karafka 1.3.7 → 1.4.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: a4a06a7ae9c8de9ada4f029b4aa8c0e5946d127eed891c0676dc6cfa92d61aaa
-  data.tar.gz: 532c931751da662fe1b0aa7b4c5361effd710d7112a39fd094c93a3846bd04e9
+  metadata.gz: 7bae0f63af5f741dc00d382e34a0a5a35262692c9b03930dc18fd05440bf418b
+  data.tar.gz: f9a82e8ee1324b62820f24e8dddd09fdd1ec153c50a32f4f31f50a9192983341
 SHA512:
-  metadata.gz: 56e43f67ef345d2a9f528be97107fa84d1a5d13f343bcc43ae3af4bc7d1af85cf01f438dd0c050ea799c5c6c55c8b680d9b9f2c4c10d37a846cefa55ffcdebf3
-  data.tar.gz: ba50c3a5b047b1afabd578708fed5ea5cd7cdf875333e31e129752ec83992c6cc8eb0d4e601438846229b273da709e8b7e1a3fd777ad0e0919dde2c7f2349293
+  metadata.gz: a8fecf4be83990908c8aaf6b1e0f262391e9cdd65079362f76f68871d9ec3f45e4c0e9ceb1b277036de7f4c6ab2c140e52ee4ad855ec150c64626e9728a650a6
+  data.tar.gz: 2b626896aad5bc96bce14f0943a8d9c279dd877d2a221994f2ae56031d0bc97a33e5696b892bc8224f1c527a27e0861a79aaedd4b7716286528f6baf8f00de04

data/.ruby-version

@@ -1 +1 @@
-2.7.1
+2.7.2

data/CHANGELOG.md

@@ -1,5 +1,27 @@
 # Karafka framework changelog
 
+## 1.4.1 (2020-12-04)
+- Return non-zero exit code when printing usage
+- Add support for :assignment_strategy for consumers
+
+## 1.4.0 (2020-09-05)
+- Rename `Karafka::Params::Metadata` to `Karafka::Params::BatchMetadata`
+` Rename consumer `#metadata` to `#batch_metadata`
+- Separate metadata (including Karafka native metadata) from the root of params (backwards compatibility preserved thanks to rabotyaga)
+- Remove metadata hash dependency
+- Remove params dependency on a hash in favour of PORO
+- Remove batch metadata dependency on a hash
+- Remove MultiJson in favour of JSON in the default deserializer
+- allow accessing all the metadata without accessing the payload
+- freeze params and underlying elements except for the mutable payload
+- provide access to raw payload after serialization
+- fixes a bug where a non-deserializable (error) params would be marked as deserialized after first unsuccessful deserialization attempt
+- fixes bug where karafka would mutate internal ruby-kafka state
+- fixes bug where topic name in metadata would not be mapped using topic mappers
+- simplifies the params and params batch API, before `#payload` usage, it won't be deserialized
+- removes the `#[]` API from params to prevent from accessing raw data in a different way than #raw_payload
+- makes the params batch operations consistent as params payload is deserialized only when accessed explicitly
+
 ## 1.3.7 (2020-08-11)
 - #599 - Allow metadata access without deserialization attempt (rabotyaga)
 - Sync with ruby-kafka `1.2.0` api

data/Gemfile.lock

@@ -1,24 +1,23 @@
 PATH
   remote: .
   specs:
-    karafka (1.3.7)
+    karafka (1.4.1)
       dry-configurable (~> 0.8)
       dry-inflector (~> 0.1)
       dry-monitor (~> 0.3)
       dry-validation (~> 1.2)
       envlogic (~> 1.1)
       irb (~> 1.0)
-      multi_json (>= 1.12)
       rake (>= 11.3)
       ruby-kafka (>= 1.0.0)
       thor (>= 0.20)
-      waterdrop (~> 1.3.0)
+      waterdrop (~> 1.4.0)
       zeitwerk (~> 2.1)
 
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (6.0.3.2)
+    activesupport (6.0.3.4)
       concurrent-ruby (~> 1.0, >= 1.0.2)
       i18n (>= 0.7, < 2)
       minitest (~> 5.1)
@@ -40,7 +39,7 @@ GEM
     dry-container (0.7.2)
       concurrent-ruby (~> 1.0)
       dry-configurable (~> 0.1, >= 0.1.3)
-    dry-core (0.4.9)
+    dry-core (0.4.10)
       concurrent-ruby (~> 1.0)
     dry-equalizer (0.3.0)
     dry-events (0.2.0)
@@ -48,8 +47,8 @@ GEM
       dry-core (~> 0.4)
       dry-equalizer (~> 0.2)
     dry-inflector (0.2.0)
-    dry-initializer (3.0.3)
-    dry-logic (1.0.6)
+    dry-initializer (3.0.4)
+    dry-logic (1.0.8)
       concurrent-ruby (~> 1.0)
       dry-core (~> 0.2)
       dry-equalizer (~> 0.2)
@@ -58,7 +57,7 @@ GEM
       dry-core (~> 0.4)
       dry-equalizer (~> 0.2)
       dry-events (~> 0.2)
-    dry-schema (1.5.2)
+    dry-schema (1.5.6)
       concurrent-ruby (~> 1.0)
       dry-configurable (~> 0.8, >= 0.8.3)
       dry-core (~> 0.4)
@@ -73,13 +72,13 @@ GEM
       dry-equalizer (~> 0.3)
       dry-inflector (~> 0.1, >= 0.1.2)
       dry-logic (~> 1.0, >= 1.0.2)
-    dry-validation (1.5.3)
+    dry-validation (1.5.6)
       concurrent-ruby (~> 1.0)
       dry-container (~> 0.7, >= 0.7.1)
       dry-core (~> 0.4)
       dry-equalizer (~> 0.2)
       dry-initializer (~> 3.0)
-      dry-schema (~> 1.5)
+      dry-schema (~> 1.5, >= 1.5.2)
     envlogic (1.1.2)
       dry-inflector (~> 0.1)
     factory_bot (6.1.0)
@@ -87,45 +86,46 @@ GEM
     i18n (1.8.5)
       concurrent-ruby (~> 1.0)
     io-console (0.5.6)
-    irb (1.2.4)
-      reline (>= 0.0.1)
+    irb (1.2.7)
+      reline (>= 0.1.5)
     king_konf (0.3.7)
-    minitest (5.14.1)
-    multi_json (1.15.0)
+    minitest (5.14.2)
     rake (13.0.1)
-    reline (0.1.4)
+    reline (0.1.9)
       io-console (~> 0.5)
-    rspec (3.9.0)
-      rspec-core (~> 3.9.0)
-      rspec-expectations (~> 3.9.0)
-      rspec-mocks (~> 3.9.0)
-    rspec-core (3.9.2)
-      rspec-support (~> 3.9.3)
-    rspec-expectations (3.9.2)
+    rspec (3.10.0)
+      rspec-core (~> 3.10.0)
+      rspec-expectations (~> 3.10.0)
+      rspec-mocks (~> 3.10.0)
+    rspec-core (3.10.0)
+      rspec-support (~> 3.10.0)
+    rspec-expectations (3.10.0)
       diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.9.0)
-    rspec-mocks (3.9.1)
+      rspec-support (~> 3.10.0)
+    rspec-mocks (3.10.0)
       diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.9.0)
-    rspec-support (3.9.3)
-    ruby-kafka (1.2.0)
+      rspec-support (~> 3.10.0)
+    rspec-support (3.10.0)
+    ruby-kafka (1.3.0)
       digest-crc
-    simplecov (0.18.5)
+    simplecov (0.20.0)
       docile (~> 1.1)
       simplecov-html (~> 0.11)
-    simplecov-html (0.12.2)
+      simplecov_json_formatter (~> 0.1)
+    simplecov-html (0.12.3)
+    simplecov_json_formatter (0.1.2)
     thor (1.0.1)
     thread_safe (0.3.6)
-    tzinfo (1.2.7)
+    tzinfo (1.2.8)
       thread_safe (~> 0.1)
-    waterdrop (1.3.4)
+    waterdrop (1.4.0)
       delivery_boy (>= 0.2, < 2.x)
       dry-configurable (~> 0.8)
       dry-monitor (~> 0.3)
       dry-validation (~> 1.2)
       ruby-kafka (>= 0.7.8)
       zeitwerk (~> 2.1)
-    zeitwerk (2.4.0)
+    zeitwerk (2.4.2)
 
 PLATFORMS
   ruby

data/README.md

@@ -2,11 +2,9 @@
 
 [![Build Status](https://travis-ci.org/karafka/karafka.svg?branch=master)](https://travis-ci.org/karafka/karafka)
 
-**Note**: Documentation presented here refers to Karafka `1.3.x`.
+**Note**: Documentation presented here refers to Karafka `1.4`.
 
-If you're upgrading from `1.2.0`, please refer to our [Upgrade Notes article](https://mensfeld.pl/2019/09/karafka-framework-1-3-0-release-notes-ruby-kafka/).
-
-If you are looking for the documentation for Karafka `1.2.x`, it can be found [here](https://github.com/karafka/wiki/tree/1.2).
+If you are looking for the documentation for Karafka `1.3.x`, it can be found [here](https://github.com/karafka/wiki/tree/1.3).
 
 ## About Karafka
 

data/config/errors.yml

@@ -37,3 +37,5 @@ en:
         Unknown consumer group
       does_not_exist:
         Given file does not exist or cannot be read
+      does_not_respond_to_call: >
+        needs to respond to a #call method

data/karafka.gemspec

@@ -23,11 +23,10 @@ Gem::Specification.new do |spec|
   spec.add_dependency 'dry-validation', '~> 1.2'
   spec.add_dependency 'envlogic', '~> 1.1'
   spec.add_dependency 'irb', '~> 1.0'
-  spec.add_dependency 'multi_json', '>= 1.12'
   spec.add_dependency 'rake', '>= 11.3'
   spec.add_dependency 'ruby-kafka', '>= 1.0.0'
   spec.add_dependency 'thor', '>= 0.20'
-  spec.add_dependency 'waterdrop', '~> 1.3.0'
+  spec.add_dependency 'waterdrop', '~> 1.4.0'
   spec.add_dependency 'zeitwerk', '~> 2.1'
 
   spec.required_ruby_version = '>= 2.5.0'

data/lib/karafka.rb

@@ -6,10 +6,10 @@
   waterdrop
   kafka
   envlogic
+  json
   thor
   forwardable
   fileutils
-  multi_json
   dry-configurable
   dry-validation
   dry/events/publisher

data/lib/karafka/assignment_strategies/round_robin.rb

@@ -0,0 +1,13 @@
+# frozen_string_literal: true
+
+module Karafka
+  # Strategies for Kafka partitions assignments
+  module AssignmentStrategies
+    # Standard RoundRobin strategy
+    class RoundRobin < SimpleDelegator
+      def initialize
+        super(Kafka::RoundRobinAssignmentStrategy.new)
+      end
+    end
+  end
+end

data/lib/karafka/attributes_map.rb

@@ -19,6 +19,7 @@ module Karafka
           consumer: %i[
             session_timeout offset_commit_interval offset_commit_threshold
             offset_retention_time heartbeat_interval fetcher_max_queue_size
+            assignment_strategy
           ],
           subscribe: %i[start_from_beginning max_bytes_per_partition],
           consumption: %i[min_bytes max_bytes max_wait_time],

data/lib/karafka/cli.rb

@@ -10,6 +10,8 @@ module Karafka
   class Cli < Thor
     package_name 'Karafka'
 
+    default_task :missingno
+
     class << self
       # Loads all Cli commands into Thor framework
       # This method should be executed before we run Karafka::Cli.start, otherwise we won't
@@ -20,6 +22,12 @@ module Karafka
         end
       end
 
+      # When there is a CLI crash, exit
+      # @return [true]
+      def exit_on_failure?
+        true
+      end
+
       private
 
       # @return [Array<Class>] Array with Cli action classes that can be used as commands

data/lib/karafka/cli/base.rb

@@ -43,16 +43,16 @@ module Karafka
         end
 
         # Allows to set description of a given cli command
-        # @param desc [String] Description of a given cli command
-        def desc(desc)
-          @desc ||= desc
+        # @param args [Array] All the arguments that Thor desc method accepts
+        def desc(*args)
+          @desc ||= args
         end
 
         # This method will bind a given Cli command into Karafka Cli
         # This method is a wrapper to way Thor defines its commands
         # @param cli_class [Karafka::Cli] Karafka cli_class
         def bind_to(cli_class)
-          cli_class.desc name, @desc
+          cli_class.desc name, *@desc
 
           (@options || []).each { |option| cli_class.option(*option) }
 

data/lib/karafka/cli/missingno.rb

@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+module Karafka
+  class Cli < Thor
+    # Command that gets invoked when no method is provided when running the CLI
+    # It allows us to exit with exit code 1 instead of default 0 to indicate that something
+    #   was missing
+    # @see https://github.com/karafka/karafka/issues/619
+    class Missingno < Base
+      desc 'Hidden command that gets invoked when no command is provided', hide: true
+
+      # Prints an error about the lack of command (nothing selected)
+      def call
+        Karafka.logger.error('No command provided')
+        exit 1
+      end
+    end
+  end
+end

data/lib/karafka/connection/api_adapter.rb

@@ -109,11 +109,13 @@ module Karafka
           # Majority of users don't use custom topic mappers. No need to change anything when it
           # is a default mapper that does not change anything. Only some cloud providers require
           # topics to be remapped
-          return [params] if Karafka::App.config.topic_mapper.is_a?(Karafka::Routing::TopicMapper)
+          return [params.metadata] if Karafka::App.config.topic_mapper.is_a?(
+            Karafka::Routing::TopicMapper
+          )
 
           # @note We don't use tap as it is around 13% slower than non-dup version
-          dupped = params.dup
-          dupped['topic'] = Karafka::App.config.topic_mapper.outgoing(params.topic)
+          dupped = params.metadata.dup
+          dupped['topic'] = Karafka::App.config.topic_mapper.outgoing(params.metadata.topic)
           [dupped]
         end
 

data/lib/karafka/connection/batch_delegator.rb

@@ -23,7 +23,11 @@ module Karafka
           ) do
             # Due to how ruby-kafka is built, we have the metadata that is stored on the batch
             # level only available for batch consuming
-            consumer.metadata = Params::Builders::Metadata.from_kafka_batch(kafka_batch, topic)
+            consumer.batch_metadata = Params::Builders::BatchMetadata.from_kafka_batch(
+              kafka_batch,
+              topic
+            )
+
             kafka_messages = kafka_batch.messages
 
             # Depending on a case (persisted or not) we might use new consumer instance per

data/lib/karafka/consumers/batch_metadata.rb

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+module Karafka
+  module Consumers
+    # Brings the batch metadata into consumers that support batch_fetching
+    module BatchMetadata
+      attr_accessor :batch_metadata
+    end
+  end
+end

data/lib/karafka/consumers/includer.rb

@@ -16,7 +16,7 @@ module Karafka
 
           bind_backend(consumer, topic)
           bind_params(consumer, topic)
-          bind_metadata(consumer, topic)
+          bind_batch_metadata(consumer, topic)
           bind_responders(consumer, topic)
         end
 
@@ -40,13 +40,14 @@ module Karafka
           consumer.extend(SingleParams)
         end
 
-        # Adds an option to work with metadata for consumer instances that have batch fetching
+        # Adds an option to work with batch metadata for consumer instances that have
+        #   batch fetching enabled
         # @param consumer [Karafka::BaseConsumer] consumer instance
         # @param topic [Karafka::Routing::Topic] topic of a consumer class
-        def bind_metadata(consumer, topic)
+        def bind_batch_metadata(consumer, topic)
           return unless topic.batch_fetching
 
-          consumer.extend(Metadata)
+          consumer.extend(BatchMetadata)
         end
 
         # Adds responders support for topics and consumers with responders defined for them

data/lib/karafka/consumers/single_params.rb

@@ -8,7 +8,7 @@ module Karafka
 
       # @return [Karafka::Params::Params] params instance for non batch consumption consumers
       def params
-        params_batch.to_a.first
+        params_batch.first
       end
     end
   end

data/lib/karafka/contracts.rb

@@ -5,6 +5,6 @@ module Karafka
   module Contracts
     # Regexp for validating format of groups and topics
     # @note It is not nested inside of the contracts, as it is used by couple of them
-    TOPIC_REGEXP = /\A(\w|\-|\.)+\z/.freeze
+    TOPIC_REGEXP = /\A(\w|-|\.)+\z/.freeze
   end
 end

data/lib/karafka/contracts/consumer_group.rb

@@ -32,6 +32,7 @@ module Karafka
         required(:offset_retention_time).maybe(:integer)
         required(:heartbeat_interval).filled { (int? | float?) & gteq?(0) }
         required(:fetcher_max_queue_size).filled(:int?, gt?: 0)
+        required(:assignment_strategy).value(:any)
         required(:connect_timeout).filled { (int? | float?) & gt?(0) }
         required(:reconnect_timeout).filled { (int? | float?) & gteq?(0) }
         required(:socket_timeout).filled { (int? | float?) & gt?(0) }
@@ -70,13 +71,13 @@ module Karafka
 
       # Uri rule to check if uri is in a Karafka acceptable format
       rule(:seed_brokers) do
-        if value&.is_a?(Array) && !value.all?(&method(:kafka_uri?))
+        if value.is_a?(Array) && !value.all?(&method(:kafka_uri?))
           key.failure(:invalid_broker_schema)
         end
       end
 
       rule(:topics) do
-        if value&.is_a?(Array)
+        if value.is_a?(Array)
           names = value.map { |topic| topic[:name] }
 
           key.failure(:topics_names_not_unique) if names.size != names.uniq.size
@@ -84,7 +85,7 @@ module Karafka
       end
 
       rule(:topics) do
-        if value&.is_a?(Array)
+        if value.is_a?(Array)
           value.each_with_index do |topic, index|
             TOPIC_CONTRACT.call(topic).errors.each do |error|
               key([:topics, index, error.path[0]]).failure(error.text)
@@ -93,6 +94,10 @@ module Karafka
         end
       end
 
+      rule(:assignment_strategy) do
+        key.failure(:does_not_respond_to_call) unless value.respond_to?(:call)
+      end
+
       rule(:ssl_client_cert, :ssl_client_cert_key) do
         if values[:ssl_client_cert] && !values[:ssl_client_cert_key]
           key(:ssl_client_cert_key).failure(:ssl_client_cert_with_ssl_client_cert_key)

data/lib/karafka/helpers/class_matcher.rb

@@ -8,7 +8,7 @@ module Karafka
     class ClassMatcher
       # Regexp used to remove any non classy like characters that might be in the consumer
       # class name (if defined dynamically, etc)
-      CONSTANT_REGEXP = %r{[?!=+\-\*/\^\|&\[\]<>%~\#\:\s\(\)]}.freeze
+      CONSTANT_REGEXP = %r{[?!=+\-*/\^|&\[\]<>%~\#:\s()]}.freeze
 
       private_constant :CONSTANT_REGEXP
 

data/lib/karafka/instrumentation/logger.rb

@@ -29,11 +29,11 @@ module Karafka
 
       # @return [Karafka::Helpers::MultiDelegator] multi delegator instance
       #   to which we will be writing logs
-      # We use this approach to log stuff to file and to the STDOUT at the same time
+      # We use this approach to log stuff to file and to the $stdout at the same time
       def target
         Karafka::Helpers::MultiDelegator
           .delegate(:write, :close)
-          .to(STDOUT, file)
+          .to($stdout, file)
       end
 
       # Makes sure the log directory exists as long as we can write to it

data/lib/karafka/instrumentation/stdout_listener.rb

@@ -43,7 +43,7 @@ module Karafka
         # so it returns a topic as a string, not a routing topic
         debug(
           <<~MSG.chomp.tr("\n", ' ')
-            Params deserialization for #{event[:caller].topic} topic
+            Params deserialization for #{event[:caller].metadata.topic} topic
             successful in #{event[:time]} ms
           MSG
         )
@@ -52,7 +52,9 @@ module Karafka
       # Logs unsuccessful deserialization attempts of incoming data
       # @param event [Dry::Events::Event] event details including payload
       def on_params_params_deserialize_error(event)
-        error "Params deserialization error for #{event[:caller].topic} topic: #{event[:error]}"
+        topic = event[:caller].metadata.topic
+        error = event[:error]
+        error "Params deserialization error for #{topic} topic: #{error}"
       end
 
       # Logs errors that occurred in a listener fetch loop

data/lib/karafka/params/batch_metadata.rb

@@ -0,0 +1,26 @@
+# frozen_string_literal: true
+
+module Karafka
+  module Params
+    # Simple batch metadata object that stores all non-message information received from Kafka
+    # cluster while fetching the data
+    # @note This metadata object refers to per batch metadata, not `#params.metadata`
+    BatchMetadata = Struct.new(
+      :batch_size,
+      :first_offset,
+      :highwater_mark_offset,
+      :unknown_last_offset,
+      :last_offset,
+      :offset_lag,
+      :deserializer,
+      :partition,
+      :topic,
+      keyword_init: true
+    ) do
+      # @return [Boolean] is the last offset known or unknown
+      def unknown_last_offset?
+        unknown_last_offset
+      end
+    end
+  end
+end

data/lib/karafka/params/builders/batch_metadata.rb

@@ -0,0 +1,30 @@
+# frozen_string_literal: true
+
+module Karafka
+  module Params
+    module Builders
+      # Builder for creating batch metadata object based on the batch informations
+      module BatchMetadata
+        class << self
+          # Creates metadata based on the kafka batch data
+          # @param kafka_batch [Kafka::FetchedBatch] kafka batch details
+          # @param topic [Karafka::Routing::Topic] topic for which we've fetched the batch
+          # @return [Karafka::Params::BatchMetadata] batch metadata object
+          def from_kafka_batch(kafka_batch, topic)
+            Karafka::Params::BatchMetadata.new(
+              batch_size: kafka_batch.messages.count,
+              first_offset: kafka_batch.first_offset,
+              highwater_mark_offset: kafka_batch.highwater_mark_offset,
+              unknown_last_offset: kafka_batch.unknown_last_offset?,
+              last_offset: kafka_batch.last_offset,
+              offset_lag: kafka_batch.offset_lag,
+              deserializer: topic.deserializer,
+              partition: kafka_batch.partition,
+              topic: topic.name
+            ).freeze
+          end
+        end
+      end
+    end
+  end
+end

data/lib/karafka/params/builders/params.rb

@@ -12,22 +12,24 @@ module Karafka
         class << self
           # @param kafka_message [Kafka::FetchedMessage] message fetched from Kafka
           # @param topic [Karafka::Routing::Topic] topic for which this message was fetched
-          # @return [Karafka::Params::Params] params object
+          # @return [Karafka::Params::Params] params object with payload and message metadata
           def from_kafka_message(kafka_message, topic)
-            Karafka::Params::Params
-              .new
-              .merge!(
-                'create_time' => kafka_message.create_time,
-                'headers' => kafka_message.headers || {},
-                'is_control_record' => kafka_message.is_control_record,
-                'key' => kafka_message.key,
-                'offset' => kafka_message.offset,
-                'deserializer' => topic.deserializer,
-                'partition' => kafka_message.partition,
-                'receive_time' => Time.now,
-                'topic' => kafka_message.topic,
-                'payload' => kafka_message.value
-              )
+            metadata = Karafka::Params::Metadata.new(
+              create_time: kafka_message.create_time,
+              headers: kafka_message.headers || {},
+              is_control_record: kafka_message.is_control_record,
+              key: kafka_message.key,
+              offset: kafka_message.offset,
+              deserializer: topic.deserializer,
+              partition: kafka_message.partition,
+              receive_time: Time.now,
+              topic: topic.name
+            ).freeze
+
+            Karafka::Params::Params.new(
+              kafka_message.value,
+              metadata
+            )
           end
         end
       end

data/lib/karafka/params/builders/params_batch.rb

@@ -12,11 +12,11 @@ module Karafka
           # @param topic [Karafka::Routing::Topic] topic for which we're received messages
           # @return [Karafka::Params::ParamsBatch<Karafka::Params::Params>] batch with params
           def from_kafka_messages(kafka_messages, topic)
-            params_array = kafka_messages.map! do |message|
+            params_array = kafka_messages.map do |message|
               Karafka::Params::Builders::Params.from_kafka_message(message, topic)
             end
 
-            Karafka::Params::ParamsBatch.new(params_array)
+            Karafka::Params::ParamsBatch.new(params_array).freeze
           end
         end
       end

data/lib/karafka/params/metadata.rb

@@ -2,34 +2,19 @@
 
 module Karafka
   module Params
-    # Simple metadata object that stores all non-message information received from Kafka cluster
-    # while fetching the data
-    class Metadata < Hash
-      # Attributes that should be accessible as methods as well (not only hash)
-      METHOD_ATTRIBUTES = %w[
-        batch_size
-        first_offset
-        highwater_mark_offset
-        last_offset
-        offset_lag
-        deserializer
-        partition
-        topic
-      ].freeze
-
-      private_constant :METHOD_ATTRIBUTES
-
-      METHOD_ATTRIBUTES.each do |attr|
-        # Defines a method call accessor to a particular hash field.
-        define_method(attr) do
-          self[attr]
-        end
-      end
-
-      # @return [Boolean] is the last offset known or unknown
-      def unknown_last_offset?
-        self['unknown_last_offset']
-      end
-    end
+    # Single message / params metadata details that can be accessed without the need for the
+    # payload deserialization
+    Metadata = Struct.new(
+      :create_time,
+      :headers,
+      :is_control_record,
+      :key,
+      :offset,
+      :deserializer,
+      :partition,
+      :receive_time,
+      :topic,
+      keyword_init: true
+    )
   end
 end

data/lib/karafka/params/params.rb

@@ -6,68 +6,44 @@ module Karafka
     # It provides lazy loading not only until the first usage, but also allows us to skip
     # using deserializer until we execute our logic. That way we can operate with
     # heavy-deserialization data without slowing down the whole application.
-    class Params < Hash
-      # Params attributes that should be available via a method call invocation for Kafka
-      # client compatibility.
-      # Kafka passes internally Kafka::FetchedMessage object and the ruby-kafka consumer
-      # uses those fields via method calls, so in order to be able to pass there our params
-      # objects, have to have same api.
-      METHOD_ATTRIBUTES = %w[
-        create_time
-        headers
-        is_control_record
-        key
-        offset
-        deserializer
-        deserialized
-        partition
-        receive_time
-        topic
-        payload
-      ].freeze
+    class Params
+      extend Forwardable
 
-      private_constant :METHOD_ATTRIBUTES
+      attr_reader :raw_payload, :metadata
 
-      METHOD_ATTRIBUTES.each do |attr|
-        # Defines a method call accessor to a particular hash field.
-        # @note Won't work for complex key names that contain spaces, etc
-        # @param key [Symbol] name of a field that we want to retrieve with a method call
-        # @example
-        #   key_attr_reader :example
-        #   params.example #=> 'my example payload'
-        define_method(attr) do
-          self[attr]
-        end
-      end
+      def_delegators :metadata, *Metadata.members
 
-      # Overrides `Hash#[]` to allow lazy deserialization of payload. This allows us to fetch
-      # metadata without actually triggering deserialization of the payload until it is needed
-      # @param key [String, Symbol] hash key
-      # @return [Object] content of a given params key
-      def [](key)
-        # Payload will be deserialized only when we request for it.
-        deserialize! if key == 'payload'
-        super
+      # @param raw_payload [Object] incoming payload before deserialization
+      # @param metadata [Karafka::Params::Metadata] message metadata object
+      def initialize(raw_payload, metadata)
+        @raw_payload = raw_payload
+        @metadata = metadata
+        @deserialized = false
+        @payload = nil
       end
 
-      # @return [Karafka::Params::Params] This method will trigger deserializer execution. If we
-      #   decide to retrieve data, deserializer will be executed to get data. Output of that will
-      #   be merged to the current object. This object will be also marked as already deserialized,
-      #   so we won't deserialize it again.
-      def deserialize!
-        return self if self['deserialized']
+      # @return [Object] lazy-deserialized data (deserialized upon first request)
+      def payload
+        return @payload if deserialized?
+
+        @payload = deserialize
+        # We mark deserialization as successful after deserialization, as in case of an error
+        # this won't be falsely set to true
+        @deserialized = true
+        @payload
+      end
 
-        self['deserialized'] = true
-        self['payload'] = deserialize
-        self
+      # @return [Boolean] did given params payload were deserialized already
+      def deserialized?
+        @deserialized
       end
 
       private
 
-      # @return [Object] deserialized data
+      # @return [Object] tries de-serializes data
       def deserialize
         Karafka.monitor.instrument('params.params.deserialize', caller: self) do
-          self['deserializer'].call(self)
+          metadata.deserializer.call(self)
         end
       rescue ::StandardError => e
         Karafka.monitor.instrument('params.params.deserialize.error', caller: self, error: e)

data/lib/karafka/params/params_batch.rb

@@ -15,47 +15,46 @@ module Karafka
         @params_array = params_array
       end
 
-      # @yieldparam [Karafka::Params::Params] each deserialized and loaded params instance
-      # @note Invocation of this method will cause loading and deserializing each param after
-      #   another. If you want to get access without deserializing, please access params_array
-      #   directly
+      # @yieldparam [Karafka::Params::Params] each params instance
+      # @note Invocation of this method will not cause loading and deserializing each param after
+      #   another.
       def each
-        @params_array.each { |param| yield(param.deserialize!) }
+        @params_array.each { |param| yield(param) }
       end
 
       # @return [Array<Karafka::Params::Params>] returns all the params in a loaded state, so they
       #   can be used for batch insert, etc. Without invoking all, up until first use, they won't
       #   be deserialized
       def deserialize!
-        each(&:itself)
+        each(&:payload)
       end
 
       # @return [Array<Object>] array with deserialized payloads. This method can be useful when
       #   we don't care about metadata and just want to extract all the data payloads from the
       #   batch
       def payloads
-        deserialize!.map(&:payload)
+        map(&:payload)
       end
 
-      # @return [Karafka::Params::Params] first element after the deserialization process
+      # @return [Karafka::Params::Params] first element
       def first
-        @params_array.first.deserialize!
+        @params_array.first
       end
 
-      # @return [Karafka::Params::Params] last element after the deserialization process
+      # @return [Karafka::Params::Params] last element
       def last
-        @params_array.last.deserialize!
-      end
-
-      # @return [Array<Karafka::Params::Params>] pure array with params (not deserialized)
-      def to_a
-        @params_array
+        @params_array.last
       end
 
       # @return [Integer] number of messages in the batch
       def size
         @params_array.size
       end
+
+      # @return [Array<Karafka::Params::Params>] pure array with params
+      def to_a
+        @params_array
+      end
     end
   end
 end

data/lib/karafka/routing/builder.rb

@@ -16,6 +16,7 @@ module Karafka
       private_constant :CONTRACT
 
       def initialize
+        super
         @draws = Concurrent::Array.new
       end
 

data/lib/karafka/routing/consumer_group.rb

@@ -8,9 +8,11 @@ module Karafka
     class ConsumerGroup
       extend Helpers::ConfigRetriever
 
-      attr_reader :topics
-      attr_reader :id
-      attr_reader :name
+      attr_reader(
+        :topics,
+        :id,
+        :name
+      )
 
       # @param name [String, Symbol] raw name of this consumer group. Raw means, that it does not
       #   yet have an application client_id namespace, this will be added here by default.

data/lib/karafka/serialization/json/deserializer.rb

@@ -17,8 +17,8 @@ module Karafka
         #   }
         #   Deserializer.call(params) #=> { 'a' => 1 }
         def call(params)
-          ::MultiJson.load(params['payload'])
-        rescue ::MultiJson::ParseError => e
+          params.raw_payload.nil? ? nil : ::JSON.parse(params.raw_payload)
+        rescue ::JSON::ParserError => e
           raise ::Karafka::Errors::DeserializationError, e
         end
       end

data/lib/karafka/setup/config.rb

@@ -89,6 +89,9 @@ module Karafka
         #   are stored for further processing. Note, that each item in the queue represents a
         #   response from a single broker
         setting :fetcher_max_queue_size, 10
+        # option assignment_strategy [Object] a strategy determining the assignment of
+        #   partitions to the consumers.
+        setting :assignment_strategy, Karafka::AssignmentStrategies::RoundRobin.new
         # option max_bytes_per_partition [Integer] the maximum amount of data fetched
         #   from a single partition at a time.
         setting :max_bytes_per_partition, 1_048_576

data/lib/karafka/version.rb

@@ -3,5 +3,5 @@
 # Main module namespace
 module Karafka
   # Current Karafka version
-  VERSION = '1.3.7'
+  VERSION = '1.4.1'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: karafka
 version: !ruby/object:Gem::Version
-  version: 1.3.7
+  version: 1.4.1
 platform: ruby
 authors:
 - Maciej Mensfeld
@@ -36,7 +36,7 @@ cert_chain:
   2DND//YJUikn1zwbz1kT70XmHd97B4Eytpln7K+M1u2g1pHVEPW4owD/ammXNpUy
   nt70FcDD4yxJQ+0YNiHd0N8IcVBM1TMIVctMNQ==
   -----END CERTIFICATE-----
-date: 2020-08-11 00:00:00.000000000 Z
+date: 2020-12-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: dry-configurable
@@ -122,20 +122,6 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.0'
-- !ruby/object:Gem::Dependency
-  name: multi_json
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '1.12'
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '1.12'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
@@ -184,14 +170,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.3.0
+        version: 1.4.0
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.3.0
+        version: 1.4.0
 - !ruby/object:Gem::Dependency
   name: zeitwerk
   requirement: !ruby/object:Gem::Requirement
@@ -241,6 +227,7 @@ files:
 - karafka.gemspec
 - lib/karafka.rb
 - lib/karafka/app.rb
+- lib/karafka/assignment_strategies/round_robin.rb
 - lib/karafka/attributes_map.rb
 - lib/karafka/backends/inline.rb
 - lib/karafka/base_consumer.rb
@@ -251,6 +238,7 @@ files:
 - lib/karafka/cli/flow.rb
 - lib/karafka/cli/info.rb
 - lib/karafka/cli/install.rb
+- lib/karafka/cli/missingno.rb
 - lib/karafka/cli/server.rb
 - lib/karafka/code_reloader.rb
 - lib/karafka/connection/api_adapter.rb
@@ -259,9 +247,9 @@ files:
 - lib/karafka/connection/client.rb
 - lib/karafka/connection/listener.rb
 - lib/karafka/connection/message_delegator.rb
+- lib/karafka/consumers/batch_metadata.rb
 - lib/karafka/consumers/callbacks.rb
 - lib/karafka/consumers/includer.rb
-- lib/karafka/consumers/metadata.rb
 - lib/karafka/consumers/responders.rb
 - lib/karafka/consumers/single_params.rb
 - lib/karafka/contracts.rb
@@ -280,7 +268,8 @@ files:
 - lib/karafka/instrumentation/monitor.rb
 - lib/karafka/instrumentation/proctitle_listener.rb
 - lib/karafka/instrumentation/stdout_listener.rb
-- lib/karafka/params/builders/metadata.rb
+- lib/karafka/params/batch_metadata.rb
+- lib/karafka/params/builders/batch_metadata.rb
 - lib/karafka/params/builders/params.rb
 - lib/karafka/params/builders/params_batch.rb
 - lib/karafka/params/metadata.rb
@@ -331,7 +320,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
+rubygems_version: 3.1.4
 signing_key: 
 specification_version: 4
 summary: Ruby based framework for working with Apache Kafka

data/lib/karafka/consumers/metadata.rb

@@ -1,10 +0,0 @@
-# frozen_string_literal: true
-
-module Karafka
-  module Consumers
-    # Brings the metadata into consumers that support batch_fetching
-    module Metadata
-      attr_accessor :metadata
-    end
-  end
-end

data/lib/karafka/params/builders/metadata.rb

@@ -1,33 +0,0 @@
-# frozen_string_literal: true
-
-module Karafka
-  module Params
-    module Builders
-      # Builder for creating metadata object based on the message or batch informations
-      # @note We have 2 ways of creating metadata based on the way ruby-kafka operates
-      module Metadata
-        class << self
-          # Creates metadata based on the kafka batch data
-          # @param kafka_batch [Kafka::FetchedBatch] kafka batch details
-          # @param topic [Karafka::Routing::Topic] topic for which we've fetched the batch
-          # @return [Karafka::Params::Metadata] metadata object
-          def from_kafka_batch(kafka_batch, topic)
-            Karafka::Params::Metadata
-              .new
-              .merge!(
-                'batch_size' => kafka_batch.messages.count,
-                'first_offset' => kafka_batch.first_offset,
-                'highwater_mark_offset' => kafka_batch.highwater_mark_offset,
-                'last_offset' => kafka_batch.last_offset,
-                'offset_lag' => kafka_batch.offset_lag,
-                'deserializer' => topic.deserializer,
-                'partition' => kafka_batch.partition,
-                'topic' => kafka_batch.topic,
-                'unknown_last_offset' => kafka_batch.unknown_last_offset?
-              )
-          end
-        end
-      end
-    end
-  end
-end