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.



4
5
6
7
8
9
 
# File 'lib/rdig/crawler.rb', line 4

def initialize(config = RDig.config, logger = RDig.logger)
  @documents = Queue.new
  @logger = logger
  @config = config
  @indexed_documents = 0
end

Instance Method Details

#add_to_index(doc) ⇒ Object 



78
79
80
81
82
 
# File 'lib/rdig/crawler.rb', line 78

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



88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
 
# File 'lib/rdig/crawler.rb', line 88

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

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

#crawl ⇒ Object 



18
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
51
52
53
 
# File 'lib/rdig/crawler.rb', line 18

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]

  # the etag filter operates on the fetched document, thats why we cannot put it into the filter chain right now.
  @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)
      @logger.info "thread #{i} running..."
      while (doc = @documents.pop) != :exit
        process_document doc, filterchain
      end
      @logger.info "thread #{i} is done."
    }
  }

  # 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
  @logger.info "indexed #{@indexer.indexed_documents} documents"
end

#process_document(doc, filterchain) ⇒ Object 



55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
 
# File 'lib/rdig/crawler.rb', line 55

def process_document(doc, filterchain)
  @logger.info "processing document #{doc.uri}"
  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}: #{$!}"
  @logger.info "Trace: #{$!.backtrace.join("\n")}"
end

#run ⇒ Object 



11
12
13
14
15
16
 
# File 'lib/rdig/crawler.rb', line 11

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