Class: CobwebModule::Crawl

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

Instance Method Summary collapse

Constructor Details

#initialize(options = {}) ⇒ Crawl

Returns a new instance of Crawl.



4
5
6
7
8
9
10
11
12
13
14
 
# File 'lib/crawl.rb', line 4

def initialize(options={})
  @options = HashUtil.deep_symbolize_keys(options)

  setup_defaults

  @redis = Redis::Namespace.new("cobweb-#{Cobweb.version}-#{@options[:crawl_id]}", :redis => Redis.new(@options[:redis_options]))
  @stats = Stats.new(@options)
  @debug = @options[:debug]
  @first_to_finish = false

end

Instance Method Details

#already_crawled?(link = @options[:url]) ⇒ Boolean

Returns true if the url requested is already in the crawled queue

Returns:

  • (Boolean) 


17
18
19
 
# File 'lib/crawl.rb', line 17

def already_crawled?(link=@options[:url])
  @redis.sismember "crawled", link
end

#already_queued?(link) ⇒ Boolean

Returns:

  • (Boolean) 


21
22
23
 
# File 'lib/crawl.rb', line 21

def already_queued?(link)
  @redis.sismember "queued", link
end

#content ⇒ Object 



125
126
127
128
 
# File 'lib/crawl.rb', line 125

def content
  raise "Content is not available" if @content.nil?
  CobwebModule::CrawlObject.new(@content, @options)
end

#crawled_base_url ⇒ Object 



213
214
215
 
# File 'lib/crawl.rb', line 213

def crawled_base_url
  @redis.get("crawled_base_url")
end

#debug_ap(value) ⇒ Object 



246
247
248
 
# File 'lib/crawl.rb', line 246

def debug_ap(value)
  ap(value) if @options[:debug]
end

#debug_puts(value) ⇒ Object 



250
251
252
 
# File 'lib/crawl.rb', line 250

def debug_puts(value)
  puts(value) if @options[:debug]
end

#finished ⇒ Object 



189
190
191
192
 
# File 'lib/crawl.rb', line 189

def finished
  set_first_to_finish
  @stats.end_crawl(@options)
end

#finished? ⇒ Boolean

Returns:

  • (Boolean) 


174
175
176
177
178
179
180
181
182
183
184
185
186
187
 
# File 'lib/crawl.rb', line 174

def finished?
  print_counters
  # if there's nothing left queued or the crawled limit has been reached and we're not still processing something
  if @options[:crawl_limit].nil? || @options[:crawl_limit] == 0
    if queue_counter == 0 && @redis.smembers("currently_running").empty?
      finished
      return true
    end
  elsif (queue_counter == 0 && @redis.smembers("currently_running").empty?) || process_counter >= @options[:crawl_limit].to_i
    finished
    return true
  end
  false
end

#finished_processing ⇒ Object 



170
171
172
 
# File 'lib/crawl.rb', line 170

def finished_processing
  @redis.srem "currently_running", @options[:url]
end

#first_to_finish? ⇒ Boolean

Returns:

  • (Boolean) 


209
210
211
 
# File 'lib/crawl.rb', line 209

def first_to_finish?
  @first_to_finish
end

#lock(key, &block) ⇒ Object 



225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
 
# File 'lib/crawl.rb', line 225

def lock(key, &block)
  debug_puts "REQUESTING LOCK [#{key}]"
  set_nx = @redis.setnx("#{key}_lock", "locked")
  debug_puts "LOCK:#{key}:#{set_nx}"
  while !set_nx
    debug_puts "===== WAITING FOR LOCK [#{key}] ====="
    sleep 0.01
    set_nx = @redis.setnx("#{key}_lock", "locked")
  end

  debug_puts "RECEIVED LOCK [#{key}]"
  @redis.expire("#{key}_lock", 10)
  begin
    result = yield
  ensure
    @redis.del("#{key}_lock")
    debug_puts "LOCK RELEASED [#{key}]"
  end
  result
end

#process(&block) ⇒ Object 



156
157
158
159
160
161
162
163
164
165
166
167
168
 
# File 'lib/crawl.rb', line 156

def process(&block)
  if @options[:crawl_limit_by_page]
    if content.mime_type.match("text/html")
      increment_process_counter
    end
  else
    increment_process_counter
  end
  @redis.sadd "enqueued", @options[:url]

  yield if block_given?
  @redis.incr("crawl_job_enqueued_count")
end

#process_links(&block) ⇒ Object 



84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
 
# File 'lib/crawl.rb', line 84

def process_links &block

  # set the base url if this is the first page
  set_base_url @redis

  @cobweb_links = CobwebLinks.new(@options)
  if within_queue_limits?
    document_links = ContentLinkParser.new(@options[:url], content.body, @options).all_links(:valid_schemes => [:http, :https])
    #get rid of duplicate links in the same page.
    document_links.uniq!
    
    # select the link if its internal
    internal_links = document_links.select{ |link| @cobweb_links.internal?(link) }

    # reject the link if we've crawled it or queued it
    internal_links.reject! { |link| @redis.sismember("crawled", link) }
    internal_links.reject! { |link| @redis.sismember("queued", link) }

    internal_links.each do |link|
      if within_queue_limits? && !already_queued?(link) && !already_crawled?(link)
        if status != CobwebCrawlHelper::CANCELLED
          yield link if block_given?
          unless link.nil?
            @redis.sadd "queued", link
            increment_queue_counter
          end
        else
          debug_puts "Cannot enqueue new content as crawl has been cancelled."
        end
      end
    end

    if @options[:store_inbound_links]
      document_links.each do |link|
        uri = URI.parse(link)
        @redis.sadd("inbound_links_#{Digest::MD5.hexdigest(uri.to_s)}", url)
      end
    end
  end
end

#redis ⇒ Object 



221
222
223
 
# File 'lib/crawl.rb', line 221

def redis
  @redis
end

#retrieve ⇒ Object 



52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
 
# File 'lib/crawl.rb', line 52

def retrieve
  unless @redis.sismember("currently_running", @options[:url])
    @redis.sadd("currently_running", @options[:url])
    unless already_crawled?
      if within_crawl_limits?
        @stats.update_status("Retrieving #{@options[:url]}...")
        @content = Cobweb.new(@options).get(@options[:url], @options)
        if @options[:url] == @redis.get("original_base_url")
          @redis.set("crawled_base_url", @content[:base_url])
        end
        update_queues

        if content.permitted_type?
          ## update statistics

          @stats.update_statistics(@content)
          return true
        end
      else
        decrement_queue_counter
      end
    else
      decrement_queue_counter
    end
  else
    debug_puts "\n\nDETECTED DUPLICATE JOB for #{@options[:url]}\n"
    debug_ap @redis.smembers("currently_running")
    decrement_queue_counter
  end
  false
end

#set_first_to_finish ⇒ Object 



194
195
196
197
198
199
200
201
202
203
204
205
206
 
# File 'lib/crawl.rb', line 194

def set_first_to_finish
  @redis.watch("first_to_finish") do
    if !@redis.exists("first_to_finish")
      @redis.multi do
        debug_puts "set first to finish"
        @first_to_finish = true
        @redis.set("first_to_finish", 1)
      end
    else
      @redis.unwatch
    end
  end
end

#statistics ⇒ Object 



217
218
219
 
# File 'lib/crawl.rb', line 217

def statistics
  @stats.get_statistics
end

#to_be_processed? ⇒ Boolean

Returns:

  • (Boolean) 


152
153
154
 
# File 'lib/crawl.rb', line 152

def to_be_processed?
  (!finished? || within_process_limits?) && !@redis.sismember("enqueued", @options[:url])
end

#update_queues ⇒ Object 



130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
 
# File 'lib/crawl.rb', line 130

def update_queues
  lock("update_queues") do
    #@redis.incr "inprogress"
    # move the url from the queued list to the crawled list - for both the original url, and the content url (to handle redirects)
    @redis.srem "queued", @options[:url]
    @redis.sadd "crawled", @options[:url]
    if content.url != @options[:url]
      @redis.srem "queued", content.url
      @redis.sadd "crawled", content.url
    end
    # increment the counter if we are not limiting by page only || we are limiting count by page and it is a page
    if @options[:crawl_limit_by_page]
      if content.mime_type.match("text/html")
        increment_crawl_counter
      end
    else
      increment_crawl_counter
    end
    decrement_queue_counter
  end
end

#within_crawl_limits? ⇒ Boolean

Returns true if the crawl count is within limits

Returns:

  • (Boolean) 


26
27
28
 
# File 'lib/crawl.rb', line 26

def within_crawl_limits?
  @options[:crawl_limit].nil? || crawl_counter < @options[:crawl_limit].to_i
end

#within_process_limits? ⇒ Boolean

Returns true if the processed count is within limits

Returns:

  • (Boolean) 


31
32
33
 
# File 'lib/crawl.rb', line 31

def within_process_limits?
  @options[:crawl_limit].nil? || process_counter < @options[:crawl_limit].to_i
end

#within_queue_limits? ⇒ Boolean

Returns true if the queue count is calculated to be still within limits when complete

Returns:

  • (Boolean) 


36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/crawl.rb', line 36

def within_queue_limits?

  # if we are limiting by page we can't limit the queue size as we don't know the mime type until retrieved
  if @options[:crawl_limit_by_page]
    return true

    # if a crawl limit is set, limit queue size to crawled + queue
  elsif @options[:crawl_limit].to_i > 0
    (queue_counter + crawl_counter) < @options[:crawl_limit].to_i

    # no crawl limit set so always within queue limit
  else
    true
  end
end