Shell Programming and Scripting

Hi perl gurus,

The below lines are the 1st few lines of my logfile to scan.
Based on this, i need to get the TOP5 UserIDs having the most activities/transactions, and the top5 tasks as well.

For example, thomsonj (appearing 4x sofar) would be the top1 UserID, and WithdrawCash (4x) would be the top1 Task.

.....
.....
04:45:48 >> ==========START of MESSAGE==========
04:45:48 >> UserID: gibsonwe
04:45:48 >> Task: CheckBalance
04:45:48 >> Processing....
04:45:50 >> Response: Current balance is [$1]

07:30:27 >> ==========START of MESSAGE==========
07:30:27 >> UserID: thomsonj
07:30:27 >> Task: WithdrawCash
07:30:27 >> Processing....
07:30:27 >> Response: [$1] dispensed

10:18:09 >> ==========START of MESSAGE==========
10:18:09 >> UserID: greendav
10:18:09 >> Task: FundTransfer
10:18:10 >> Response: Successfully transferred [$1]

13:47:23 >> ==========START of MESSAGE==========
13:47:23 >> UserID: thomsonj
13:47:23 >> Task: WithdrawCash
13:47:23 >> Processing....
13:47:26 >> Response: [$1] dispensed

15:06:15 >> ==========START of MESSAGE==========
15:06:15 >> UserID: thomsonj
15:06:15 >> Task: CheckBalance
15:06:15 >> Processing....
15:06:15 >> Response: Current balance is [$1]

17:44:50 >> ==========START of MESSAGE==========
17:44:50 >> UserID: greendav
17:44:50 >> Task: BillPayment
17:44:50 >> Processing....
17:44:50 >> Response: Successfully paid [$1]

20:25:29 >> ==========START of MESSAGE==========
20:25:29 >> UserID: thomsonj
20:25:29 >> Task: WithdrawCash
20:25:29 >> Processing....
20:25:29 >> Response: [$1] dispensed
.....
.....


Any help would be greatly appreciated.
Thanks much.

what u can do is use hash in perl
1. open file in read mode
2. use regex to get filed after "User Id:"
3. put this "user Id" as key in Hash if not present
4. every time you find user present increase key value by 1
5. at the end sort keys of hash and do print.

following link might help

Perl Hash Howto

Giving a try:

Code:

perl -lne'
  $usr{$1}++ if /UserID: (\w+)/;
  $task{$1}++ if /Task: (\w+)/;
  if (eof) {
    $, = "\n";
    print "\nTop 5 users:\n", "=" x 12, 
      (sort { $usr{$b} <=> $usr{$a} } keys %usr)[0..4];
    print "\nTop 5 tasks:\n",  "=" x 12, 
      (sort { $task{$b} <=> $task{$a} } keys %task)[0..4]; 
    }' input

Thanks so much radoulov, it worked

But do u mind explaining a bit on the code.
Appreciate it mate.

people should really do their own school/class/course work

KevinADC is right,
I'm not sure if this is a homework (it may be) and this is not helpful (give vs teach a man a/to fish).
The interpretation of the code is left as an exercise to the OP.

Code:

sed 's/.*>>//' a > a.new
awk 'BEGIN{
FS=":"
un=0
tn=0}
{
if (index($1,"UserID")!=0)
	user[$2]++
if (index($1,"Task")!=0)
	task[$2]++
}
END{
	for (i in user)
		if(user[i]>un){un=user[i];u=i}
	for(j in task)
		if(task[j]>tn){tn=task[j];t=j}
	print u"  - "un
	print t"  - "tn
}' a.new
rm a.new