inky · July 23, 2009 11:59 · Jul 22, 2009
diff --git a/pagerank.rb b/pagerank.rb
@@ -0,0 +1,96 @@
+#!/usr/bin/env ruby
+# quick Google PageRank lookup
+# usage: ~/bin/pagerank jamiedubs.com
+# => 6
+#
+# by Vsevolod S. Balashov <vsevolod @ balashovREMOVETHIS.name>
+# hacks/binify by Jamie Dubs <http://jamiedubs.com>
+# based on 3rd party code snippets (see comments)
+# TODO: make into a gem... the existing googlePR one is broken!
+
+require 'uri'
+require 'open-uri'
+
+# http://blog.outer-court.com/archive/2004_06_27_index.html#108834386239051706
+module SEO
+  class GooglePR
+
+    def initialize(uri)
+      @uri = uri
+    end
+
+    M=0x100000000 # modulo for unsigned int 32bit(4byte)
+
+    def m1(a,b,c,d)
+      (((a+(M-b)+(M-c))%M)^(d%M))%M # mix/power mod
+    end
+
+    def i2c(i)
+      [i&0xff, i>>8&0xff, i>>16&0xff, i>>24&0xff]
+    end
+
+    def c2i(s,k=0)
+      ((s[k+3].to_i*0x100+s[k+2].to_i)*0x100+s[k+1].to_i)*0x100+s[k].to_i
+    end
+
+    def mix(a,b,c)
+      a = a%M; b = b%M; c = c%M
+      a = m1(a,b,c, c >> 13); b = m1(b,c,a, a <<  8); c = m1(c,a,b, b >> 13)
+      a = m1(a,b,c, c >> 12); b = m1(b,c,a, a << 16); c = m1(c,a,b, b >>  5)
+      a = m1(a,b,c, c >>  3); b = m1(b,c,a, a << 10); c = m1(c,a,b, b >> 15)
+      [a, b, c]
+    end
+
+    def old_cn(iurl = 'info:' + @uri)
+      a = 0x9E3779B9; b = 0x9E3779B9; c = 0xE6359A60
+      len = iurl.size 
+      k = 0
+      while (len >= k + 12) do
+        a += c2i(iurl,k); b += c2i(iurl,k+4); c += c2i(iurl,k+8)
+        a, b, c = mix(a, b, c)
+        k = k + 12
+      end
+      a += c2i(iurl,k); b += c2i(iurl,k+4); c += (c2i(iurl,k+8) << 8) + len
+      a,b,c = mix(a,b,c)
+      return c
+    end
+
+    def cn
+      ch = old_cn
+      ch = ((ch/7) << 2) | ((ch-(ch/13).floor*13)&7)
+      new_url = []
+      20.times { i2c(ch).each { |i| new_url << i }; ch -= 9 }
+      ('6' + old_cn(new_url).to_s).to_i
+    end
+
+    def request_uri
+      # http://www.bigbold.com/snippets/posts/show/1260 + _ -> %5F
+      "http://toolbarqueries.google.com/search?client=navclient-auto&hl=en&ch=#{cn}&ie=UTF-8&oe=UTF-8&features=Rank&q=info:#{URI.escape(@uri, /[^-.!~*'()a-zA-Z\d]/)}"
+    end
+
+    def page_rank(uri = @uri)
+      @uri = uri if uri != @uri
+      begin
+        open(request_uri) { |f| return $1.to_i if f.string =~ /Rank_1:\d:(\d+)/ }
+      rescue OpenURI::HTTPError => e
+        puts "(!!) Error fetching URL: #{e} -- #{e.backtrace.join("\n")}"
+        sleep 2
+      end
+      nil
+    end
+
+    # calculate the checksum for the given URL    
+    private :m1, :i2c, :c2i, :mix, :old_cn
+    attr_accessor :uri
+
+  end
+end
+
+
+# run
+if __FILE__ == $0
+  url = ARGV.first
+  c = SEO::GooglePR.new(url)
+  puts c.page_rank
+end
+