Shell Programming and Scripting

Hi guys,

I have a log file that looks like this
Date Time Act Big Con
01/01/02 01:01:01 12 20 -34
01/23/03 01:23:42 21 -8 22
02/21/03 09:34:32 -87 33 29

I need to count how many positive entries were made within a give date range per column (i.e per Act, Big, Con).

Does anyone has any idea on how to go about this?

Thanks,

Odogbolu98

Of course, but it depends on your system, the tools you're using (for example, GNU `date` has more features than standard UnixWare `date`), and whether or not the file has the same format every time...

Thanks LF,

The tool I am using is Korn Shell and we are really stress on the box right now that performance is fast becoming an issue.

Any help on this issue will be highly appreciated.

You still didn't provide very much extra info though..

Is the format of every line in the log file the same as the few lines you've shown? Are the words Date Time Act Big Con actually on the first line of the file? Try explaining in as much detail what you'd like this program to do and how you'd like it to do it as you've got laid out in your mind.. Why'd you decide you'd like to be able to search between a range of dates? Why are you adding only the positive numbers (and what do Act Big and Con even mean)?

Just a few questions that will hopefully make it easier for everyone to help you make a script that does what you want..

Thanks oombera,

The data I provided was just a sample one, but here is a real live data.

Date Time W220A WI220B W220C W223A
3/26/2003 0:00:04 1 1 1 7
3/26/2003 0:15:02 1 1 2 8
3/26/2003 0:30:03 1 1 1 47
3/26/2003 0:45:01 1 0 1 6
3/26/2003 1:00:03 1 4 1 -7
3/26/2003 1:15:02 27 7 1 4
3/26/2003 1:30:02 1 1 1 8
3/26/2003 1:45:02 1 1 2 6
3/26/2003 2:00:03 1 -1 1 6
3/26/2003 2:15:00 1 1 1 8

The is log file for indicating the date and time of polls for each transaction indiated about. W220A - W223A are all transaction names. What I plan to do, is to count all the positive entries per transaction for the month and indicate the number of success(i.e 80% succes poll for W220A, etc). I hope this help.

I will appreciate any help or clue as to the most efficient way to do this.

Thanks,

Odogbolu98

Here is an example of how to read the file and select only the lines for a given month:

Code:

#! /usr/bin/ksh
givenmonth=3
exec < datafile
read titleline
while read Date Time W220A WI220B W220C W223A ; do
        month=${Date%%/*}
        ((month != givenmonth)) && continue
        echo $Date $W220A $WI220B $W220C $W223A
done
exit 0