powertrack 1.0.3 → 1.1.0

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    ZjgwZmUwMGNlNDJmMWFkZmQ2NWI1Y2I5NTJmZDIyMmVkODliZDI5YQ==
+    NjZjMjc0OWE1ODU4NTUyNTY2Zjk1YzdiMjYxZDgxMjYwMTI0MmQyMg==
   data.tar.gz: !binary |-
-    NzkxMWY4YzRjZjY4NTkwMjljZDBlY2YzNTQxYzEwNWM0YWIxMTE4ZQ==
+    MmEzNDQ0YWYzZGYwODcwNDM0MDU5MDc1YjdlZTIwMTRjMmRhNzZlYQ==
 SHA512:
   metadata.gz: !binary |-
-    OTA4YTE2OTAwNmUwZmQyYzNhODM3MTI3Y2ViMzYyODhmNmY3ZThkYTQ4ZTYx
-    MTkwNjYyMzZhNmY1NjRhZmY3YjU2YmE2YjMwZjBiNGQ3M2RkODEyNmM4OTY5
-    OGI1NmJhOTZhZjAyODQxMmM5NzM2ZTJkZWNmOTg4NDcyMDQxZGQ=
+    NTU5ZmE3MzI3NzUzYTYxYzgyNGZiYzBkYjllODdiMTA5M2YyYWIyN2Y3M2E2
+    M2ZhNDRmZWFlNTE1ZTI4YTA0OGZiNDg4NDFiYWQwOWRiNGMwNjhlMTIzZWRi
+    NGNhNzFiZTIyM2Q3Y2I4OTUzYmQ0ZmRmYjQ4ODRiODhkMjUzMzg=
   data.tar.gz: !binary |-
-    MmRiZmFiOWM3NjI4NWMzYWE5NDA1MDU2YjZjMjIzNGMwMjhmYzcwZDc5Yzcw
-    Mjc0MDM4YmQ4ZGUxMTI4OTZmMjM0YzNiMDVhZjExNTMwNWQzODI0YzNjODNm
-    N2Q4YzE4MDAxMWM4YjAzNTc1YTI5NTgxOGJjZTZhNDQyNThjMGI=
+    YzQxZjcxMDVmYTQzZmRlODNjMWIyNmNlNDczNWUyMTQzNzM5MmRlZDk2ZDJh
+    ZDhkNmU3MzRmMTFiZjBlM2RiZjYzZmQ4MzQ1MjI1YThjZGVmYjdjMWZjZTI0
+    OGUxYzYwY2JlY2Q3NDMwOWQ1N2NlMjg1MDI1ZjU2NzU5NGQzOGE=

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    powertrack (1.0.3)
+    powertrack (1.1.0)
       em-http-request (~> 1.1)
       eventmachine (~> 1.0)
       exponential-backoff (~> 0.0.2)

data/History.txt

@@ -1,10 +1,15 @@
-1.0.3
------
+v1.1.0
+------
+
+  * Add support for Replay feature
+
+v1.0.3
+------
 
   * Retrieve the on_system function from options
 
-1.0.2
------
+v1.0.2
+------
 
   * to_json can get options in order for powertrack to support the standard
     JSON library (through the MultiJson gem). Thanks to @duncanita for the PR.

data/README.md

@@ -90,8 +90,8 @@ The ```:stop_timeout``` may be fine-tune when passing options to the tracker.
 
 As highly recommended by GNIP, the PowerTrack::Stream client manages an exponential
 backoff retry mechanism when a disconnection happens. The reconnections can be
-fine-tuned through the ```max_retries``` and ```backoff``` options passed to the
-```track``` call.
+fine-tuned through the ```:max_retries``` and ```:backoff``` options passed to
+the ```track``` call.
 
 ## Backfill
 
@@ -102,6 +102,19 @@ last 5 minutes when reconnecting.
 Provide a (numerical) client id as the last (but optional) argument of the
 PowerTrack::Stream constructor to enable this feature.
 
+## Replay
+
+Replay is a feature provided by GNIP to recover lost activities over the last
+5 days. The Replay stream lives aside the realtime stream and is activated
+by setting the ```:replay``` option to true when building a ```PowerTrack::Stream```
+object.
+
+Once Replay is activated, you use the stream as previously, starting by
+configuring some rules that define which activities you will recover. Once done,
+you can track the stream by specifying a timeframe with the ```:from```
+and ```:to options```. By default, replay happens over 30 minutes, starting 1
+hour ago.
+
 ## Errors
 
 All the errors that come from PowerTrack are defined through an ad-hoc exception

data/TODO.md

@@ -78,5 +78,6 @@ See [Managing disconnections](http://support.gnip.com/articles/disconnections-ex
 ## Other features
 
 * _[DONE]_ Support test and development streams
+* _[DONE]_ Support Replay mode (5-days back history)
 * Support status dashboard
 * Support Historical Powertrack

data/lib/powertrack/streaming/retrier.rb

@@ -56,7 +56,7 @@ module PowerTrack
     # retrier.
     def retry(&block)
       # TODO: manage exceptions
-      while @continue && @retries < @max_retries
+      while @continue && @retries <= @max_retries
         res = yield
         if @continue
           @retries += 1

data/lib/powertrack/streaming/stream.rb

@@ -18,7 +18,7 @@ module PowerTrack
     include VoidLogger::LoggerMixin
 
     # The format of the URLs to connect to the various stream services
-    FEATURE_URL_FORMAT = "https://%s:%s/accounts/%s/publishers/%s/streams/track/%s%s.json".freeze
+    FEATURE_URL_FORMAT = "https://%s:%s/accounts/%s/publishers/%s/%s/track/%s%s.json".freeze
 
     # The default timeout on a connection to PowerTrack. Can be overriden per call.
     DEFAULT_CONNECTION_TIMEOUT = 30
@@ -32,16 +32,22 @@ module PowerTrack
       connect_timeout: DEFAULT_CONNECTION_TIMEOUT,
       inactivity_timeout: DEFAULT_INACTIVITY_TIMEOUT,
       # use a client id if you want to leverage the Backfill feature
-      client_id: nil
+      client_id: nil,
+      # enable the replay mode to get activities over the last 5 days
+      # see http://support.gnip.com/apis/replay/api_reference.html
+      replay: false
     }
 
     DEFAULT_OK_RESPONSE_STATUS = 200
 
-    # the patterns used to identify the various types of message received from GNIP
+    # The patterns used to identify the various types of message received from GNIP
     # everything else is an activity
     HEARTBEAT_MESSAGE_PATTERN = /\A\s*\z/
     SYSTEM_MESSAGE_PATTERN = /\A\s*\{\s*"(info|warn|error)":/mi
 
+    # The format used to send UTC timestamps in Replay mode
+    REPLAY_TIMESTAMP_FORMAT = '%Y%m%d%H%M'
+
     attr_reader :username, :account_name, :data_source, :label
 
     def initialize(username, password, account_name, data_source, label, options=nil)
@@ -52,6 +58,8 @@ module PowerTrack
       @label = label
       @options = DEFAULT_STREAM_OPTIONS.merge(options || {})
       @client_id = @options[:client_id]
+      @replay = !!@options[:replay]
+      @stream_mode = @replay ? 'replay' : 'streams'
     end
 
     # Adds many rules to your PowerTrack stream’s ruleset.
@@ -105,7 +113,7 @@ module PowerTrack
       # receive GZip-compressed payloads ?
       compressed: true,
       # max number of retries after a disconnection
-      max_retries: 3,
+      max_retries: 2,
       # advanced options to configure exponential backoff used for retries
       backoff: nil,
       # max number of seconds to wait for last message handlers to complete
@@ -113,6 +121,10 @@ module PowerTrack
       # pass message in raw form (JSON formatted string) instead of JSON-decoded
       # Ruby objects to message handlers
       raw: false,
+      # the starting date from which the activities will be recovered (replay mode only)
+      from: nil,
+      # the ending date to which the activities will be recovered (replay mode only)
+      to: nil,
       # called for each message received, except heartbeats
       on_message: nil,
       # called for each activity received
@@ -158,6 +170,7 @@ module PowerTrack
                 gnip_server_port,
                 @account_name,
                 @data_source,
+                @stream_mode,
                 @label,
                 feature ]
 
@@ -187,6 +200,7 @@ module PowerTrack
     # Opens a new connection to GNIP PowerTrack.
     def connect(hostname, feature=nil)
       url = feature_url(hostname, feature)
+      logger.debug("Connecting to '#{url}' with headers #{connection_headers}...")
       EventMachine::HttpRequest.new(url, connection_headers)
     end
 
@@ -283,14 +297,14 @@ module PowerTrack
       handle_api_response(resp_status, resp_error, resp_body, options[:ok])
     end
 
-    # Returns the type of message received on the stream, nil when the type
-    # cannot be identified.
+    # Returns the type of message received on the stream, together with a
+    # level indicator in case of a system message, nil otherwise.
     def message_type(message)
       case message
-      when HEARTBEAT_MESSAGE_PATTERN then :heartbeat
-      when SYSTEM_MESSAGE_PATTERN then :system
+      when HEARTBEAT_MESSAGE_PATTERN then [ :heartbeat, nil ]
+      when SYSTEM_MESSAGE_PATTERN then [ :system, $1.downcase.to_sym ]
       else
-        :activity
+        [ :activity, nil ]
       end
     end
 
@@ -322,7 +336,25 @@ module PowerTrack
       EM.run do
         logger.info "Starting the reactor..."
         con = connect('stream')
-        http = con.get(head: track_req_headers(options[:compressed]))
+        get_opts = { head: track_req_headers(options[:compressed]) }
+
+        # add a timeframe in replay mode
+        if @replay
+          now = Time.now
+          # start 1 hour ago by default
+          from = options[:from] || (now - 60*60)
+          # stop 30 minutes ago by default
+          to = options[:to] || (now - 30*60)
+
+          get_opts[:query] = {
+            'fromDate' => from.utc.strftime(REPLAY_TIMESTAMP_FORMAT),
+            'toDate' => to.utc.strftime(REPLAY_TIMESTAMP_FORMAT)
+          }
+
+          logger.info "Replay mode enabled from '#{from}' to '#{to}'"
+        end
+
+        http = con.get(get_opts)
 
         # polls to see if the connection should be closed
         close_watcher = EM.add_periodic_timer(1) do
@@ -352,19 +384,21 @@ module PowerTrack
             next
           end
 
-          # reset retries when some (valid) data are received
-          if retrier.retrying?
-            logger.info "Resetting retries..."
-            retrier.reset!
-          end
-
           # process the chunk
           buffer.process(chunk) do |raw|
             logger.debug "New message received"
+
+            # get the message type and its (optional) level
+            m_type, m_level = message_type(raw)
+
+            # reset retries when some (valid) data are received
+            if retrier.retrying? && m_level != :error
+              logger.info "Resetting retries..."
+              retrier.reset!
+            end
+
             EM.defer do
               # select the right message handler(s) according to the message type
-              m_type = message_type(raw)
-
               if m_type == :heartbeat
                 on_heartbeat.call if on_heartbeat
               else
@@ -401,6 +435,7 @@ module PowerTrack
             resp_status = http_client.response_header.status || DEFAULT_OK_RESPONSE_STATUS
             resp_error = http_client.error
             resp_body = http_client.response
+
             wait_til_defers_finish_and_stop(stop_timeout)
           end
         end

data/lib/powertrack/version.rb

@@ -1,3 +1,3 @@
 module PowerTrack
-  VERSION = '1.0.3'
+  VERSION = '1.1.0'
 end

data/test/minitest_helper.rb

@@ -30,12 +30,13 @@ class Minitest::Test
   end
 
   # Returns a brand-new stream based on the config found in test/powertrack.yml.
-  def new_stream
+  def new_stream(replay=false)
     PowerTrack::Stream.new(
       powertrack_config[:username],
       powertrack_config[:password],
       powertrack_config[:account_name],
       powertrack_config[:data_source],
-      powertrack_config[:stream_label])
+      replay ? 'prod' : powertrack_config[:stream_label],
+      replay: replay)
   end
 end

data/test/test_manage_rules.rb

@@ -5,7 +5,15 @@ require 'multi_json'
 class TestManageRules < Minitest::Test
 
   def test_add_then_delete_a_single_rule
-    stream = new_stream
+    add_then_delete_a_single_rule(false)
+  end
+
+  def test_add_then_delete_a_single_rule_in_replay_mode
+    add_then_delete_a_single_rule(true)
+  end
+
+  def add_then_delete_a_single_rule(replay)
+    stream = new_stream(replay)
 
     rule = PowerTrack::Rule.new('coke')
     assert rule.valid?

data/test/test_track_stream.rb

@@ -4,8 +4,16 @@ require 'multi_json'
 
 class TestTrackStream < Minitest::Test
 
-  def test_track_simple_stream
-    stream = new_stream
+  def test_track_realtime_stream
+    track_simple_stream(false)
+  end
+
+  def test_track_replay_stream
+    track_simple_stream(true)
+  end
+
+  def track_simple_stream(replay)
+    stream = new_stream(replay)
 
     # add a logger
     stream.logger = Logger.new(STDERR)
@@ -23,6 +31,8 @@ class TestTrackStream < Minitest::Test
       received = 0
       tweeted = 0
       closed = false
+      from = nil
+      to = nil
 
       # ready to track
       on_message = lambda do |message|
@@ -34,28 +44,48 @@ class TestTrackStream < Minitest::Test
       on_activity = lambda do |tweet|
         tweeted += 1
       end
+      on_system = lambda do |message|
+        $stderr.puts message.inspect
+      end
 
       close_now = lambda { closed }
 
-      delay = 60
-      Thread.new do
-        $stderr.puts "Time-bomb thread running for #{delay} seconds..."
-        sleep delay
-        $stderr.puts "Time to shut down !"
-        closed = true
+      if replay
+        now = Time.now
+        from = now - 31*60
+        to = now - 30*60
+        delay = to - from
+      else
+        delay = 60
+        Thread.new do
+          $stderr.puts "Time-bomb thread running for #{delay} seconds..."
+          sleep delay
+          $stderr.puts "Time to shut down !"
+          closed = true
+        end
       end
 
       started_at = Time.now
       res = stream.track(on_message: on_message,
                          on_heartbeat: on_heartbeat,
                          on_activity: on_activity,
+                         on_system: on_system,
                          close_now: close_now,
-                         max_retries: 3,
-                         fake_disconnections: 20)
+                         max_retries: replay ? 0 : 2,
+                         fake_disconnections: replay ? nil : 20,
+                         from: from,
+                         to: to)
+
+      ended_at = Time.now
 
       assert_nil res
-      assert closed, 'Stream not closed'
-      assert Time.now - started_at >= delay
+      assert replay || closed, 'Stream not closed'
+
+      if replay
+        assert (ended_at - started_at) <= delay
+      else
+        assert (ended_at - started_at) >= delay
+      end
 
       assert heartbeats > 0, 'No heartbeat received'
       puts "#{heartbeats} heartbeats received"

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: powertrack
 version: !ruby/object:Gem::Version
-  version: 1.0.3
+  version: 1.1.0
 platform: ruby
 authors:
 - Laurent Farcy
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2016-02-22 00:00:00.000000000 Z
+date: 2016-07-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler