Class: RDig::Crawler

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

Instance Method Summary collapse

Constructor Details

#initialize(config = RDig.config, logger = RDig.logger) ⇒ Crawler

Returns a new instance of Crawler.



6
7
8
9
10
 
# File 'lib/rdig/crawler.rb', line 6

def initialize(config = RDig.config, logger = RDig.logger)
  @documents = Queue.new
  @logger = logger
  @config = config
end

Instance Method Details

#add_to_index(doc) ⇒ Object 



75
76
77
 
# File 'lib/rdig/crawler.rb', line 75

def add_to_index(doc)
  @indexer << doc if doc.needs_indexing?
end

#add_url(url, filterchain, referring_document = nil) ⇒ Object

pipes a new document pointing to url through the filter chain, if it survives that, it gets added to the documents queue for further processing



83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
 
# File 'lib/rdig/crawler.rb', line 83

def add_url(url, filterchain, referring_document = nil)
  return if url.nil? || url.empty?

  @logger.debug "add_url #{url}"
  doc = if referring_document
    referring_document.create_child(url)
  else
    Document.create(url)
  end

  doc = filterchain.apply(doc)
    
  if doc
    @documents << doc
    @logger.debug "url #{url} survived filterchain"
  end
rescue
  nil
end

#crawl ⇒ Object 



19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
 
# File 'lib/rdig/crawler.rb', line 19

def crawl
  raise 'no start urls given!' if @config.crawler.start_urls.empty?
  # check whether we are indexing on-disk or via http
  url_type = @config.crawler.start_urls.first =~ /^file:\/\// ? :file : :http
  chain_config = RDig.filter_chain[url_type]

  @etag_filter = ETagFilter.new
  filterchain = UrlFilters::FilterChain.new(chain_config)
  @config.crawler.start_urls.each { |url| add_url(url, filterchain) }

  num_threads = @config.crawler.num_threads
  group = ThreadsWait.new
  num_threads.times { |i|
    group.join_nowait Thread.new("fetcher #{i}") {
      filterchain = UrlFilters::FilterChain.new(chain_config)
      while (doc = @documents.pop) != :exit
        process_document doc, filterchain
      end
    }
  }

  # check for an empty queue every now and then 
  sleep_interval = @config.crawler.wait_before_leave
  begin 
    sleep sleep_interval
  end until @documents.empty?
  # nothing to do any more, tell the threads to exit
  num_threads.times { @documents << :exit }

  @logger.info "waiting for threads to finish..."
  group.all_waits
end

#process_document(doc, filterchain) ⇒ Object 



52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
 
# File 'lib/rdig/crawler.rb', line 52

def process_document(doc, filterchain)
  @logger.debug "processing document #{doc}"
  doc.fetch
  case doc.status
  when :success
    if @etag_filter.apply(doc)
      # add links from this document to the queue
      doc.content[:links].each { |url| 
        add_url(url, filterchain, doc) 
      } unless doc.content[:links].nil?
      add_to_index doc
    end        
  when :redirect
    @logger.debug "redirect to #{doc.content}"
    add_url(doc.content, filterchain, doc)
  else
    @logger.error "unknown doc status #{doc.status}: #{doc}"
  end
rescue
  @logger.error "error processing document #{doc.uri.to_s}: #{$!}"
  @logger.debug "Trace: #{$!.backtrace.join("\n")}"
end

#run ⇒ Object 



12
13
14
15
16
17
 
# File 'lib/rdig/crawler.rb', line 12

def run
  @indexer = Index::Indexer.new(@config.index)
  crawl
ensure
  @indexer.close if @indexer
end