#!/perl/bin
use strict;


# rand.pl - Version 1.0 (Aug 3th, 1998)
#
# Perl-Script by Kajetan Hinner. This program is Mailware, every user should write to Kajetan@Hinner.com
# 
# Also note that there is no taint checking AT ALL. This could be a security problem, if you don't
# know the quality of input data fed into rand.pl.
# 
#
# rand.pl was developed using Linux 2.1.x, ircII 4.4, Perl 5.004 and KDE 4. As editor for large files I recommend nedit. 
#
#
# How this script works:       
#
# This perl program was developed for tracedomains.pl (www.hinner.com/programs). But it's also
# useful to randomly select a certain number of lines from an ASCII input file.
# Not that the input file is read at once, so supply enough memory (RAM) in your machine.
#
# You must supply a file named Domains.txt. Then you type perl rand.pl <number>. <number>
# is the number of lines you want to select randomly from the input file Domains.txt.
# Two random numbers are taken, the first two and the last character are cut off and those
# strings are concatenated. From the middle there is the actual random number taken. If it's
# to large (index for line of input line), it's divided by two until it fits. This specific
# line number of the input file (which must be named Domains.txt) is written to the outfile
# Domains.rand
#
#
# Author: Kajetan Hinner, University of Rostock, Sociology Dep., Kajetan@Hinner.com
#










srand;			#set random seed
use Math::TrulyRandom;

my $i = 1; 		# counter

my $zeile = 1;		# aktuelle Zeile

my $domain = ""; 	# zusammengebastelter Domainname
my $domaincounter =0;	# number of valid domains in infile

my @linesread;		# eingelesene datei, liste

my $file = "Domains.txt";		# sample infile
my $randfile = "Domains.rand";	# outfile, random sample of $outfile


my $wanteddomains;			# number of random domains wanted


#scan command line arguments

my $self = $0;
$self =~ s!^.*/!!;
if (@ARGV != 1) { die "Usage: perl $self <wanted random domains#>\nInput file must be Domains.txt in current path!\n"; }

$wanteddomains = $ARGV[0];		# get parameter


open (FILEH, "< $file") or die "can't open $file: $!";

@linesread = <FILEH>;		# read whole files in one big slurp

close FILEH;

die "error while reading $file\n" unless @linesread;


open (FILEOUT, "> $randfile") or die "can't open $randfile: $!";

my @randomvalues;
my @selected;		# list; every already used random is flagged
my @index;		# all random values which are used (index to infile line#)

my $value;		# random value

my $i=0;		# while-loop-counter

my $lines = @linesread;	# number of lines

while ($i<$wanteddomains) {

 my $value1=truly_random_value();
 my $value2=truly_random_value();
 
 print ("Plain random values: $value1, $value2\n");
 

 # chop off first two and last digit
 $value1 =~ /..(.*)./; $value1=$1;
 $value2 =~ /..(.*)./; $value2=$1;
 $value = $value1 . $value2;		# concatenate both values.
 print "built figure: $value, $_\n";
 
 $randomvalues[$i]=$value;		# save for later analysis
 
 # print ("calculated value: $value\n");

 my $length = length($value);
 my $index = substr ($value, $length/3, length($lines));
 # print ("length $length, index $index\n");
 
 while ($index > $lines)
      {
       $index /= 2;
      }

 $index =~ s/\.\d*$//;
       
 
 if (!($selected[$index]))
   {
    $selected[$index]=1;
    $index[$i]=$index;
    $i++;		# next value
    print ("$i: selected $index\n");
   }
 
 }			# while loop


for ($i=0; $i<$wanteddomains; $i++)

 {
 (print FILEOUT "$linesread[$index[$i]]")   or die "can't write to $randfile $!";    
 }     
   
close (FILEOUT);

my @randsorted = sort {$a <=> $b} @randomvalues;         # sort numerically ascending

print ("Some statistical information about used random values\n");

my $min=$randsorted[0];
my $max=$randsorted[@randsorted-1];
my $range=$max-$min;

print ("min: $min, max: $max, median: $randsorted[@randsorted/2]\n");