Module: BlackStack::Netting

Defined in:
lib/functions.rb

Overview

Network

Defined Under Namespace

Classes: ApiCallException

Constant Summary collapse

CALL_METHOD_GET = 
'get'
CALL_METHOD_POST = 
'post'
DEFAULT_SSL_VERIFY_MODE = 
OpenSSL::SSL::VERIFY_NONE
SUCCESS = 
'success'
@@lockfiles = 
[]
@@max_api_call_channels = 
10

Class Method Summary collapse

Class Method Details

.add_param(url, param_name, param_value) ⇒ Object

Add a parameter to the url. It doesn’t validate if the param already exists.



673
674
675
676
677
678
679
680
681
682
683
684
685
 
# File 'lib/functions.rb', line 673

def self.add_param(url, param_name, param_value)
  uri = URI(url)
  params = URI.decode_www_form(uri.query || '')
  
  if (params.size==0)
    params << [param_name, param_value]
    uri.query = URI.encode_www_form(params)
    return uri.to_s
  else
    uri.query = URI.encode_www_form(params)
    return uri.to_s + "&" + param_name + "=" + param_value    
  end
end

.api_call(url, params = {}, method = BlackStack::Netting::CALL_METHOD_POST, ssl_verify_mode = BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE, max_retries = 5) ⇒ Object 



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
 
# File 'lib/functions.rb', line 552

def self.api_call(url, params={}, method=BlackStack::Netting::CALL_METHOD_POST, ssl_verify_mode=BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE, max_retries=5)
  nTries = 0
  bSuccess = false
  parsed = nil
  sError = ""
  while (nTries < max_retries && bSuccess == false)
    begin
      nTries = nTries + 1
      uri = URI(url)
      res = BlackStack::Netting::call_post(uri, params, ssl_verify_mode) if method==BlackStack::Netting::CALL_METHOD_POST
      res = BlackStack::Netting::call_get(uri, params, ssl_verify_mode) if method==BlackStack::Netting::CALL_METHOD_GET
      parsed = JSON.parse(res.body)
      if (parsed['status']==BlackStack::Netting::SUCCESS)
        bSuccess = true
      else
        sError = "Status: #{parsed['status'].to_s}. Description: #{parsed['value'].to_s}." 
      end
    rescue Errno::ECONNREFUSED => e
      sError = "Errno::ECONNREFUSED:" + e.to_console
    rescue => e2
      sError = "Exception:" + e2.to_console
    end
  end # while

  if (bSuccess==false)
    raise "#{sError}"
  end
end

.call_get(url, params = {}, ssl_verify_mode = BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE) ⇒ Object

New call_get



487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
 
# File 'lib/functions.rb', line 487

def self.call_get(url, params = {}, ssl_verify_mode=BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE) 
  uri = URI(url)
  uri.query = URI.encode_www_form(params)
  Net::HTTP.start(uri.host, uri.port, :use_ssl => uri.scheme == 'https', :verify_mode => ssl_verify_mode) do |http|
    req = Net::HTTP::Get.new uri
    #req.body = body if !body.nil?
    res = http.request req
    case res
    when Net::HTTPSuccess then res
    when Net::HTTPRedirection then BlackStack::Netting::call_get(URI(res['location']), params)
    else
      res.error!
    end
  end
end

.call_post(url, params = {}, ssl_verify_mode = BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE, use_lockfile = true) ⇒ Object

Call the API and return th result. url: valid internet address params: hash of params to attach in the call ssl_verify_mode: you can disabele SSL verification here. max_channels: this method use lockfiles to prevent an excesive number of API calls from each datacenter. There is not allowed more simultaneous calls than max_channels. TODO: setup max_simultaneus_calls in the configurtion file.



509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
 
# File 'lib/functions.rb', line 509

def self.call_post(url, params = {}, ssl_verify_mode=BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE, use_lockfile=true)
  # build the lockfile name
#      raise "Max Channels cannot be higher than #{BlackStack::Netting.lockfiles.size.to_s}" if BlackStack::Netting.max_api_call_channels > BlackStack::Netting.lockfiles.size
#      x = rand(BlackStack::Netting.max_api_call_channels)
#      # lock the file
#      BlackStack::Netting.lockfiles[x].flock(File::LOCK_EX) if use_lockfile
  
  begin
    
    # do the call
    uri = URI(url)
    ret = Net::HTTP.start(uri.host, uri.port, :use_ssl => uri.scheme == 'https', :verify_mode => ssl_verify_mode) do |http|
      req = Net::HTTP::Post.new(uri)
      req['Content-Type'] = 'application/json'
      req.set_form_data(params)
      #req.body = body if !body.nil?
      res = http.request req
      case res 
      when Net::HTTPSuccess then res
      when Net::HTTPRedirection then BlackStack::Netting::call_post(URI(res['location']), params, BlackStack::Netting::DEFAULT_SSL_VERIFY_MODE, false)
      else
        res.error!
      end
    end
    
    # release the file
#        BlackStack::Netting.lockfiles[x].flock(File::LOCK_UN) if use_lockfile
  rescue => e
    # release the file
#        BlackStack::Netting.lockfiles[x].flock(File::LOCK_UN) if use_lockfile
    
    # elevo la excepcion
    raise e
  ensure
    # release the file
#        BlackStack::Netting.lockfiles[x].flock(File::LOCK_UN) if use_lockfile
  end
  
  # return 
  ret
end

.change_param(url, param_name, param_value) ⇒ Object

Changes the value of a parameter in the url. It doesn’t validate if the param already exists.



688
689
690
691
692
693
694
695
 
# File 'lib/functions.rb', line 688

def self.change_param(url, param_name, param_value)
  uri = URI(url)
#  params = URI.decode_www_form(uri.query || [])
  params = CGI.parse(uri.query)
  params["start"] = param_value
  uri.query = URI.encode_www_form(params)
  uri.to_s
end

.download(url, to) ⇒ Object

Download a file from an url to a local folder. url: must be somedomain.net instead of somedomain.net/, otherwise, it will throw exception. to: must be a valid path to a folder.



584
585
586
587
588
589
590
591
592
593
594
595
 
# File 'lib/functions.rb', line 584

def self.download(url, to)
  uri = URI(url)
  domain = uri.host.start_with?('www.') ? uri.host[4..-1] : uri.host    
  path = uri.path
  filename = path.split("/").last
  Net::HTTP.start(domain) do |http|
    resp = http.get(path)
    open(to, "wb") do |file|
      file.write(resp.body)
    end
  end
end

.file_age(filename) ⇒ Object

returns the age in days of the given file



620
621
622
 
# File 'lib/functions.rb', line 620

def self.file_age(filename)
  (Time.now - File.ctime(filename))/(24*3600)
end

.get_host_without_www(url) ⇒ Object

Removes the ‘www.’ from an URL.



604
605
606
607
608
 
# File 'lib/functions.rb', line 604

def self.get_host_without_www(url)
  url = "http://#{url}" if URI.parse(url).scheme.nil?
  host = URI.parse(url).host.downcase
  host.start_with?('www.') ? host[4..-1] : host
end

.get_redirect(url) ⇒ Object

Get the final URL if a web page is redirecting.



611
612
613
614
615
616
617
 
# File 'lib/functions.rb', line 611

def self.get_redirect(url)
  uri = URI.parse(url)
  protocol = uri.scheme
  host = uri.host.downcase
  res = Net::HTTP.get_response(uri)
  "#{protocol}://#{host}#{res['location']}"
end

.get_url_extension(url) ⇒ Object

Return the extension of the last path into an URL. Example: get_url_extension(“connect.data.com/sitemap_index.xml?foo_param=foo_value”) => “.xml”



599
600
601
 
# File 'lib/functions.rb', line 599

def self.get_url_extension(url)
  return File.extname(URI.parse(url).path.to_s)
end

.getDomainFromEmail(email) ⇒ Object 



725
726
727
728
729
730
731
 
# File 'lib/functions.rb', line 725

def self.getDomainFromEmail(email)
  if email.email?
    return email.split("@").last
  else
    raise "getDomainFromEmail: Wrong email format."
  end
end

.getDomainFromUrl(url) ⇒ Object

get the domain from any url



709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
 
# File 'lib/functions.rb', line 709

def self.getDomainFromUrl(url)
  if (url !~ /^http:\/\//i && url !~ /^https:\/\//i) 
    url = "http://#{url}"
  end
  
  if (URI.parse(url).host == nil)
    raise "Cannot get domain for #{url}" 
  end

  if (url.to_s.length>0)
    return URI.parse(url).host.sub(/^www\./, '')
  else
    return nil
  end
end

.getWhoisDomains(domain, allow_heuristic_to_avoid_hosting_companies = false) ⇒ Object 



733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
 
# File 'lib/functions.rb', line 733

def self.getWhoisDomains(domain, allow_heuristic_to_avoid_hosting_companies=false)
  a = Array.new
  c = Whois::Client.new
  r = c.lookup(domain)

  res = r.to_s.scan(/Registrant Email: (#{BlackStack::Strings::MATCH_EMAIL})/).first
  if (res!=nil)
    a << BlackStack::Netting::getDomainFromEmail(res[0].downcase)
  end

  res = r.to_s.scan(/Admin Email: (#{BlackStack::Strings::MATCH_EMAIL})/).first
  if (res!=nil)
    a << BlackStack::Netting::getDomainFromEmail(res[0].downcase)
  end

  res = r.to_s.scan(/Tech Email: (#{BlackStack::Strings::MATCH_EMAIL})/).first
  if (res!=nil)
    a << BlackStack::Netting::getDomainFromEmail(res[0].downcase)
  end

  # remover duplicados
  a = a.uniq

  # 
  if (allow_heuristic_to_avoid_hosting_companies==true)
    # TODO: develop this feature
  end

  return a
end

.lockfiles ⇒ Object 



458
459
460
 
# File 'lib/functions.rb', line 458

def self.lockfiles()
  @@lockfiles
end

.max_api_call_channels ⇒ Object 



454
455
456
 
# File 'lib/functions.rb', line 454

def self.max_api_call_channels()
  @@max_api_call_channels
end

.params(url) ⇒ Object

returns a hash with the parametes in the url



660
661
662
663
664
665
666
667
668
669
670
 
# File 'lib/functions.rb', line 660

def self.params(url)
  # TODO: Corregir este parche:
  # => El codigo de abajo usa la URL de una busqueda en google. Esta url generara una excepcion cuando se intenta parsear sus parametros.
  # => Ejecutar las 2 lineas de abajo para verificar.
  # => url = "https://www.google.com/webhp#q=[lead+generation]+%22John%22+%22Greater+New+York+City+Area+*+Financial+Services%22+site:linkedin.com%2Fpub+-site:linkedin.com%2Fpub%2Fdir"
  # => p = CGI::parse(URI.parse(url).query)
  # => La linea de abajo hace un gsbub que hace que esta url siga funcionando como busqueda de google, y ademas se posible parsearla.
  url = url.gsub("webhp#q=", "webhp?q=")

  return CGI::parse(URI.parse(url).query)
end

.set(h) ⇒ Object 



462
463
464
465
466
467
468
469
470
471
 
# File 'lib/functions.rb', line 462

def self.set(h)
  @@max_api_call_channels = h[:max_api_call_channels]
  @@lockfiles = []
  
  i = 0
  while i<@@max_api_call_channels
    @@lockfiles << File.open("./apicall.channel_#{i.to_s}.lock", "w")
    i+=1
  end
end

.set_param(url, param_name, param_value) ⇒ Object

Change or add the value of a parameter in the url, depending if the parameter already exists or not.



698
699
700
701
702
703
704
705
706
 
# File 'lib/functions.rb', line 698

def self.set_param(url, param_name, param_value)
  params = BlackStack::Netting::params(url) 
  if ( params.has_key?(param_name) == true )
    newurl = BlackStack::Netting::change_param(url, param_name, param_value)
  else
    newurl = BlackStack::Netting::add_param(url, param_name, param_value)
  end
  return newurl
end