trackablaze 0.1.2 → 0.1.3
Sign up to get free protection for your applications and to get access to all the features.
- data/trackers/pagerank.rb +144 -0
- data/trackers/pagerank.yml +12 -0
- data/version.rb +1 -1
- metadata +3 -1
@@ -0,0 +1,144 @@
|
|
1
|
+
require 'socket'
|
2
|
+
|
3
|
+
module Trackablaze
|
4
|
+
class Pagerank < Tracker
|
5
|
+
def get_metrics(tracker_items)
|
6
|
+
tracker_items.collect {|tracker_item| get_metrics_single(tracker_item)}
|
7
|
+
end
|
8
|
+
|
9
|
+
def get_metrics_single(tracker_item)
|
10
|
+
metrics = {}
|
11
|
+
|
12
|
+
if (tracker_item.params["domain"].nil? || tracker_item.params["domain"].empty?)
|
13
|
+
add_error(metrics, "No domain supplied", "domain")
|
14
|
+
return metrics
|
15
|
+
end
|
16
|
+
|
17
|
+
pagerank = nil
|
18
|
+
begin
|
19
|
+
pagerank = getpr(tracker_item.params["domain"])
|
20
|
+
rescue
|
21
|
+
end
|
22
|
+
|
23
|
+
if (pagerank.nil?)
|
24
|
+
add_error(metrics, "Could not find pagerank. Is domain specified properly (e.g. google.com)?", "domain")
|
25
|
+
return metrics
|
26
|
+
end
|
27
|
+
|
28
|
+
metrics['pagerank'] = pagerank
|
29
|
+
|
30
|
+
metrics
|
31
|
+
end
|
32
|
+
|
33
|
+
#------------------------------
|
34
|
+
# Ruby conversion by Amol Kelkar of
|
35
|
+
# PageRank Lookup v1.1 by HM2K at http://www.hm2k.com/projects/pagerank
|
36
|
+
# based on an alogoritham found here: http://pagerank.gamesaga.net/
|
37
|
+
#------------------------------
|
38
|
+
|
39
|
+
#convert a string to a 32-bit integer
|
40
|
+
def str_to_num(str, check, magic)
|
41
|
+
int32Unit = 4294967296; # 2^32
|
42
|
+
|
43
|
+
length = str.length;
|
44
|
+
(0..length-1).each do |i|
|
45
|
+
check *= magic;
|
46
|
+
|
47
|
+
#If the float is beyond the boundaries of integer (usually +/- 2.15e+9 = 2^31),
|
48
|
+
# the result of converting to integer is undefined
|
49
|
+
# refer to http://www.php.net/manual/en/language.types.integer.php
|
50
|
+
if (check >= int32Unit)
|
51
|
+
check = (check - int32Unit * (check / int32Unit).to_i);
|
52
|
+
#if the check less than -2^31
|
53
|
+
check = (check < -2147483648) ? (check + int32Unit) : check;
|
54
|
+
end
|
55
|
+
check += str[i].ord
|
56
|
+
end
|
57
|
+
return check
|
58
|
+
end
|
59
|
+
|
60
|
+
# genearate a hash for a url
|
61
|
+
def hash_url(str)
|
62
|
+
check1 = str_to_num(str, 0x1505, 0x21);
|
63
|
+
check2 = str_to_num(str, 0, 0x1003F);
|
64
|
+
|
65
|
+
check1 >>= 2;
|
66
|
+
check1 = ((check1 >> 4) & 0x3FFFFC0 ) | (check1 & 0x3F);
|
67
|
+
check1 = ((check1 >> 4) & 0x3FFC00 ) | (check1 & 0x3FF);
|
68
|
+
check1 = ((check1 >> 4) & 0x3C000 ) | (check1 & 0x3FFF);
|
69
|
+
|
70
|
+
t1 = ((((check1 & 0x3C0) << 4) | (check1 & 0x3C)) <<2 ) | (check2 & 0xF0F );
|
71
|
+
t2 = ((((check1 & 0xFFFFC000) << 4) | (check1 & 0x3C00)) << 0xA) | (check2 & 0xF0F0000 );
|
72
|
+
|
73
|
+
return t1 | t2;
|
74
|
+
end
|
75
|
+
|
76
|
+
# genearate a checksum for the hash string
|
77
|
+
def check_hash(hashnum)
|
78
|
+
checkByte = 0
|
79
|
+
flag = 0
|
80
|
+
|
81
|
+
hashStr = sprintf('%u', hashnum)
|
82
|
+
length = hashStr.length;
|
83
|
+
|
84
|
+
(length - 1).downto(0) do |i|
|
85
|
+
re = hashStr[i].to_i;
|
86
|
+
if (1 === (flag % 2))
|
87
|
+
re += re
|
88
|
+
re = (re / 10).to_i + (re % 10).to_i
|
89
|
+
end
|
90
|
+
checkByte += re;
|
91
|
+
flag += 1
|
92
|
+
end
|
93
|
+
|
94
|
+
checkByte %= 10;
|
95
|
+
if (0 != checkByte)
|
96
|
+
checkByte = 10 - checkByte;
|
97
|
+
if (1 === (flag % 2) )
|
98
|
+
if (1 === (checkByte % 2))
|
99
|
+
checkByte += 9
|
100
|
+
end
|
101
|
+
checkByte >>= 1
|
102
|
+
end
|
103
|
+
end
|
104
|
+
|
105
|
+
return "7#{checkByte}#{hashStr}"
|
106
|
+
end
|
107
|
+
|
108
|
+
# return the pagerank checksum hash
|
109
|
+
def getch(url)
|
110
|
+
return check_hash(hash_url(url));
|
111
|
+
end
|
112
|
+
|
113
|
+
# return the pagerank figure
|
114
|
+
def getpr(url)
|
115
|
+
googlehost='toolbarqueries.google.com';
|
116
|
+
googleua='Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.0.6) Gecko/20060728 Firefox/1.5';
|
117
|
+
|
118
|
+
ch = getch(url)
|
119
|
+
fp = TCPSocket.open(googlehost, 80)
|
120
|
+
if (fp)
|
121
|
+
out = "GET /search?client=navclient-auto&ch=#{ch}&features=Rank&q=info:#{url} HTTP/1.1\r\n"
|
122
|
+
|
123
|
+
out = out + "User-Agent: #{googleua}\r\n"
|
124
|
+
out = out + "Host: #{googlehost}\r\n"
|
125
|
+
out = out + "Connection: Close\r\n\r\n"
|
126
|
+
|
127
|
+
fp.puts out
|
128
|
+
|
129
|
+
while line = fp.gets do
|
130
|
+
pos = line.index("Rank_");
|
131
|
+
unless(pos.nil?)
|
132
|
+
pr = line.slice(pos + 9, 2);
|
133
|
+
return pr.to_i;
|
134
|
+
end
|
135
|
+
end
|
136
|
+
fp.close
|
137
|
+
end
|
138
|
+
end
|
139
|
+
|
140
|
+
end
|
141
|
+
end
|
142
|
+
|
143
|
+
|
144
|
+
|
data/version.rb
CHANGED
metadata
CHANGED
@@ -2,7 +2,7 @@
|
|
2
2
|
name: trackablaze
|
3
3
|
version: !ruby/object:Gem::Version
|
4
4
|
prerelease:
|
5
|
-
version: 0.1.
|
5
|
+
version: 0.1.3
|
6
6
|
platform: ruby
|
7
7
|
authors:
|
8
8
|
- Amol Kelkar
|
@@ -118,6 +118,8 @@ files:
|
|
118
118
|
- lib/trackablaze.rb
|
119
119
|
- trackers/facebook_page.rb
|
120
120
|
- trackers/facebook_page.yml
|
121
|
+
- trackers/pagerank.rb
|
122
|
+
- trackers/pagerank.yml
|
121
123
|
- trackers/twitter.rb
|
122
124
|
- trackers/twitter.yml
|
123
125
|
- README.markdown
|