Class: Glib::JsonCrawler::Router

Inherits:
Object
  • Object
show all
Defined in:
lib/glib/json_crawler/router.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ Router

Returns a new instance of Router.



15
16
17
18
19
20
21
22
 
# File 'lib/glib/json_crawler/router.rb', line 15

def initialize
  @depth = -1
  @logger = ''
  @visitor = Glib::Json::Traversal::Visitor.new
  @read_only_actions = Set.new
  # default rails's development host
  @host ||= 'localhost:3000'
end

Instance Attribute Details

#host ⇒ Object

Returns the value of attribute host.



5
6
7
 
# File 'lib/glib/json_crawler/router.rb', line 5

def host
  @host
end

#logger ⇒ Object (readonly)

Returns the value of attribute logger.



4
5
6
 
# File 'lib/glib/json_crawler/router.rb', line 4

def logger
  @logger
end

#read_only_actions ⇒ Object (readonly)

Returns the value of attribute read_only_actions.



4
5
6
 
# File 'lib/glib/json_crawler/router.rb', line 4

def read_only_actions
  @read_only_actions
end

Instance Method Details

#allowed?(url) ⇒ Boolean

Returns:

  • (Boolean) 


96
97
98
99
 
# File 'lib/glib/json_crawler/router.rb', line 96

def allowed?(url)
  regex = Regexp.new("#{host}.+(?<!\.pdf)$")
  regex.match(url)
end

#crawl_multiple(views, block) ⇒ Object 



92
93
94
 
# File 'lib/glib/json_crawler/router.rb', line 92

def crawl_multiple(views, block)
  @visitor.traverse_multiple views, block
end

#follow(http, target_router) ⇒ Object 



84
85
86
87
88
89
90
 
# File 'lib/glib/json_crawler/router.rb', line 84

def follow(http, target_router)
  @depth += 1
  target_router.read_only_actions.each do |crawler_action|
    action, url = crawler_action
    http.get(url, action, {}, false)
  end
end

#log(action, url, response = nil) ⇒ Object 



7
8
9
10
11
12
13
 
# File 'lib/glib/json_crawler/router.rb', line 7

def log(action, url, response = nil)
  @logger +='  ' * @depth + [
    action,
    response.present? ? response.code : nil,
    url
  ].compact.join(' :: ') + "\n"
end

#process_action(http, spec) ⇒ Object 



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
 
# File 'lib/glib/json_crawler/router.rb', line 44

def process_action(http, spec)
  action = spec['action']
  params = spec

  if action.present?
    @depth += 1
    case action
    when 'initiate_navigation'
      @read_only_actions.add([action, params['url']])
      JsonCrawler::NavInitiate.new(http, params, action)
    when 'runMultiple-v1', 'runMultiple'
      JsonCrawler::RunMultiple.new(http, params, action)
    when 'windows/open-v1', 'dialogs/open-v1', 'windows/reload-v1', 'windows/open', 'dialogs/open', 'windows/reload', 'windows/openWeb', 'windows/openWeb-v1'
      if allowed?(params['url'])
        @read_only_actions.add([action, params['url']])
        JsonCrawler::WindowsOpen.new(http, params, action)
      else
        self.log action, params['url']
      end
    when 'sheets/select-v1', 'sheets/select'
      JsonCrawler::Menu.new(http, params, action)
    when 'http/post-v1', 'http/post'
      JsonCrawler::ActionHttp.new(:post, http, params, action)
    when 'forms/submit-v1', 'forms/submit'
      forms = @visitor.forms
      # raise 'Submit action needs to be inside a form' if forms.size < 1
      JsonCrawler::FormsSubmit.new(http, forms.last)
    else
      unless [
        'http/delete-v1', 'dialogs/oauth-v1',
        'http/delete', 'dialogs/oauth'
      ].include?(action)
        @read_only_actions.add([action, params['url']])
      end
      self.log action, params['url']
    end
    @depth -= 1
  end
end

#step(http, args) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
 
# File 'lib/glib/json_crawler/router.rb', line 24

def step(http, args)
  # TODO: Refactor
  case args['view']
  when 'fields/submit-v1', 'fields/submit'
    @depth += 1
    forms = @visitor.forms
    JsonCrawler::FormsSubmit.new(http, forms.last)
    @depth -= 1
    return
  end

  if args.is_a?(Hash) && args['rel'] != 'nofollow'
    if (on_click = (args.fetch('onClick', nil)))
      process_action(http, on_click)
    end
  end

  @read_only_actions.replace(@read_only_actions.sort_by { |e| e[1].to_s })
end