linkedindata 0.0.1 → 0.0.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/lib/linkedindata.rb +9 -6
- metadata +2 -2
data/lib/linkedindata.rb
CHANGED
|
@@ -6,6 +6,7 @@ class LinkedinData
|
|
|
6
6
|
def initialize(input)
|
|
7
7
|
@input = input
|
|
8
8
|
@output = Array.new
|
|
9
|
+
@startindex = 10
|
|
9
10
|
end
|
|
10
11
|
|
|
11
12
|
# Searches for links on Google
|
|
@@ -16,11 +17,9 @@ class LinkedinData
|
|
|
16
17
|
page = agent.submit(gform, gform.buttons.first)
|
|
17
18
|
examine(page)
|
|
18
19
|
end
|
|
19
|
-
|
|
20
|
+
|
|
20
21
|
# Examines a search page
|
|
21
22
|
def examine(page)
|
|
22
|
-
startindex = 0
|
|
23
|
-
|
|
24
23
|
page.links.each do |link|
|
|
25
24
|
if (link.href.include? "linkedin.com") && (!link.href.include? "webcache") && (!link.href.include? "site:linkedin.com/pub+")
|
|
26
25
|
saveurl = link.href.split("?q=")
|
|
@@ -35,10 +34,11 @@ class LinkedinData
|
|
|
35
34
|
url1 = link.href.split("&start=")
|
|
36
35
|
url2 = url1[1].split("&sa=N")
|
|
37
36
|
|
|
38
|
-
if url2[0].to_i
|
|
37
|
+
if url2[0].to_i == @startindex
|
|
38
|
+
sleep(20)
|
|
39
|
+
@startindex += 10
|
|
39
40
|
agent = Mechanize.new
|
|
40
41
|
examine(agent.get("http://google.com" + link.href))
|
|
41
|
-
else startindex = url2[0].to_i
|
|
42
42
|
end
|
|
43
43
|
end
|
|
44
44
|
end
|
|
@@ -64,6 +64,9 @@ class LinkedinData
|
|
|
64
64
|
# Gets all data and returns in JSON
|
|
65
65
|
def getData
|
|
66
66
|
search
|
|
67
|
-
return @output
|
|
67
|
+
return JSON.pretty_generate(@output)
|
|
68
68
|
end
|
|
69
69
|
end
|
|
70
|
+
|
|
71
|
+
l = LinkedinData.new("National Security Agency")
|
|
72
|
+
puts l.getData
|
metadata
CHANGED
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: linkedindata
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.0.
|
|
4
|
+
version: 0.0.2
|
|
5
5
|
prerelease:
|
|
6
6
|
platform: ruby
|
|
7
7
|
authors:
|
|
@@ -9,7 +9,7 @@ authors:
|
|
|
9
9
|
autorequire:
|
|
10
10
|
bindir: bin
|
|
11
11
|
cert_chain: []
|
|
12
|
-
date: 2014-
|
|
12
|
+
date: 2014-02-14 00:00:00.000000000 Z
|
|
13
13
|
dependencies: []
|
|
14
14
|
description: Scrapes all LinkedIn profiles including terms you specify.
|
|
15
15
|
email: shidash@shidash.com
|