Class: Scrappy::Agent
- Inherits:
-
Object
- Object
- Scrappy::Agent
- Defined in:
- lib/scrappy/agent/agent.rb
Constant Summary collapse
- Options =
OpenStruct.new :format=>:yarf, :format_header=>true, :depth=>0, :delay=>0, :workers=>10
- ContentTypes =
{ :png => 'image/png', :rdfxml => 'application/rdf+xml', :rdf => 'application/rdf+xml' }
Instance Attribute Summary collapse
-
#id ⇒ Object
Returns the value of attribute id.
-
#kb ⇒ Object
Returns the value of attribute kb.
-
#options ⇒ Object
Returns the value of attribute options.
Class Method Summary collapse
Instance Method Summary collapse
-
#initialize(args = {}) ⇒ Agent
constructor
A new instance of Agent.
- #map(args, queue = nil) ⇒ Object
-
#observe(uris) ⇒ Object
Method to observe several webs, and extract the data periodically.
- #proxy(args = {}) ⇒ Object
- #reduce(results) ⇒ Object
- #request(args = {}) ⇒ Object
Methods included from BlindAgent
#html, #html_data?, #uri, #uri=
Methods included from Cached
Methods included from MapReduce
Methods included from Optimizer
#optimize_extractors, #optimize_patterns
Methods included from Trainer
Methods included from Extractor
#extract, #extract_graph, #fragments_for
Constructor Details
#initialize(args = {}) ⇒ Agent
Returns a new instance of Agent.
24 25 26 27 28 29 30 31 32 33 34 35 |
# File 'lib/scrappy/agent/agent.rb', line 24 def initialize args={} super() @cluster_count = args[:workers] || Options.workers @cluster_options = [ { :referenceable=>Options.referenceable, :agent=>Options.agent, :workers=>1, :window=>false } ] @cluster = args[:parent] @id = args[:id] || Agent.pool.keys.size Agent.pool[@id] = self @kb = args[:kb] || Options.kb @options = Options.clone @repository = args[:repository] || Options.repository end |
Instance Attribute Details
#id ⇒ Object
Returns the value of attribute id.
22 23 24 |
# File 'lib/scrappy/agent/agent.rb', line 22 def id @id end |
#kb ⇒ Object
Returns the value of attribute kb.
22 23 24 |
# File 'lib/scrappy/agent/agent.rb', line 22 def kb @kb end |
#options ⇒ Object
Returns the value of attribute options.
22 23 24 |
# File 'lib/scrappy/agent/agent.rb', line 22 def @options end |
Class Method Details
.[](id) ⇒ Object
18 19 20 |
# File 'lib/scrappy/agent/agent.rb', line 18 def self.[] id pool[id] || Agent.new(:id=>id) end |
.pool ⇒ Object
15 16 17 |
# File 'lib/scrappy/agent/agent.rb', line 15 def self.pool @pool ||= {} end |
Instance Method Details
#map(args, queue = nil) ⇒ Object
37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 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 83 84 85 86 87 88 89 90 91 92 93 |
# File 'lib/scrappy/agent/agent.rb', line 37 def map args, queue=nil depth = args[:depth] || .depth request = { :method=>args[:method]||:get, :uri=>complete_uri(args[:uri]), :inputs=>args[:inputs]||{} } # Expire cache cache.expire! 300 # 5 minutes # Lookup in cache triples = if cache[request] puts "Retrieving cached #{request[:uri]}...done!" if .debug cache[request][:response] elsif @repository # Extracts from the repository request_from_repository(request) else # Perform the request request_uncached(request) end # If previous cache exists, do not cache it again unless cache[request] # Cache the request cache[request] = { :time=>Time.now, :response=>triples } cache[request.merge(:uri=>self.uri)] = { :time=>Time.now, :response=>triples } if self.uri end # Enqueue subresources # Pages are enqueued without reducing depth pages = triples.select { |s,p,o| p==ID("rdf:type") and o==ID("sc:Page") }.map{|s,p,o| s}.select{|n| n.is_a?(Symbol)} # All other URIS are enqueued with depth reduced uris = if depth != 0 (triples.map { |s, p, o| [s,o] }.flatten - [ID(self.uri)] - pages).select{|n| n.is_a?(Symbol)} else [] end # Recently created URIs are not followed nofollow = triples.select { |s,p,o| p==ID("rdf:type") and o==ID("sc:NewUri") }.map{|s,p,o| s}.select{|n| n.is_a?(Symbol)} pages -= nofollow uris -= nofollow items = ( pages.map { |uri| {:uri=>uri.to_s, :depth=>[-1, depth].max} } + uris.map { |uri| {:uri=>uri.to_s, :depth=>[-1, depth-1].max} } ). uniq.select { |item| !RDF::ID.bnode?(item[:uri]) } items.each { |item| puts "Enqueuing (depth = #{item[:depth]}): #{item[:uri]}" if !queue or !(queue.history + queue.items).include?(item) } if .debug if queue items.each { |item| queue.push_unless_done item } else triples += process items end triples unless .dump end |
#observe(uris) ⇒ Object
Method to observe several webs, and extract the data periodically
140 141 142 143 144 145 146 147 148 149 150 151 |
# File 'lib/scrappy/agent/agent.rb', line 140 def observe uris while true time_init = Time.now.to_i uris.each do |uri| puts "Pinging #{uri}..." request :uri=>uri end time = .repository.time * 60 - (Time.now.to_i - time_init) puts "Sleeping until #{Time.now + time}..." sleep time end end |
#proxy(args = {}) ⇒ Object
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 |
# File 'lib/scrappy/agent/agent.rb', line 114 def proxy args={} request = { :method=>:get, :inputs=>{}, :format=>.format, :depth=>.depth }.merge(args) response = self.request(request) output = if .dump "" else if .debug print "Serializing..."; $stdout.flush end output = response.serialize request[:format], .format_header puts 'done!'if .debug output end OpenStruct.new :output => output, :content_type => ContentTypes[request[:format]] || 'text/plain', :uri => self.uri, :status => self.html_data? ? (self.uri == request[:uri] ? :ok : :redirect) : :error end |
#reduce(results) ⇒ Object
95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
# File 'lib/scrappy/agent/agent.rb', line 95 def reduce results return [] if .dump if .debug print "Merging results..."; $stdout.flush end triples = []; results.each { |result| triples += result } triples.uniq! puts 'done!'if .debug triples end |
#request(args = {}) ⇒ Object
110 111 112 |
# File 'lib/scrappy/agent/agent.rb', line 110 def request args={} RDF::Graph.new clean(map(args) || []) end |