When polling fails
@@ -673,10 +674,6 @@ module Rdkafka
     end
 
     private
-    def monotonic_now
-      # needed because Time.now can go backwards
-      Process.clock_gettime(Process::CLOCK_MONOTONIC)
-    end
 
     def closed_consumer_check(method)
       raise Rdkafka::ClosedConsumerError.new(method) if closed?

data/lib/rdkafka/helpers/time.rb

@@ -0,0 +1,14 @@
+# frozen_string_literal: true
+
+module Rdkafka
+  # Namespace for some small utilities used in multiple components
+  module Helpers
+    # Time related methods used across Karafka
+    module Time
+      # @return [Float] current monotonic time in seconds with microsecond precision
+      def monotonic_now
+        ::Process.clock_gettime(::Process::CLOCK_MONOTONIC)
+      end
+    end
+  end
+end

data/lib/rdkafka/producer.rb

@@ -1,10 +1,10 @@
 # frozen_string_literal: true
 
-require "objspace"
-
 module Rdkafka
   # A producer for Kafka messages. To create a producer set up a {Config} and call {Config#producer producer} on that.
   class Producer
+    include Helpers::Time
+
     # Cache partitions count for 30 seconds
     PARTITIONS_COUNT_TTL = 30
 
@@ -167,18 +167,16 @@ module Rdkafka
     end
 
     # Partition count for a given topic.
-    # NOTE: If 'allow.auto.create.topics' is set to true in the broker, the topic will be auto-created after returning nil.
     #
     # @param topic [String] The topic name.
+    # @return [Integer] partition count for a given topic
     #
-    # @return partition count [Integer,nil]
-    #
-    # We cache the partition count for a given topic for given time
-    # This prevents us in case someone uses `partition_key` from querying for the count with
-    # each message. Instead we query once every 30 seconds at most
+    # @note If 'allow.auto.create.topics' is set to true in the broker, the topic will be
+    #   auto-created after returning nil.
     #
-    # @param topic [String] topic name
-    # @return [Integer] partition count for a given topic
+    # @note We cache the partition count for a given topic for given time.
+    #   This prevents us in case someone uses `partition_key` from querying for the count with
+    #   each message. Instead we query once every 30 seconds at most
     def partition_count(topic)
       closed_producer_check(__method__)
 
@@ -308,11 +306,6 @@ module Rdkafka
 
     private
 
-    def monotonic_now
-      # needed because Time.now can go backwards
-      Process.clock_gettime(Process::CLOCK_MONOTONIC)
-    end
-
     def closed_producer_check(method)
       raise Rdkafka::ClosedProducerError.new(method) if closed?
     end

data/lib/rdkafka/version.rb

@@ -1,7 +1,7 @@
 # frozen_string_literal: true
 
 module Rdkafka
-  VERSION = "0.13.8"
+  VERSION = "0.13.10"
   LIBRDKAFKA_VERSION = "2.2.0"
   LIBRDKAFKA_SOURCE_SHA256 = "af9a820cbecbc64115629471df7c7cecd40403b6c34bfdbb9223152677a47226"
 end

data/lib/rdkafka.rb

@@ -1,7 +1,12 @@
 # frozen_string_literal: true
 
-require "rdkafka/version"
+require "logger"
+require "objspace"
+require "ffi"
+require "json"
 
+require "rdkafka/version"
+require "rdkafka/helpers/time"
 require "rdkafka/abstract_handle"
 require "rdkafka/admin"
 require "rdkafka/admin/create_topic_handle"
@@ -24,3 +29,7 @@ require "rdkafka/native_kafka"
 require "rdkafka/producer"
 require "rdkafka/producer/delivery_handle"
 require "rdkafka/producer/delivery_report"
+
+# Main Rdkafka namespace of this gem
+module Rdkafka
+end

data/spec/rdkafka/abstract_handle_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::AbstractHandle do
   let(:response) { 0 }
   let(:result) { -1 }

data/spec/rdkafka/admin/create_topic_handle_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Admin::CreateTopicHandle do
   let(:response) { 0 }
 

data/spec/rdkafka/admin/create_topic_report_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Admin::CreateTopicReport do
   subject { Rdkafka::Admin::CreateTopicReport.new(
       FFI::MemoryPointer.from_string("error string"),

data/spec/rdkafka/admin/delete_topic_handle_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Admin::DeleteTopicHandle do
   let(:response) { 0 }
 

data/spec/rdkafka/admin/delete_topic_report_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Admin::DeleteTopicReport do
   subject { Rdkafka::Admin::DeleteTopicReport.new(
       FFI::MemoryPointer.from_string("error string"),

data/spec/rdkafka/admin_spec.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
 require "ostruct"
 
 describe Rdkafka::Admin do

data/spec/rdkafka/bindings_spec.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
 require 'zlib'
 
 describe Rdkafka::Bindings do

data/spec/rdkafka/callbacks_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Callbacks do
 
   # The code in the call back functions is 100% covered by other specs.  Due to

data/spec/rdkafka/config_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Config do
   context "logger" do
     it "should have a default logger" do
@@ -115,6 +113,14 @@ describe Rdkafka::Config do
       consumer.close
     end
 
+    it "should create a consumer with consumer_poll_set set to false" do
+      config = rdkafka_consumer_config
+      config.consumer_poll_set = false
+      consumer = config.consumer
+      expect(consumer).to be_a Rdkafka::Consumer
+      consumer.close
+    end
+
     it "should raise an error when creating a consumer with invalid config" do
       config = Rdkafka::Config.new('invalid.key' => 'value')
       expect {

data/spec/rdkafka/consumer/headers_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Consumer::Headers do
   let(:headers) do
     { # Note String keys!

data/spec/rdkafka/consumer/message_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Consumer::Message do
   let(:native_client) { new_native_client }
   let(:native_topic) { new_native_topic(native_client: native_client) }

data/spec/rdkafka/consumer/partition_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Consumer::Partition do
   let(:offset) { 100 }
   let(:err) { 0 }

data/spec/rdkafka/consumer/topic_partition_list_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::Consumer::TopicPartitionList do
   it "should create a new list and add unassigned topics" do
     list = Rdkafka::Consumer::TopicPartitionList.new

data/spec/rdkafka/consumer_spec.rb

@@ -1,6 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
 require "ostruct"
 require 'securerandom'
 
@@ -55,6 +54,30 @@ describe Rdkafka::Consumer do
         consumer.subscription
       }.to raise_error(Rdkafka::RdkafkaError)
     end
+
+    context "when using consumer without the poll set" do
+      let(:consumer) do
+        config = rdkafka_consumer_config
+        config.consumer_poll_set = false
+        config.consumer
+      end
+
+      it "should subscribe, unsubscribe and return the subscription" do
+        expect(consumer.subscription).to be_empty
+
+        consumer.subscribe("consume_test_topic")
+
+        expect(consumer.subscription).not_to be_empty
+        expected_subscription = Rdkafka::Consumer::TopicPartitionList.new.tap do |list|
+          list.add_topic("consume_test_topic")
+        end
+        expect(consumer.subscription).to eq expected_subscription
+
+        consumer.unsubscribe
+
+        expect(consumer.subscription).to be_empty
+      end
+    end
   end
 
   describe "#pause and #resume" do
@@ -1077,6 +1100,29 @@ describe Rdkafka::Consumer do
     end
   end
 
+  # Only relevant in case of a consumer with separate queues
+  describe '#events_poll' do
+    let(:stats) { [] }
+
+    before { Rdkafka::Config.statistics_callback = ->(published) { stats << published } }
+
+    after { Rdkafka::Config.statistics_callback = nil }
+
+    let(:consumer) do
+      config = rdkafka_consumer_config('statistics.interval.ms': 100)
+      config.consumer_poll_set = false
+      config.consumer
+    end
+
+    it "expect to run events_poll, operate and propagate stats on events_poll and not poll" do
+      consumer.subscribe("consume_test_topic")
+      consumer.poll(1_000)
+      expect(stats).to be_empty
+      consumer.events_poll(-1)
+      expect(stats).not_to be_empty
+    end
+  end
+
   describe "a rebalance listener" do
     let(:consumer) do
       config = rdkafka_consumer_config

data/spec/rdkafka/error_spec.rb

@@ -1,7 +1,5 @@
 # frozen_string_literal: true
 
-require "spec_helper"
-
 describe Rdkafka::RdkafkaError do
   it "should raise a type error for a nil response" do
     expect {