Class: Probot

Inherits:
Object
  • Object
show all
Defined in:
lib/probot.rb,
lib/probot/version.rb

Overview

Two main parts of this class:

Parse a robots.txt file
Find the most specific rule for a given URL. We use the length of the regexp as a proxy for specificity.

Defined Under Namespace

Classes: ParsedLine

Constant Summary collapse

VERSION = 
"0.2.0"

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(data, agent: "*") ⇒ Probot

Returns a new instance of Probot.

Raises:

  • (ArgumentError) 


25
26
27
28
29
30
31
32
33
34
35
36
 
# File 'lib/probot.rb', line 25

def initialize(data, agent: "*")
  raise ArgumentError, "The first argument must be a string" unless data.is_a?(String)
  @agent = agent

  @rules = {}
  @current_agents = ["*"]
  @current_agents.each { |agent| @rules[agent] ||= {"disallow" => [], "allow" => [], "crawl_delay" => 0} }
  @sitemaps = []
  @site = URI(data) if data.start_with?("http")
  @doc = @site.nil? ? data : fetch_robots_txt(@site)
  parse(@doc)
end

Instance Attribute Details

#agent ⇒ Object

Returns the value of attribute agent.



23
24
25
 
# File 'lib/probot.rb', line 23

def agent
  @agent
end

#doc ⇒ Object (readonly)

Returns the value of attribute doc.



22
23
24
 
# File 'lib/probot.rb', line 22

def doc
  @doc
end

#rules ⇒ Object (readonly)

Returns the value of attribute rules.



22
23
24
 
# File 'lib/probot.rb', line 22

def rules
  @rules
end

#site ⇒ Object

Returns the value of attribute site.



23
24
25
 
# File 'lib/probot.rb', line 23

def site
  @site
end

#sitemaps ⇒ Object

Returns the value of attribute sitemaps.



23
24
25
 
# File 'lib/probot.rb', line 23

def sitemaps
  @sitemaps
end

Class Method Details

.allowed?(url, agent: "*") ⇒ Boolean

Returns:

  • (Boolean) 


149
 
# File 'lib/probot.rb', line 149

def self.allowed?(url, agent: "*") = Probot.new(url, agent: agent).allowed?(url)

Instance Method Details

#allowed ⇒ Object 



52
 
# File 'lib/probot.rb', line 52

def allowed = rules.dig(@agent, "allow") || rules.dig("*", "allow")

#allowed?(url) ⇒ Boolean

If a URL is not disallowed, it is allowed - so we check if it is explictly disallowed and if not, it’s allowed.

Returns:

  • (Boolean) 


67
 
# File 'lib/probot.rb', line 67

def allowed?(url) = !disallowed?(url)

#allowed_best(url) ⇒ Object 



62
 
# File 'lib/probot.rb', line 62

def allowed_best(url) = allowed_matches(url).max_by { |k, v| v }

#allowed_matches(url) ⇒ Object 



56
 
# File 'lib/probot.rb', line 56

def allowed_matches(url) = allowed.select { |allowed_url| url.match?(allowed_url) }.to_h { |rule| [rule, pattern_length(rule)] }

#crawl_delay ⇒ Object 



46
 
# File 'lib/probot.rb', line 46

def crawl_delay = rules.dig(@agent, "crawl_delay")

#disallowed ⇒ Object 



50
 
# File 'lib/probot.rb', line 50

def disallowed = rules.dig(@agent, "disallow") || rules.dig("*", "disallow")

#disallowed?(url) ⇒ Boolean

Returns:

  • (Boolean) 


69
 
# File 'lib/probot.rb', line 69

def disallowed?(url) = matching_rule(url)&.keys&.first == :disallow

#disallowed_best(url) ⇒ Object 



60
 
# File 'lib/probot.rb', line 60

def disallowed_best(url) = disallowed_matches(url).max_by { |k, v| v }

#disallowed_matches(url) ⇒ Object 



54
 
# File 'lib/probot.rb', line 54

def disallowed_matches(url) = disallowed.select { |disallowed_url| url.match?(disallowed_url) }.to_h { |rule| [rule, pattern_length(rule)] }

#fetch_robots_txt(url) ⇒ Object 



40
41
42
43
44
 
# File 'lib/probot.rb', line 40

def fetch_robots_txt(url)
  Net::HTTP.get(URI(url).tap { |u| u.path = "/robots.txt" }, request_headers)
rescue
  ""
end

#found_agents ⇒ Object 



48
 
# File 'lib/probot.rb', line 48

def found_agents = rules.keys

#matches(url) ⇒ Object 



58
 
# File 'lib/probot.rb', line 58

def matches(url) = {disallowed: disallowed_matches(url), allowed: allowed_matches(url)}

#matching_rule(url) ⇒ Object 



64
 
# File 'lib/probot.rb', line 64

def matching_rule(url) = (disallowed_best(url)&.last.to_i > allowed_best(url)&.last.to_i) ? {disallow: disallowed_best(url)&.first} : {allow: allowed_best(url)&.first}

#parse(doc) ⇒ Object 



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
109
110
111
112
113
114
115
116
 
# File 'lib/probot.rb', line 71

def parse(doc)
  # We need to handle consective user-agent lines, which are considered to be part of the same record.
  subsequent_agent = false

  doc.lines.each do |line|
    next if line.start_with?("#") || !line.include?(":") || line.split(":").length < 2

    data = ParsedLine.new(line)

    if data.agent?
      if subsequent_agent
        @current_agents << data.value
      else
        @current_agents = [data.value]
        subsequent_agent = true
      end

      @current_agents.each { |agent| rules[agent] ||= {"disallow" => [], "allow" => [], "crawl_delay" => 0} }
      next
    end

    # All Regex characters are escaped, then we unescape * and $ as they may used in robots.txt
    if data.allow? || data.disallow?
      @current_agents.each { |agent| rules[agent][data.key] << Regexp.new(Regexp.escape(data.value).gsub('\*', ".*").gsub('\$', "$")) }

      # When user-agent strings are found on consecutive lines, they are considered to be part of the same record. Google ignores crawl_delay.
      subsequent_agent = false
      next
    end

    if data.crawl_delay?
      @current_agents.each { |agent| rules[agent][data.key] = data.value }
      next
    end

    # Ensure we have an absolute URL
    if data.sitemap?
      sitemap_uri = URI(data.value)
      sitemap_uri = sitemap_uri.host.nil? ? URI.join(*[site, sitemap_uri].compact) : sitemap_uri
      @sitemaps << sitemap_uri.to_s
      next
    end

    @current_agents.each { |agent| rules[agent][data.key] = data.value }
  end
end

#pattern_length(regexp) ⇒ Object 



118
 
# File 'lib/probot.rb', line 118

def pattern_length(regexp) = regexp.source.gsub(/(\\[\*\$\.])/, "*").length

#request_headers ⇒ Object 



38
 
# File 'lib/probot.rb', line 38

def request_headers = (agent == "*") ? {} : {"User-Agent" => @agent}