Class: FeedTools::FeedUpdater

Inherits:
Object
  • Object
show all
Defined in:
lib/feed_updater.rb

Overview

A simple daemon for scheduled updating of feeds.

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.on_begin(&block) ⇒ Object

Declares an on_begin event. The given block will be called before the update sequence runs to allow for any setup required. The block is not passed any arguments.



124
125
126
127
 
# File 'lib/feed_updater.rb', line 124

def self.on_begin(&block)
  raise "No block supplied for on_begin." if block.nil?
  @@on_begin = block
end

.on_complete(&block) ⇒ Object

Declares an on_complete event. The given block will be called after all feeds have been updated. The block is passed a list of feeds that FeedUpdater attempted to update.



148
149
150
151
 
# File 'lib/feed_updater.rb', line 148

def self.on_complete(&block)
  raise "No block supplied for on_complete." if block.nil?
  @@on_complete = block
end

.on_error(&block) ⇒ Object

Declares an on_error event. The given block will be called after an error occurs during a feed update. The block is passed the href of the feed that errored out, and the exception object that was raised.



140
141
142
143
 
# File 'lib/feed_updater.rb', line 140

def self.on_error(&block)
  raise "No block supplied for on_error." if block.nil?
  @@on_error = block
end

.on_update(&block) ⇒ Object

Declares an on_update event. The given block will be called after every feed update. The block is passed the feed object that was loaded and the time it took in seconds to successfully load it.



132
133
134
135
 
# File 'lib/feed_updater.rb', line 132

def self.on_update(&block)
  raise "No block supplied for on_update." if block.nil?
  @@on_update = block
end

Instance Method Details

#application ⇒ Object

Returns a reference to the daemon application.



277
278
279
280
 
# File 'lib/feed_updater.rb', line 277

def application()
  @application = nil if !defined?(@application)
  return @application
end

#daemon_options ⇒ Object

Returns a hash of the currently set daemon options.



262
263
264
265
266
267
268
269
270
271
272
273
274
 
# File 'lib/feed_updater.rb', line 262

def daemon_options()
  if !defined?(@daemon_options) || @daemon_options.nil?
    @daemon_options = {
      :app_name => "feed_updater_daemon",
      :dir_mode => :normal,
      :dir => self.pid_file_dir,
      :backtrace => true,
      :ontop => false
    }
  end
  @daemon_options[:dir] = self.pid_file_dir
  return @daemon_options
end

#feed_href_list ⇒ Object

Returns a list of feeds to be updated.



221
222
223
224
225
226
227
228
229
230
231
232
233
 
# File 'lib/feed_updater.rb', line 221

def feed_href_list()
  if !defined?(@feed_href_list) || @feed_href_list.nil?
    @feed_href_list_override = false
    if self.updater_options.nil?
      @feed_href_list = nil
    elsif self.updater_options[:feed_href_list].kind_of?(Array)
      @feed_href_list = self.updater_options[:feed_href_list]
    else
      @feed_href_list = nil
    end
  end
  return @feed_href_list
end

#feed_href_list=(new_feed_href_list) ⇒ Object

Sets a list of feeds to be updated.



236
237
238
239
 
# File 'lib/feed_updater.rb', line 236

def feed_href_list=(new_feed_href_list)
  @feed_href_list_override = true
  @feed_href_list = new_feed_href_list
end

#initial_directory ⇒ Object

Returns the initial directory that the daemon was started from.



154
155
156
157
 
# File 'lib/feed_updater.rb', line 154

def initial_directory()
  @initial_directory = nil if !defined?(@initial_directory)
  return @initial_directory
end

#log_file ⇒ Object

Returns the path to the log file.



182
183
184
185
186
187
188
189
190
191
192
 
# File 'lib/feed_updater.rb', line 182

def log_file()
  if !defined?(@log_file) || @log_file.nil?
    if self.log_file_dir.nil?
      @log_file = File.expand_path("./feed_updater.log")
    else
      @log_file = File.expand_path(
        self.log_file_dir + "/feed_updater.log")
    end
  end
  return @log_file
end

#log_file_dir ⇒ Object

Returns the directory where the log files are stored.



171
172
173
174
 
# File 'lib/feed_updater.rb', line 171

def log_file_dir()
  @log_file_dir = nil if !defined?(@log_file_dir)
  return @log_file_dir
end

#log_file_dir=(new_log_file_dir) ⇒ Object

Sets the directory where the log files are stored.



177
178
179
 
# File 'lib/feed_updater.rb', line 177

def log_file_dir=(new_log_file_dir)
  @log_file_dir = new_log_file_dir
end

#logger ⇒ Object

Returns the logger object.



195
196
197
198
199
200
201
202
203
204
 
# File 'lib/feed_updater.rb', line 195

def logger()
  if !defined?(@logger) || @logger.nil?
    @logger = FeedUpdaterLogger.new(self.log_file)
    @logger.level = self.updater_options[:log_level]
    @logger.datetime_format = nil
    @logger.progname = nil
    @logger.prefix = "FeedUpdater".ljust(20)
  end
  return @logger
end

#pid ⇒ Object

Returns the process id of the daemon. This should return nil if the daemon is not running.



284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
 
# File 'lib/feed_updater.rb', line 284

def pid()
  if File.exists?(File.expand_path(pid_file_dir + "/feed_updater.pid"))
    begin
      pid_file = File.open(
        File.expand_path(pid_file_dir + "/feed_updater.pid"), "r")
      return pid_file.read.to_s.strip.to_i
    rescue Exception
      return nil
    end
  else
    return nil if self.application.nil?
    begin
      return self.application.pid.pid
    rescue Exception
      return nil
    end
  end
end

#pid_file_dir ⇒ Object

Returns the directory where the pid files are stored.



160
161
162
163
 
# File 'lib/feed_updater.rb', line 160

def pid_file_dir()
  @pid_file_dir = nil if !defined?(@pid_file_dir)
  return @pid_file_dir
end

#pid_file_dir=(new_pid_file_dir) ⇒ Object

Sets the directory where the pid files are stored.



166
167
168
 
# File 'lib/feed_updater.rb', line 166

def pid_file_dir=(new_pid_file_dir)
  @pid_file_dir = new_pid_file_dir
end

#progress_precentage ⇒ Object 



518
519
520
521
522
523
524
525
526
527
528
529
530
 
# File 'lib/feed_updater.rb', line 518

def progress_precentage()
  if !defined?(@remaining_href_list) || !defined?(@feed_href_list)
    return nil
  end
  if @remaining_href_list.nil? || @feed_href_list.nil?
    return nil
  end
  if @feed_href_list == []
    return nil
  end
  return 100.0 - (100.0 *
    (@remaining_href_list.size.to_f / @feed_href_list.size.to_f))
end

#restart ⇒ Object

Restarts the daemon.



513
514
515
516
 
# File 'lib/feed_updater.rb', line 513

def restart()
  self.stop()
  self.start()
end

#restart_logger ⇒ Object

Restarts the logger object. This needs to be done after the program forks.



208
209
210
211
212
213
214
215
216
217
218
 
# File 'lib/feed_updater.rb', line 208

def restart_logger()
  begin
    self.logger.close()
  rescue IOError
  end
  @logger = FeedUpdaterLogger.new(self.log_file)
  @logger.level = self.updater_options[:log_level]
  @logger.datetime_format = nil
  @logger.progname = nil
  @logger.prefix = "FeedUpdater".ljust(20)
end

#start ⇒ Object

Starts the daemon.



314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
 
# File 'lib/feed_updater.rb', line 314

def start()
  self.logger.prefix = "FeedUpdater".ljust(20)
  if !defined?(@@on_update) || @@on_update.nil?
    raise "No on_update handler block given."
  end
  if self.status == :running || self.pid != nil
    puts "FeedUpdater is already running."
    return self.pid
  end
  if defined?(ActiveRecord)
    ActiveRecord::Base.allow_concurrency = true
  end

  @initial_directory = File.expand_path(".")

  if @application.nil?
    self.logger()
    feed_update_proc = lambda do
      # Reestablish correct location
      Dir.chdir(File.expand_path(self.initial_directory))

      sleep(1)

      self.restart_logger()
      self.logger.info("Using environment: #{FEED_TOOLS_ENV}")
      
      if FeedTools.configurations[:feed_cache].nil?
        FeedTools.configurations[:feed_cache] =
          "FeedTools::DatabaseFeedCache"
      end

      if !FeedTools.feed_cache.connected?
        FeedTools.configurations[:feed_cache] =
          "FeedTools::DatabaseFeedCache"
        FeedTools.feed_cache.initialize_cache()
        begin
          ActiveRecord::Base.default_timezone = :utc
          ActiveRecord::Base.connection
        rescue Exception
          config_path = FileStalker.hunt([
            "./config/database.yml",
            "../config/database.yml",
            "../../config/database.yml",
            "./database.yml",
            "../database.yml",
            "../../database.yml"
          ])
          if config_path != nil
            require 'yaml'
            config_hash = File.open(config_path) do |file|
              config_hash = YAML.load(file.read)
              unless config_hash[FEED_TOOLS_ENV].nil?
                config_hash = config_hash[FEED_TOOLS_ENV]
              end
              config_hash
            end
            self.logger.info(
              "Attempting to manually load the database " +
              "connection described in:")
            self.logger.info(config_path)
            begin
              ActiveRecord::Base.configurations = config_hash
              ActiveRecord::Base.establish_connection(config_hash)
              ActiveRecord::Base.connection
            rescue Exception => error
              self.logger.fatal(
                "Error initializing database:")
              self.logger.fatal(error)
              exit
            end
          else
            self.logger.fatal(
              "The database.yml file does not appear " +
              "to exist.")
            exit
          end
        end
      end
      if !FeedTools.feed_cache.connected?
        self.logger.fatal("Not connected to the feed cache.")
        if FeedTools.feed_cache.respond_to?(:table_exists?)
          if !FeedTools.feed_cache.table_exists?
            self.logger.fatal(
              "The FeedTools cache table is missing.")
          end
        end
        exit
      end
      
      # A random start delay is introduced so that we don't have multiple
      # feed updater daemons getting kicked off at the same time by
      # multiple users.
      if self.updater_options[:start_delay]
        delay = (rand(45) + 15)
        self.logger.info("Startup delay set for #{delay} minutes.")
        sleep(delay.minutes)
      end
      
      # The main feed update loop.
      loop do
        result = nil
        sleepy_time = self.updater_options[:sleep_time].to_i.minutes
        begin
          result = Benchmark.measure do
            self.update_feeds()
          end
          self.logger.info(
            "#{@feed_href_list.size} feed(s) updated " +
            "in #{result.real.round} seconds.")
          sleepy_time =
            (self.updater_options[:sleep_time].to_i.minutes -
              result.real.round)
        rescue Exception => error
          self.logger.error("Feed update sequence errored out.")
          self.logger.error(error.class.name + ": " + error.message)
          self.logger.error("\n" + error.backtrace.join("\n").to_s)
        end

        @feed_href_list = nil
        @feed_href_list_override = false
        ObjectSpace.garbage_collect()
        if sleepy_time > 0
          self.logger.info(
            "Sleeping for #{(sleepy_time / 1.minute.to_f).round} " +
            "minutes...")
          sleep(sleepy_time)
        else
          self.logger.info(
            "Update took more than " +
            "#{self.updater_options[:sleep_time].to_i} minutes, " +
            "restarting immediately.")
        end
      end
    end
    options = self.daemon_options.merge({
      :proc => feed_update_proc,
      :mode => :proc
    })
    @application = Daemons::ApplicationGroup.new(
      'feed_updater', options).new_application(options)
    @application.start()
  else
    @application.start()
  end
  self.logger.prefix = nil
  self.logger.level = 0
  self.logger.info("-" * 79)
  self.logger.info("Daemon started, " +
    "FeedUpdater #{FeedTools::FEED_UPDATER_VERSION::STRING} / " +
    "FeedTools #{FeedTools::FEED_TOOLS_VERSION::STRING}")
  self.logger.info(" @ #{Time.now.utc.to_s}")
  self.logger.info("-" * 79)
  self.logger.level = self.updater_options[:log_level]
  self.logger.prefix = "FeedUpdater".ljust(20)
  @status = :running
  return self.pid
end

#status ⇒ Object

Returns either :running or :stopped depending on the daemon’s current status.



243
244
245
246
 
# File 'lib/feed_updater.rb', line 243

def status()
  @status = :stopped if @status.nil?
  return @status
end

#stop ⇒ Object

Stops the daemon.



473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
 
# File 'lib/feed_updater.rb', line 473

def stop()
  if self.pid.nil?
    puts "FeedUpdater isn't running."
  end
  begin
    # Die.
    Process.kill('TERM', self.pid)
  rescue Exception
  end
  begin
    # No, really, I mean it.  You need to die.
    system("kill #{self.pid} 2> /dev/null")
    
    # Perhaps I wasn't clear somehow?
    system("kill -9 #{self.pid} 2> /dev/null")
  rescue Exception
  end
  begin
    # Clean the pid file up.
    if File.exists?(
        File.expand_path(self.pid_file_dir + "/feed_updater.pid"))
      File.delete(
        File.expand_path(self.pid_file_dir + "/feed_updater.pid"))
    end
  rescue Exception
  end
  self.logger.prefix = "FeedUpdater".ljust(20)
  self.logger.level = 0
  self.logger.info("Daemon stopped.")
  self.logger.level = self.updater_options[:log_level]
  @status = :stopped
  return nil if self.application.nil?
  begin
    self.application.stop()
  rescue Exception
  end
  return nil
end

#update_feeds ⇒ Object

Updates all of the feeds.



533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
 
# File 'lib/feed_updater.rb', line 533

def update_feeds()
  self.logger.level = 0
  self.logger.prefix = "FeedUpdater".ljust(20)
  ObjectSpace.garbage_collect()
  if defined?(@@on_begin) && @@on_begin != nil
    self.logger.info("Running custom startup event...")
    self.cloaker(&(@@on_begin)).bind(self).call()
  end
  if defined?(@feed_href_list_override) && @feed_href_list_override
    self.logger.info("Using custom feed list...")
    self.feed_href_list()
  else
    self.logger.info("Loading default feed list...")
    begin
      expire_time = (Time.now - 1.hour).utc
      expire_time_string = sprintf('%04d-%02d-%02d %02d:%02d:%02d',
        expire_time.year, expire_time.month, expire_time.day,
        expire_time.hour, expire_time.min, expire_time.sec)
      @feed_href_list =
        FeedTools.feed_cache.connection.execute(
          "SELECT href FROM cached_feeds WHERE " +
          "last_retrieved < '#{expire_time_string}'").to_a.flatten
    rescue Exception
      self.logger.warn("Default feed list failed, using fallback.")
      @feed_href_list =
        FeedTools.feed_cache.find(:all).collect do |feed|
          feed.href
        end
      self.logger.warn(
        "Fallback succeeded. Custom feed list override recommended.")
    end
  end
  self.logger.info("Updating #{@feed_href_list.size} feed(s)...")
  self.logger.level = self.updater_options[:log_level]
  
  @threads = []
  @remaining_href_list = @feed_href_list.dup

  ObjectSpace.garbage_collect()

  begin_updating = false
  self.logger.info(
    "Starting up #{self.updater_options[:threads]} thread(s)...")
  
  mutex = Mutex.new
  for i in 0...self.updater_options[:threads]
    updater_thread = Thread.new do
      self.logger.level = self.updater_options[:log_level]
      self.logger.datetime_format = "%s"
      self.logger.progname = "FeedUpdater".ljust(20)
      
      while !Thread.current.respond_to?(:thread_id) &&
          begin_updating == false
        Thread.pass
      end
      mutex.synchronize do
        self.logger.prefix =
          "Thread #{Thread.current.thread_id} ".ljust(20)
        self.logger.info("Thread started.")
        
        begin
          FeedTools.feed_cache.initialize_cache()
          if !FeedTools.feed_cache.set_up_correctly?
            self.logger.info("Problem with cache connection...")
          end
        rescue Exception => error
          self.logger.info(error)
        end
      end
      
      ObjectSpace.garbage_collect()
      Thread.pass

      while @remaining_href_list.size > 0
        progress = nil
        mutex.synchronize do
          Thread.current.progress = self.progress_precentage()
          Thread.current.href = @remaining_href_list.shift()
          progress = sprintf("%.2f", Thread.current.progress)
        end
        begin
          begin
            Thread.current.feed = nil
            feed_load_benchmark = Benchmark.measure do
              Thread.current.feed =
                FeedTools::Feed.open(Thread.current.href)
            end
            Thread.pass
            if Thread.current.feed.live?
              unless @@on_update.nil?
                mutex.synchronize do
                  progress = sprintf("%.2f", Thread.current.progress)
                  self.logger.prefix = 
                    ("Thread #{Thread.current.thread_id} (#{progress}%)"
                      ).ljust(20)
                  self.cloaker(&(@@on_update)).bind(self).call(
                    Thread.current.feed, feed_load_benchmark.real)
                end
              end
            else
              mutex.synchronize do
                progress = sprintf("%.2f", Thread.current.progress)
                self.logger.prefix = 
                  ("Thread #{Thread.current.thread_id} (#{progress}%)"
                    ).ljust(20)
                self.logger.info(
                  "'#{Thread.current.href}' unchanged " +
                  "or unavailable, skipping.")
              end
            end
          rescue Exception => error
            mutex.synchronize do
              progress = sprintf("%.2f", Thread.current.progress)
              self.logger.prefix = 
                ("Thread #{Thread.current.thread_id} (#{progress}%)"
                  ).ljust(20)
              if @@on_error != nil
                self.cloaker(&(@@on_error)).bind(self).call(
                  Thread.current.href, error)
              else
                self.logger.error(
                  "Error updating '#{Thread.current.href}':")
                self.logger.error(error.class.name + ": " + error.message)
                self.logger.error(error.class.backtrace)
              end
            end
          end
        rescue Exception => error
          mutex.synchronize do
            progress = sprintf("%.2f", Thread.current.progress)
            self.logger.prefix = ("Thread #{Thread.current.thread_id} " +
              "(#{progress}%)"
              ).ljust(20)
            self.logger.fatal("Critical unhandled error.")
            self.logger.fatal("Error updating '#{Thread.current.href}':")
            self.logger.fatal(error.class.name + ": " + error.message)
            self.logger.fatal(error.class.backtrace)
          end
        end
        ObjectSpace.garbage_collect()
        Thread.pass
      end          
    end
    @threads << updater_thread        
    class <<updater_thread
      attr_accessor :thread_id
      attr_accessor :progress
      attr_accessor :href
      attr_accessor :feed
    end
    updater_thread.thread_id = i
  end
  mutex.synchronize do
    self.logger.prefix = "FeedUpdater".ljust(20)
    self.logger.info(
      "#{@threads.size} thread(s) successfully started...")
    begin_updating = true
  end
  Thread.pass
  
  ObjectSpace.garbage_collect()
  Thread.pass
  for i in 0...@threads.size
    mutex.synchronize do
      self.logger.prefix = "FeedUpdater".ljust(20)
      self.logger.info(
        "Joining on thread #{@threads[i].thread_id}...")
    end
    @threads[i].join
  end
  self.logger.prefix = "FeedUpdater".ljust(20)
  ObjectSpace.garbage_collect()
  
  self.logger.progname = nil
  unless @@on_complete.nil?
    self.cloaker(&(@@on_complete)).bind(self).call(@feed_href_list)
  end
  self.logger.level = 0
  self.logger.info("Finished updating.")
  self.logger.level = self.updater_options[:log_level]
end

#updater_options ⇒ Object

Returns a hash of the currently set updater options.



249
250
251
252
253
254
255
256
257
258
259
 
# File 'lib/feed_updater.rb', line 249

def updater_options()
  if !defined?(@updater_options) || @updater_options.nil?
    @updater_options = {
      :start_delay => true,
      :threads => 1,
      :sleep_time => 60,
      :log_level => 0
    }
  end
  return @updater_options
end