Shell Programming and Scripting

Hi All
I have a CSV file where 3rd and 4th fields are date and time
it's a session log file printing the date and time per session
Now, I want to create a statistics of how many sessions per date and Time
I managed a script for sessions per hour

Code:

#!/bin/bash

#This Shell Script recieves as input a csv file, removes the first line of the file and calculates the number of sessions instances per hour
#Enjoy :-)


#deleting the statistics file and file with no first line from the system

rm -rf /tmp/scripts/stat_file
rm -rf /tmp/scripts/New_File_Name

#Inputing the File name to be analyzed

read -p "Enter The File Name to Analyze -> " File_Name

#Removing the first line which contains the field names

tail +2 $File_Name > /tmp/scripts/New_File_Name

#printing the time field from the csv file, printing just the hour, counting instances of each hour and redirecting to a statistics file

awk -F ',' '{ print $4 }' /tmp/scripts/New_File_Name | awk -F ':' '{ print $1 }' | uniq -c >> /tmp/scripts/stat_file

#printing the first and second field from the output file with explanations.

awk '{ print "there were "$1" sessions at "$2" AM" }' /tmp/scripts/stat_file


now, I want to add the dates also. date is field $3 in the original file and is in dd-mm-yy format.

I'm stuck :-(
Any ideas?

thanks in advance
The owl

How do you want to "add" the dates? Do you want to count how many times each day appears?

Code:

awk -F, '{ date_count[$4]++; } END { for (d in date_count) { print d,"-",date_count[d]; }}'

After this, you can sort it however you want. You should first, however, convert the months:

Code:

awk 'sub("Jan",1) || sub("Feb",2) || sub("Mar",3)  (... etc)' | \
sort -t - -k 4n -k 3n -k 2n

Hi
This is command I'm using

awk -F ',' '{ print $3, $4 }' /tmp/scripts/New_File_Name | awk '{ print $1, $2 }'

This is the output

27-01-08 03:44:41.018
27-01-08 03:47:19.711
27-01-08 03:49:22.438
27-01-08 03:54:07.578
27-01-08 03:56:05.518
27-01-08 04:06:24.786
27-01-08 04:08:17.956
27-01-08 04:08:26.964
27-01-08 04:09:33.353
27-01-08 04:11:28.766
27-01-08 04:11:35.338
27-01-08 04:11:41.080
27-01-08 04:12:47.780
27-01-08 04:12:55.726
27-01-08 04:13:02.687
27-01-08 04:13:08.742
27-01-08 04:13:17.915
27-01-08 04:22:59.725
27-01-08 04:24:14.726
27-01-08 04:26:21.373
27-01-08 04:27:37.617
27-01-08 04:29:49.979
27-01-08 04:31:28.018
27-01-08 04:32:00.765
27-01-08 04:38:26.139
27-01-08 04:40:18.062
27-01-08 04:41:35.777
27-01-08 04:42:39.187
27-01-08 04:44:52.026
27-01-08 04:51:33.561
27-01-08 04:53:49.810
27-01-08 05:01:08.307
27-01-08 05:02:02.499
27-01-08 05:03:03.190
27-01-08 05:03:50.279
27-01-08 05:04:53.827
27-01-08 05:05:48.014
27-01-08 05:07:38.179
27-01-08 05:08:26.018
27-01-08 05:09:57.883
27-01-08 05:11:32.100
27-01-08 05:54:34.025
27-01-08 05:55:39.882
27-01-08 05:56:44.151
27-01-08 05:58:53.207
27-01-08 06:01:05.527
27-01-08 06:02:35.431
27-01-08 07:10:04.648
28-01-08 07:11:04.648
28-01-08 07:12:04.648

What I want to do is create a responce which sais:
on the 27-01-08 the following lines appeard:
5 lines at 03:00
26 lines at 04:00

etc...

how do I do that?

thanks again
The Owl

Create an array element for each date and hour. So for simplicity, filter this output to awk again:

Code:

awk -F"[: ]" '{ ts[$1 "." $2]++;}  END { for (key in ts) { print ts[key] " lines at " key ":00"; }'

Do you see it now? Array indexes in awk are just strings (called associative arrays or hashes). So you create the index based on the first field (the date) and the second field (the hour). If the array index already exists, the code increments what's already there. Otherwise, it just becomes 1. You can also do two-dimension arrays in awk. Check the man pages for how to do that, though.