I have a access log file on the webserver. the log contains info on userids , date/time and page acccessed. I need help with a shell scipt that can parse the file and compile a consilidated list of userids, and page accessed by time. Here's an example of 2 records fom the file. 182.178.112.13 - - [23/May/2002:07:39:01 -0400] "GET /mai-cgi/Home.cgi?userid=114736 HTTP/1.0" 200 12098 182.178.112.13 - - [23/May/2002:07:39:01 -0400] "GET /mai-cgi/TopMenuBar4HomePage.cgi?userid=114736 HTTP/1.0" 200 1413 Thanks in advance.

This is a job for awk, I am not going to write you all the code: the date reformatting would take a separate function in any case to arrange the data ordinally. A brief, untested, partial idea is this:: awk ' /\[.*:[0-9]+:[0-9]+[0-9]+-.*\]/ { sub(/[.*:/,"",$0) sub(/-.*\]/,"",$0) dates[cnt++","NR] = $0 } /\?userid=[0-9]+/ { userids[ucnt++","NR] = $0 } /\/[a-z]+-[a-z]\/[A-Za-z]+.cgi/ { pages[pcnt++",",NR] = $0 } Which gathers your data into three separate arrays and is now ready for you to sort and recalculate the data. see if someone on comp.lang.awk wants to write your program for you.