Class: LinkedinData

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

Instance Method Summary collapse

Constructor Details

#initialize(input, todegree) ⇒ LinkedinData

Returns a new instance of LinkedinData.



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

def initialize(input, todegree)
  @input = input
  @output = Array.new
  @startindex = 10
  @numhops = todegree
end

Instance Method Details

#examine(page) ⇒ Object

Examines a search page



30
31
32
33
34
35
36
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
 
# File 'lib/linkedindata.rb', line 30

def examine(page)
  # Separate getting profile links and going to next page
    # Method for getting links to all result pages
    # Different method for getting all profile links on page and scraping (split to new thread for this)
       # Has own output set, merge into full one at end (make sure threadsafe)
    
  # Have own input and output
  page.links.each do |link|
    if (link.href.include? "linkedin.com") && (!link.href.include? "webcache") && (!link.href.include? "site:linkedin.com/pub+")
      saveurl = link.href.split("?q=")
      
      if saveurl[1]
        url = saveurl[1].split("&")
        begin
          scrape(url[0], 0)
        rescue
        end
      end
    end

    # Find the link to the next page and go to it
    if (link.href.include? "&sa=N") && (link.href.include? "&start=")
      url1 = link.href.split("&start=")
      url2 = url1[1].split("&sa=N")

      if url2[0].to_i == @startindex
        sleep(rand(30..90))
        @startindex += 10
        agent = Mechanize.new
        examine(agent.get("http://google.com" + link.href))
      end
    end
  end
end

#getData ⇒ Object

Gets all data and returns in JSON



135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
 
# File 'lib/linkedindata.rb', line 135

def getData
  search

  # Get related profiles
  @numhops.times do
    @output.each do |o|
      if o[:degree] < @numhops

        if o[:related_people]
          o[:related_people].each do |i|
            if @output.select { |obj| obj[:name] == i[:name]}.empty?
              scrape(i[:url], o[:degree]+1)
            end
          end
        end

      end
    end
  end

  formatted_json = JSON.pretty_generate(relScore(showAllKeys(@output)))
  return formatted_json
end

#relScore(data) ⇒ Object

Add a score to each profile based on the # of times it appears in “people also viewed”



105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
 
# File 'lib/linkedindata.rb', line 105

def relScore(data)

  # Make list of profiles
  profiles = Hash.new
  data.each do |d|
    profiles[d["profile_url"]] = 0
  end

  # Get degree for each profile
  data.each do |i|
    if i["related_people"]
      i["related_people"].each do |p|
        if profiles[p["url"]]
          # Calculate degree- (2/d*2) except when degree is 0
          degree_divide = i["degree"] == 0 ? 1 : i["degree"]*2
          profiles[p["url"]] += (2.0/degree_divide)
        end
      end
    end
  end

  # Merge scores back into dataset
  data.each do |m|
    m.merge!(:score => profiles[m["profile_url"]])
  end

  return data
end

#scrape(url, curhops) ⇒ Object

Scrapes profile



66
67
68
69
70
71
72
73
74
75
76
77
78
79
 
# File 'lib/linkedindata.rb', line 66

def scrape(url, curhops)
  # Download profile and rescue on error
  begin
    url.gsub!("https", "http")
    profile = Linkedin::Profile.get_profile(url)
  rescue
  end
  
  # Parse profile if returned
  if profile
    p = ParseProfile.new(profile, url, curhops)
    @output.concat(p.parse)
  end
end

#search ⇒ Object

Searches for profiles on Google



20
21
22
23
24
25
26
27
 
# File 'lib/linkedindata.rb', line 20

def search
  agent = Mechanize.new
  agent.user_agent_alias = 'Linux Firefox'
  gform = agent.get("http://google.com").form("f")
  gform.q = "site:linkedin.com/pub " + @input
  page = agent.submit(gform, gform.buttons.first)
  examine(page)
end

#showAllKeys(data) ⇒ Object

Make sure all keys that occur occur in each item (even if nil)



82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
 
# File 'lib/linkedindata.rb', line 82

def showAllKeys(data)
  # Get all keys
  fields = Set.new
  data.map { |o| fields.merge(o.keys) }

  # Make sure all items have all keys
  datarr = Array.new
  data.each do |d|
    temphash = Hash.new
    fields.each do |f|
      if !d[f]
        temphash[f] = nil
      else
        temphash[f] = d[f]
      end
    end
    datarr.push(temphash)
  end

  return datarr
end