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 => RedisConnection.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_handled?(link) ⇒ Boolean

Returns:

  • (Boolean) 


29
30
31
 
# File 'lib/crawl.rb', line 29

def already_handled?(link)
  already_crawled?(link) || already_queued?(link) || already_running?(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

#already_running?(link) ⇒ Boolean

Returns:

  • (Boolean) 


25
26
27
 
# File 'lib/crawl.rb', line 25

def already_running?(link)
  @redis.sismember "currently_running", link
end

#cancelled? ⇒ Boolean

Returns:

  • (Boolean) 


33
34
35
 
# File 'lib/crawl.rb', line 33

def cancelled?
  @stats.get_statistics[:current_status] == "Cancelled"
end

#content ⇒ Object 



157
158
159
160
 
# File 'lib/crawl.rb', line 157

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

#crawled_base_url ⇒ Object 



260
261
262
 
# File 'lib/crawl.rb', line 260

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

#debug_ap(value) ⇒ Object 



293
294
295
 
# File 'lib/crawl.rb', line 293

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

#debug_puts(value) ⇒ Object 



297
298
299
 
# File 'lib/crawl.rb', line 297

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

#finish ⇒ Object 



229
230
231
232
233
234
235
236
237
238
239
 
# File 'lib/crawl.rb', line 229

def finish
  debug_puts ""
  debug_puts "========================================================================"
  debug_puts "finished crawl on #{@options[:url]}"
  print_counters
  debug_puts "========================================================================"
  debug_puts ""

  set_first_to_finish
  @stats.end_crawl(@options)
end

#finished? ⇒ Boolean

Returns:

  • (Boolean) 


208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
 
# File 'lib/crawl.rb', line 208

def finished?
  print_counters
  debug_puts @stats.get_status
  if @stats.get_status == CobwebCrawlHelper::FINISHED
    debug_puts "Already Finished!"
  end  
  # 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?
      debug_puts "queue_counter is 0 and currently_running is empty so we're done"
      #finished
      return true
    end
  elsif (queue_counter == 0 || process_counter >= @options[:crawl_limit].to_i) && @redis.smembers("currently_running").empty?
    #finished
    debug_puts "queue_counter: #{queue_counter}, @redis.smembers(\"currently_running\").empty?: #{@redis.smembers("currently_running").empty?}, process_counter: #{process_counter}, @options[:crawl_limit].to_i: #{@options[:crawl_limit].to_i}"
    return true
  end
  false
end

#finished_processing ⇒ Object 



204
205
206
 
# File 'lib/crawl.rb', line 204

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

#first_to_finish? ⇒ Boolean

Returns:

  • (Boolean) 


256
257
258
 
# File 'lib/crawl.rb', line 256

def first_to_finish?
  @first_to_finish
end

#lock(key, &block) ⇒ Object 



272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
 
# File 'lib/crawl.rb', line 272

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", 30)
  begin
    result = yield
  ensure
    @redis.del("#{key}_lock")
    #debug_puts "LOCK RELEASED [#{key}]"
  end
  result
end

#process(&block) ⇒ Object 



188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
 
# File 'lib/crawl.rb', line 188

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

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

#process_links(&block) ⇒ Object 



110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
 
# File 'lib/crawl.rb', line 110

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) }

    # if the site has the same content for http and https then normalize to http 
    if @options[:treat_https_as_http]
      internal_links.map!{|link| link.gsub(/^https/, "http")}
    end

    # reject the link if we've crawled it or queued it
    internal_links.reject! { |link| already_handled?(link)}

    lock("internal-links") do
      internal_links.each do |link|
        if within_queue_limits? && !already_handled?(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
    end

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

#redis ⇒ Object 



268
269
270
 
# File 'lib/crawl.rb', line 268

def redis
  @redis
end

#retrieve ⇒ Object 



64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
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
 
# File 'lib/crawl.rb', line 64

def retrieve
  unless cancelled?
    unless already_running? @options[:url]
      unless already_crawled? @options[:url]
        update_queues
        if within_crawl_limits?
          @redis.sadd("currently_running", @options[:url])
          @stats.update_status("Retrieving #{@options[:url]}...")
          @content = Cobweb.new(@options).get(@options[:url], @options)
          update_counters

          if @options[:url] == @redis.get("original_base_url")
            @redis.set("crawled_base_url", @content[:base_url])
          end

          if content.permitted_type?
            ## update statistics

            @stats.update_statistics(@content)
            return true
          end
        else
          puts "======================================="
          puts "OUTWITH CRAWL LIMITS"
          puts "======================================="
          decrement_queue_counter
        end
      else
        puts "======================================="
        puts "ALREADY CRAWLED"
        puts "======================================="
        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
  else
    puts "======================================="
    puts "CRAWL CANCELLED"
    puts "======================================="
  end
  false
end

#set_first_to_finish ⇒ Object 



241
242
243
244
245
246
247
248
249
250
251
252
253
 
# File 'lib/crawl.rb', line 241

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 



264
265
266
 
# File 'lib/crawl.rb', line 264

def statistics
  @stats.get_statistics
end

#to_be_processed? ⇒ Boolean

Returns:

  • (Boolean) 


184
185
186
 
# File 'lib/crawl.rb', line 184

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

#update_counters ⇒ Object 



173
174
175
176
177
178
179
180
181
182
 
# File 'lib/crawl.rb', line 173

def update_counters
  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

#update_queues ⇒ Object 



162
163
164
165
166
167
168
169
170
171
 
# File 'lib/crawl.rb', line 162

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]

    # increment the counter if we are not limiting by page only || we are limiting count by page and it is a page
  end
end

#within_crawl_limits? ⇒ Boolean

Returns true if the crawl count is within limits

Returns:

  • (Boolean) 


38
39
40
 
# File 'lib/crawl.rb', line 38

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) 


43
44
45
 
# File 'lib/crawl.rb', line 43

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) 


48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
 
# File 'lib/crawl.rb', line 48

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