UNIX for Dummies Questions & Answers

Hey all, a relative bash/script newbie trying solve a problem.

I've got a text file with lots of lines that I've been able to clean up and format with awk/sed/cut, but now I'd like to remove the lines with duplicate usernames based on time stamp. Here's what the data looks like

Code:

2007-11-03 user1 xxxxxxxx
2008-06-30 user1 yyyyyyyy
2008-01-26 user2 bbbbbbbb
2007-11-03 user3 aaaaaaaa
2007-11-02 user4 cccccccc
2008-07-09 user4 dddddddd
2008-06-30 user4 eeeeeeee

What I'm trying to do is pull only the lines with the most recent date with unique accounts.

For example, I'd like the output to be:

Code:

2008-06-30 user1 yyyyyyyy
2008-01-26 user2 bbbbbbbb
2007-11-03 user3 aaaaaaaa
2008-07-09 user4 eeeeeeee

Some accounts are listed only 2 times, some more often. How can I delete the 'oldest' lines with that username?

Thanks for your help.

Code:

awk '!arr[$1 $2]++'  inputfile > newfile

anbu23,

Perfect! Thank you.