Class: Robots::ParsedRobots

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

Instance Method Summary collapse

Constructor Details

#initialize(uri, user_agent) ⇒ ParsedRobots

Returns a new instance of ParsedRobots.



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
 
# File 'lib/robots.rb', line 37

def initialize(uri, user_agent)
  @last_accessed = Time.at(1)
  
  io = Robots.get_robots_txt(uri, user_agent)
  
  if !io || io.content_type != "text/plain" || io.status != ["200", "OK"]
    io = StringIO.new("User-agent: *\nAllow: /\n")
  end

  @other = {}
  @disallows = {}
  @allows = {}
  @delays = {} # added delays to make it work
  agent = /.*/
  io.each do |line|
    next if line =~ /^\s*(#.*|$)/
    arr = line.split(":")
    key = arr.shift
    value = arr.join(":").strip
    value.strip!
    case key
    when "User-agent"
      agent = to_regex(value)
    when "Allow"
      @allows[agent] ||= []
      @allows[agent] << to_regex(value)
    when "Disallow"
      @disallows[agent] ||= []
      @disallows[agent] << to_regex(value)
    when "Crawl-delay"
      @delays[agent] = value.to_i
    else
      @other[key] ||= []
      @other[key] << value
    end
  end
  
  @parsed = true
end

Instance Method Details

#allowed?(uri, user_agent) ⇒ Boolean

Returns:

  • (Boolean) 


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
109
110
 
# File 'lib/robots.rb', line 77

def allowed?(uri, user_agent)
  return true unless @parsed
  allowed = true
  path = uri.request_uri
  
  @disallows.each do |key, value|
    if user_agent =~ key
      value.each do |rule|
        if path =~ rule
          allowed = false
        end
      end
    end
  end
  
  @allows.each do |key, value|
    unless allowed      
      if user_agent =~ key
        value.each do |rule|
          if path =~ rule
            allowed = true
          end
        end
      end
    end
  end
  
  if allowed && @delays[user_agent]
    sleep @delays[user_agent] - (Time.now - @last_accessed)
    @last_accessed = Time.now
  end
  
  return allowed
end

#other_values ⇒ Object 



112
113
114
 
# File 'lib/robots.rb', line 112

def other_values
  @other
end