geohashing

   1 #!/usr/bin/env perl
   2 #
   3 # Reference implementation for the geohashing algorithm described in xkcd #426
   4 # See http://wiki.xkcd.com/ProjectRainbow/Main_Page
   5 #
   6 # Dan Boger (zigdon@gmail.com)
   7 # 2008-05-21
   8
   9 use strict;
  10 use warnings;
  11 use Digest::MD5 qw/md5_hex/;
  12 use Date::Manip;
  13 use Getopt::Std;
  14
  15 my %opts;
  16 getopts('e', \%opts);
  17
  18 # get the date from the commandline, or assume it's today's date
  19 my $date = shift || "";
  20 $date ||= 'today';
  21 my $datestring = UnixDate( ParseDate($date), "%Y-%m-%d" );
  22
  23 die "Usage: $0 [-e] <date> [DJI opening]"
  24   unless $datestring and $datestring =~ /^\d\d\d\d-\d\d-\d\d$/;
  25
  26 # get the DJIA from the commandline, or try to retrieve it from google
  27 my $djia = shift;
  28 unless ($djia) {
  29     $djia = &download_djia($date, $opts{e});
  30 }
  31
  32 print "Date: $datestring, DJIA: $djia\n";
  33
  34 # calculate the MD5 of the combined date and DJIA
  35 my $md5 = md5_hex("$datestring-$djia");
  36 print "MD5($datestring-$djia): $md5\n";
  37
  38 # split into two
  39 my ( $md5x, $md5y ) = ( substr( $md5, 0, 16 ), substr( $md5, 16, 16 ) );
  40 print "Split: $md5x, $md5y\n";
  41
  42 # transform into a fraction
  43 my ( $fx, $fy ) = ( 0, 0 );
  44 while ( length $md5x or length $md5y ) {
  45     my $d = substr( $md5x, -1, 1, "" );
  46     $fx += hex $d;
  47     $fx /= 16;
  48
  49     $d = substr( $md5y, -1, 1, "" );
  50     $fy += hex $d;
  51     $fy /= 16;
  52 }
  53 printf "Fractions: %0.16f, %0.16f\n", $fx, $fy;
  54
  55 sub download_djia {
  56     my $date = shift;
  57     my $use_30w_rule = shift;
  58
  59     if ($use_30w_rule) {
  60        print "Adjusting for 30W\n";
  61        $date = DateCalc(ParseDate($datestring), "-1 day");
  62     }
  63
  64     my $datestring = UnixDate( ParseDate($date), "%Y-%m-%d" );
  65     my $djia;
  66
  67     require LWP::Simple;
  68     import LWP::Simple 'get';
  69     require HTML::TreeBuilder;
  70
  71     my $URL =
  72 'http://finance.google.com/finance/historical?cid=983582&startdate=%s&enddate=%s';
  73     $URL = sprintf( $URL,
  74         UnixDate( DateCalc( $date, "- 7 days" ), "%b+%d,+%Y" ),
  75         UnixDate( $date, "%b+%d,+%Y" ) );
  76     print "Downloading DJIA from google: $URL\n";
  77     my $page = get($URL);
  78     die "Failed to get DJIA from google!" unless $page;
  79     my $tree = HTML::TreeBuilder->new_from_content($page)
  80       or die "Failed to parse google DJIA page!";
  81
  82     my $data = $tree->look_down( id => "prices" ) or die "No data div: $!";
  83     foreach my $td ( $data->look_down( class => "firstcol" ) ) {
  84         next if $td->as_text eq 'Date';
  85         my @date = localtime;
  86         $djia = $td->right->as_text;
  87         $djia =~ s/,//g;
  88         print "DJIA opening for $datestring is $djia\n";
  89         last;
  90     }
  91
  92     $tree->delete;
  93
  94     die "Failed to retrieve DJIA from google!" unless $djia;
  95
  96     return $djia;
  97 }