I have the following file that looks simular to this:

68 1 2 10
68 3
68 4
107 1
107 10
5 1 10

How can I make the file look like this:

5 1 10
68 1 2 3 4 10
107 1 10

#!/opt/third-party/bin/perl

my %fileHash;
my @arr;

open(FILE, "<", "a") || die "Unable to open file 'a'<$!>\n";

while(<FILE>) {
  chomp;
  @arr = split(/ /);
  for(my $i = 1; $i <= ($#arr + 1); $i++ ) {
    $fileHash{$arr[0]} .= " $arr[$i]";
  }
}

close(FILE);

foreach my $key ( sort keys %fileHash ) {
  print "$key ==> $fileHash{$key}\n";
}

exit 0

Thanks for all your help! My trouble is that I have never used perl before. Is there a way to do this using awk?

if order is not important

awk '{  col1 = $1
	    $1=""
	    array[col1]=$0array[col1] 
	 }
END{
    for ( i in array ) print i " "array[i]	
}' "file"

# ./test.sh
5  1 10
68  4 3 1 2 10
107  10 1

something to start with:

nawk -f jojo.awk myFile.txt

jojo.awk:

{
 vars=substr($0, index($0, FS)+1)
 arr[$1] = ($1 in arr) ? arr[$1] FS vars : vars
}
END {
  for (i in arr)
    print i, arr[i]
}

Here is a Korn shell version:

mTemp='Temp.txt'
mLast=''
sort -n input_file | \
while read mLine
do
  set -- `echo $mLine`
  if [ "$1" != "${mLast}" ]; then
    if [ "${mLast}" != "" ]; then
      (echo ${mLast};sort -n ${mTemp}) | paste -d' ' -s -
    fi
    rm -f ${mTemp}
    mLast=$1
  fi
  shift
  echo $@ | tr ' ' '\n' >> ${mTemp}
done
if [ "${mLast}" != "" ]; then
  (echo ${mLast};sort -n ${mTemp}) | paste -d' ' -s -
  rm -f ${mTemp}
fi

This will work for me, thanks for all the help.