#!/usr/bin/perl
# File  : fixdups.pl
# Author: Lyndon Hill, http://www.lyndonhill.com
# Note  : This script is free to use. If you make significant improvements please
#         send me a copy.
#
# Script to fix duplicate entries (generate duplicate list first)
# Usage:
# fixdups.pl
#
# A temporary file "dupscript" will be created. Please delete it.
# The output of the script will be a file with the name dictionary-new
# where "dictionary" is the name you specify in the $dictionary variable
# a few lines below.

# User editable variables

# the dictionary file with duplicates
$dictionary = "en-ka.bedic";

# the name of the duplicate list file
$duplicatelist = "duplicate-list";

# End of user editable variables

# Load in duplicate list
open(DUPLIST, $duplicatelist);

$duplicates = 0;
$i = 0;

while(<DUPLIST>)
{
  $line = $_; chop $line;
  ($blah1, $cutstart[$duplicates], $blah2, $cutend[$duplicates], $blah3, $blah4,
  $entrystart, $blah4, $destination[$duplicates]) = split / /, $line; # find parameters
  $duplicates++;
}

close(DUPLIST);

# modify line numbers to take into account cut and paste
for($i = 0; $i < $duplicates; $i++)
{
  for($j = $i+1; $j < $duplicates; $j++)
  {
    # shift every following duplicate up by two lines
    $cutstart[$j] -= 2;
    $cutend[$j] -= 2;
 
    if($destination[$j] > $destination[$i])
    {
      if($destination[$j] > $cutend[$i])
      {
        # small shift if straightforward merging
        $destination[$j] -= 2;
      }
      else
      {
        # shift down destinations for entries after paste position (ordered merging)
        $destination[$j] += ($cutend[$i] - $cutstart[$i]);
      }
    }

    # make multiple merges point to end of current merge
    if($destination[$j] == $destination[$i])
    {
      $destination[$j] = $destination[$j] + ($cutend[$i] - $cutstart[$i]);
    }

  }
}

# output modification file
open(DUPMODIFY, ">dupscript");

print DUPMODIFY "H\n";

for($i = 0; $i < $duplicates; $i++)
{
  print DUPMODIFY ($cutstart[$i]-1) . ",$cutstart[$i]d\n"; # delete entry line at duplicate and blank line before
  print DUPMODIFY ($cutstart[$i]-1) . "," . ($cutend[$i]-2) . "m" . ($destination[$i]) . "\n"; # move body of duplicate to end of original entry
}

print DUPMODIFY "w $dictionary-new\n";
print DUPMODIFY "q\n";
close(DUPMODIFY);

# perform the moves using ed
`ed - $dictionary < dupscript`;

print "Dictionary file duplicate entries now joined together.\n";
exit;