snowplow-tracker 0.4.2 → 0.5.0

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    MDU0N2EyMjU3MDk1ODBhMGZlMzRlNjg1Mzg4ZmQ3OTg1N2QwZGFhMw==
+    NDkzMWJkNzQyZjVmNjk0NmQ0YjI5ODA5YzkxM2Y1NmZhZmEwNTgzNA==
   data.tar.gz: !binary |-
-    NjgxMmRlZGNjNThlZWZkYTIwNDk5Y2ViZTE4ZjA4NmEyYzMzOThiMw==
+    NTU5NDI2M2UzYTQ0M2ZhNGEyYjM2ZDdjMzQwZjA4ZGYyNDQ3OWU0OA==
 SHA512:
   metadata.gz: !binary |-
-    MTg4MDk0ZGM5ZjI0MjhiZmI2NzE0N2E3ZjJlMjNkNjVlZmVkZTg5MzhlMDQx
-    NTFhNmIwYWMxMGRhMmM0ODVkMzJkZTcwM2JmZjg4ODFlYTMwNDM0Y2MxNjE3
-    YjE3YWZkYTVhNjkwN2M3MWNiYmRlOTUyZGFlMGYyMzc3OGEwNDA=
+    NjJiM2Y1ZWM2Y2Y3MTNkZDNmNWM0N2IwMGNkOTAzMTlhNjAxZjk5MTk3ZmUz
+    MjcyNTYwNmZjZjhmMWVlZGNiMmJiNGZiOGZlMmJlY2YxNDE1MjY0MGFlYzI0
+    M2E2OGY5YTkwM2M4Y2M1NjQyNGEyNzIxNzYzNDk1ZDY1MGU0NDg=
   data.tar.gz: !binary |-
-    MDBmMDgwOGM4ZWI1Y2IyMTNlZWQzMThjMWM1ZDk5MWQwZTJhNGIyMGIxOTQw
-    YTc5YmJjZjRjYzg2NDJmNGNlODFkMGFmYjkyNzQxZjc1MGI4MzZhMGUzYjkz
-    NzU4MTk0MTBmN2M4Njg1ODczNzIzM2Q3YTQ4YmRmNTQyOWVkMjA=
+    YjgxODJjMWY0NmIwOTNjODNmZmQ1NTZkOTMwNDBkODMxODZhMzc1Njk2ODc1
+    NDkzZDYzNWMyZDQ0MzBjYzRiNzI3ZTdkNjI4OWY5ZjUyMTgzMjI3ZWYyOTI4
+    NGZhYTljNTFkOTRhYWQ5MzEyMGFmMmU0ZGNhNGFjZWQ0YzQ2NmI=

data/README.md

@@ -1,6 +1,6 @@
 # Ruby Analytics for Snowplow 
 [![Gem Version](https://badge.fury.io/rb/snowplow-tracker.svg)](http://badge.fury.io/rb/snowplow-tracker)
-[![Build Status](https://travis-ci.org/snowplow/snowplow-ruby-tracker.png)](https://travis-ci.org/snowplow/snowplow-ruby-tracker)
+[![Build Status](https://travis-ci.org/snowplow/snowplow-ruby-tracker.png?branch=master)](https://travis-ci.org/snowplow/snowplow-ruby-tracker)
 [![Code Climate](https://codeclimate.com/github/snowplow/snowplow-ruby-tracker.png)](https://codeclimate.com/github/snowplow/snowplow-ruby-tracker)
 [![Coverage Status](https://coveralls.io/repos/snowplow/snowplow-ruby-tracker/badge.png)](https://coveralls.io/r/snowplow/snowplow-ruby-tracker)
 [![License][license-image]][license]

data/lib/snowplow-tracker.rb

@@ -15,6 +15,7 @@
 
 require 'snowplow-tracker/contracts.rb'
 require 'snowplow-tracker/version.rb'
+require 'snowplow-tracker/self_describing_json.rb'
 require 'snowplow-tracker/payload.rb'
 require 'snowplow-tracker/subject.rb'
 require 'snowplow-tracker/emitters.rb'

data/lib/snowplow-tracker/contracts.rb

@@ -14,7 +14,6 @@
 # License:: Apache License Version 2.0
 
 require 'contracts'
-include Contracts
 
 module SnowplowTracker
 

data/lib/snowplow-tracker/emitters.rb

@@ -17,7 +17,6 @@ require 'net/https'
 require 'set'
 require 'logger'
 require 'contracts'
-include Contracts
 
 module SnowplowTracker
 
@@ -26,13 +25,16 @@ module SnowplowTracker
 
   class Emitter
 
+    include Contracts
+
     @@ConfigHash = ({
       :protocol => Maybe[Or['http', 'https']],
       :port => Maybe[Num],
       :method => Maybe[Or['get', 'post']],
       :buffer_size => Maybe[Num],
       :on_success => Maybe[Func[Num => Any]],
-      :on_failure => Maybe[Func[Num, Hash => Any]]
+      :on_failure => Maybe[Func[Num, Hash => Any]],
+      :thread_count => Maybe[Num]
     })
 
     @@StrictConfigHash = And[@@ConfigHash, lambda { |x|
@@ -47,19 +49,19 @@ module SnowplowTracker
     Contract String, @@StrictConfigHash => lambda { |x| x.is_a? Emitter }
     def initialize(endpoint, config={})
       config = @@DefaultConfig.merge(config)
+      @lock = Monitor.new
       @collector_uri = as_collector_uri(endpoint, config[:protocol], config[:port], config[:method])
       @buffer = []
       if not config[:buffer_size].nil?
         @buffer_size = config[:buffer_size]
       elsif config[:method] == 'get'
-        @buffer_size = 0
+        @buffer_size = 1
       else
         @buffer_size = 10
       end
       @method = config[:method]
       @on_success = config[:on_success]
       @on_failure = config[:on_failure]
-      @threads = []
       LOGGER.info("#{self.class} initialized with endpoint #{@collector_uri}")
 
       self
@@ -80,9 +82,11 @@ module SnowplowTracker
     Contract Hash => nil
     def input(payload)
       payload.each { |k,v| payload[k] = v.to_s}
-      @buffer.push(payload)
-      if @buffer.size > @buffer_size
-        flush
+      @lock.synchronize do
+        @buffer.push(payload)
+        if @buffer.size >= @buffer_size
+          flush
+        end
       end
 
       nil
@@ -91,58 +95,70 @@ module SnowplowTracker
     # Flush the buffer
     #
     Contract Bool => nil
-    def flush(sync=false)
-      send_requests
-
+    def flush(async=true)
+      @lock.synchronize do
+        send_requests(@buffer)
+        @buffer = []
+      end
       nil
     end
 
     # Send all events in the buffer to the collector
     #
-    Contract None => nil
-    def send_requests
-      LOGGER.info("Attempting to send #{@buffer.size} request#{@buffer.size == 1 ? '' : 's'}")
-      temp_buffer = @buffer
-      @buffer = []
+    Contract ArrayOf[Hash] => nil
+    def send_requests(evts)
+      if evts.size < 1
+        LOGGER.info("Skipping sending events since buffer is empty")
+        return
+      end
+      LOGGER.info("Attempting to send #{evts.size} request#{evts.size == 1 ? '' : 's'}")
+
+      if @method == 'post'
+        post_succeeded = false
+        begin
+          request = http_post(SelfDescribingJson.new(
+            'iglu:com.snowplowanalytics.snowplow/payload_data/jsonschema/1-0-2',
+            evts
+          ).to_json)
+          post_succeeded = is_good_status_code(request.code)
+        rescue StandardError => se
+          LOGGER.warn(se)
+        end
+        if post_succeeded
+          unless @on_success.nil?
+            @on_success.call(evts.size)
+          end
+        else
+          unless @on_failure.nil?
+            @on_failure.call(0, evts)
+          end
+        end
 
-      if @method == 'get'
+      elsif @method == 'get'
         success_count = 0
         unsent_requests = []
-        temp_buffer.each do |payload|
-          request = http_get(payload)
-          if request.code.to_i == 200
-            success_count += 1
-          else
-            unsent_requests.push(payload)
+        evts.each do |evt|
+          get_succeeded = false
+          begin
+            request = http_get(evt)
+            get_succeeded = is_good_status_code(request.code)
+          rescue StandardError => se
+            LOGGER.warn(se)
           end
-          if unsent_requests.size == 0
-            unless @on_success.nil?
-              @on_success.call(success_count)
-            end
+          if get_succeeded
+            success_count += 1
           else
-            unless @on_failure.nil?
-              @on_failure.call(success_count, unsent_requests)
-            end
+            unsent_requests << evt
           end
         end
-
-      elsif @method == 'post'
-        if temp_buffer.size > 0
-          request = http_post({
-            'schema' => 'iglu:com.snowplowanalytics.snowplow/payload_data/jsonschema/1-0-2',
-            'data' => temp_buffer
-          })
-
-          if request.code.to_i == 200
-            unless @on_success.nil?
-              @on_success.call(temp_buffer.size)
-            end
-          else
-            unless @on_failure.nil?
-              @on_failure.call(0, temp_buffer)
-            end
+        if unsent_requests.size == 0
+          unless @on_success.nil?
+            @on_success.call(success_count)
+          end
+        else
+          unless @on_failure.nil?
+            @on_failure.call(success_count, unsent_requests)
           end
-
         end
       end
 
@@ -162,7 +178,7 @@ module SnowplowTracker
         http.use_ssl = true
       end
       response = http.request(request)
-      LOGGER.add(response.code == '200' ? Logger::INFO : Logger::WARN) {
+      LOGGER.add(is_good_status_code(response.code) ? Logger::INFO : Logger::WARN) {
         "GET request to #{@collector_uri} finished with status code #{response.code}"
       }
 
@@ -184,13 +200,20 @@ module SnowplowTracker
       request.body = payload.to_json
       request.set_content_type('application/json; charset=utf-8')
       response = http.request(request)
-      LOGGER.add(response.code == '200' ? Logger::INFO : Logger::WARN) {
+      LOGGER.add(is_good_status_code(response.code) ? Logger::INFO : Logger::WARN) {
         "POST request to #{@collector_uri} finished with status code #{response.code}"
       }
 
       response
     end
 
+    # Only 2xx and 3xx status codes are considered successes
+    #
+    Contract String => Bool
+    def is_good_status_code(status_code)
+      status_code.to_i >= 200 && status_code.to_i < 400
+    end
+
     private :as_collector_uri,
             :http_get,
             :http_post
@@ -200,27 +223,54 @@ module SnowplowTracker
 
   class AsyncEmitter < Emitter
 
-    # Flush the buffer in a new thread
-    #  If sync is true, block until all flushing threads have exited
-    #
-    def flush(sync=false)
-      t = Thread.new do
-        send_requests
-      end
-      t.abort_on_exception = true
-      @threads.select!{ |thread| thread.alive?}
-      @threads.push(t)
-
-      if sync
-        LOGGER.info('Starting synchronous flush')
-        @threads.each do |thread|
-          thread.join(10)
+    Contract String, @@StrictConfigHash => lambda { |x| x.is_a? Emitter }
+    def initialize(endpoint, config={})
+      @queue = Queue.new()
+      # @all_processed_condition and @results_unprocessed are used to emulate Python's Queue.task_done()
+      @queue.extend(MonitorMixin)
+      @all_processed_condition = @queue.new_cond
+      @results_unprocessed = 0
+      (config[:thread_count] || 1).times do
+        t = Thread.new do
+          consume
         end
       end
+      super(endpoint, config)
+    end
 
-      nil
+    def consume
+      loop do
+        work_unit = @queue.pop
+        send_requests(work_unit)
+        @queue.synchronize do
+          @results_unprocessed -= 1
+          @all_processed_condition.broadcast
+        end
+      end
     end
 
+    # Flush the buffer
+    #  If async is false, block until the queue is empty
+    #
+    def flush(async=true)
+      loop do
+        @lock.synchronize do
+          @queue.synchronize do
+            @results_unprocessed += 1
+          end
+          @queue << @buffer
+          @buffer = []
+        end
+        if not async
+          LOGGER.info('Starting synchronous flush')
+          @queue.synchronize do
+            @all_processed_condition.wait_while { @results_unprocessed > 0 }
+            LOGGER.info('Finished synchronous flush')
+          end
+        end
+        break if @buffer.size < 1
+      end
+    end
   end
 
 end

data/lib/snowplow-tracker/payload.rb

@@ -17,12 +17,13 @@ require 'base64'
 require 'json'
 require 'net/http'
 require 'contracts'
-include Contracts
 
 module SnowplowTracker
 
   class Payload
 
+    include Contracts
+
     attr_reader :context
 
     Contract nil => Payload

data/lib/snowplow-tracker/self_describing_json.rb

@@ -0,0 +1,34 @@
+# Copyright (c) 2013-2014 Snowplow Analytics Ltd. All rights reserved.
+#
+# This program is licensed to you under the Apache License Version 2.0,
+# and you may not use this file except in compliance with the Apache License Version 2.0.
+# You may obtain a copy of the Apache License Version 2.0 at http://www.apache.org/licenses/LICENSE-2.0.
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the Apache License Version 2.0 is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the Apache License Version 2.0 for the specific language governing permissions and limitations there under.
+
+# Author:: Alex Dean, Fred Blundun (mailto:support@snowplowanalytics.com)
+# Copyright:: Copyright (c) 2013-2014 Snowplow Analytics Ltd
+# License:: Apache License Version 2.0
+
+module SnowplowTracker
+
+  class SelfDescribingJson
+
+    def initialize(schema, data)
+      @schema = schema
+      @data = data
+    end
+
+    def to_json
+      {
+        :schema => @schema,
+        :data => @data
+      }
+    end
+
+  end
+
+end

data/lib/snowplow-tracker/subject.rb

@@ -14,12 +14,13 @@
 # License:: Apache License Version 2.0
 
 require 'contracts'
-include Contracts
 
 module SnowplowTracker
 
   class Subject
 
+    include Contracts
+
     @@default_platform = 'srv'
     @@supported_platforms = ['pc', 'tv', 'mob', 'cnsl', 'iot']
 
@@ -52,6 +53,14 @@ module SnowplowTracker
       self
     end
 
+    # Set fingerprint for the user
+    #
+    Contract Num => Subject
+    def set_fingerprint(fingerprint)
+      @standard_nv_pairs['fp'] = fingerprint
+      self
+    end
+
     # Set the screen resolution for a device
     #
     Contract Num, Num => Subject

data/lib/snowplow-tracker/tracker.rb

@@ -15,14 +15,14 @@
 
 require 'contracts'
 require 'set'
-include Contracts
-
 require 'uuid'
 
 module SnowplowTracker
 
   class Tracker
 
+    include Contracts
+
     @@EmitterInput = Or[lambda {|x| x.is_a? Emitter}, ArrayOf[lambda {|x| x.is_a? Emitter}]]
 
     @@required_transaction_keys =   Set.new(%w(order_id total_value))
@@ -55,22 +55,14 @@ module SnowplowTracker
         augmented_item_keys.subset? @@recognised_augmented_item_keys
     }
 
-    @@SelfDescribingJson = Or[{
-      schema: String,
-      data: Any
-    }, {
-      'schema' => String,
-      'data' => Any
-    }]
-
-    @@ContextsInput = ArrayOf[@@SelfDescribingJson]
+    @@ContextsInput = ArrayOf[SelfDescribingJson]
 
     @@version = TRACKER_VERSION
     @@default_encode_base64 = true
 
     @@base_schema_path = "iglu:com.snowplowanalytics.snowplow"
     @@schema_tag = "jsonschema"
-    @@context_schema = "#{@@base_schema_path}/contexts/#{@@schema_tag}/1-0-0"
+    @@context_schema = "#{@@base_schema_path}/contexts/#{@@schema_tag}/1-0-1"
     @@unstruct_event_schema = "#{@@base_schema_path}/unstruct_event/#{@@schema_tag}/1-0-0"
 
     Contract @@EmitterInput, Maybe[Subject], Maybe[String], Maybe[String], Bool => Tracker
@@ -119,12 +111,12 @@ module SnowplowTracker
 
     # Builds a self-describing JSON from an array of custom contexts
     #
-    Contract @@ContextsInput => @@SelfDescribingJson
+    Contract @@ContextsInput => Hash
     def build_context(context)
-      {
-        schema: @@context_schema,
-        data: context
-      }
+      SelfDescribingJson.new(
+        @@context_schema,
+        context.map {|c| c.to_json}
+        ).to_json
     end
 
     # Tracking methods
@@ -259,7 +251,8 @@ module SnowplowTracker
         screen_view_properties['id'] = id
       end
       screen_view_schema = "#{@@base_schema_path}/screen_view/#{@@schema_tag}/1-0-0"
-      event_json = {schema: screen_view_schema, data: screen_view_properties}
+
+      event_json = SelfDescribingJson.new(screen_view_schema, screen_view_properties)
 
       self.track_unstruct_event(event_json, context, tstamp)
 
@@ -268,16 +261,14 @@ module SnowplowTracker
 
     # Track an unstructured event
     #
-    Contract @@SelfDescribingJson, Maybe[@@ContextsInput], Maybe[Num] => Tracker
+    Contract SelfDescribingJson, Maybe[@@ContextsInput], Maybe[Num] => Tracker
     def track_unstruct_event(event_json, context=nil, tstamp=nil)
       pb = Payload.new
       pb.add('e', 'ue')
       
-      envelope = {
-        schema: @@unstruct_event_schema,
-        data: event_json
-      }
-      pb.add_json(envelope, @config['encode_base64'], 'ue_px', 'ue_pr')
+      envelope = SelfDescribingJson.new(@@unstruct_event_schema, event_json.to_json)
+
+      pb.add_json(envelope.to_json, @config['encode_base64'], 'ue_px', 'ue_pr')
 
       unless context.nil?
         pb.add_json(build_context(context), @config['encode_base64'], 'cx', 'co')
@@ -296,9 +287,9 @@ module SnowplowTracker
     # Flush all events stored in all emitters
     #
     Contract Bool => Tracker
-    def flush(sync=false)
+    def flush(async=false)
       @emitters.each do |emitter|
-        emitter.flush(sync)
+        emitter.flush(async)
       end
 
       self

data/lib/snowplow-tracker/version.rb

@@ -14,6 +14,6 @@
 # License:: Apache License Version 2.0
 
 module SnowplowTracker
-  VERSION = '0.4.2'
+  VERSION = '0.5.0'
   TRACKER_VERSION = "rb-#{VERSION}"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: snowplow-tracker
 version: !ruby/object:Gem::Version
-  version: 0.4.2
+  version: 0.5.0
 platform: ruby
 authors:
 - Alexander Dean
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-04-08 00:00:00.000000000 Z
+date: 2015-08-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: contracts
@@ -17,14 +17,20 @@ dependencies:
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '0.3'
+        version: '0.7'
+    - - <=
+      - !ruby/object:Gem::Version
+        version: '0.11'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ~>
       - !ruby/object:Gem::Version
-        version: '0.3'
+        version: '0.7'
+    - - <=
+      - !ruby/object:Gem::Version
+        version: '0.11'
 - !ruby/object:Gem::Dependency
   name: uuid
   requirement: !ruby/object:Gem::Requirement
@@ -80,6 +86,7 @@ files:
 - lib/snowplow-tracker/contracts.rb
 - lib/snowplow-tracker/emitters.rb
 - lib/snowplow-tracker/payload.rb
+- lib/snowplow-tracker/self_describing_json.rb
 - lib/snowplow-tracker/subject.rb
 - lib/snowplow-tracker/tracker.rb
 - lib/snowplow-tracker/version.rb
@@ -103,7 +110,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.4.6
+rubygems_version: 2.4.8
 signing_key: 
 specification_version: 4
 summary: Ruby Analytics for Snowplow