cut
Picking out columns.
Many files are organised in columns. Our log has a date, a time, a level, a user, and a message. Often you only need one column. Cut picks it out.

- Files with columns
- You need just one
- cut picks it out
Cutting fields
Cut dash D tells cut what separates the columns. Here, a space. Dash F picks the field number. Field four is the user. So this prints just the user names, one per line, for the first five lines of the log.
head -n 5 logs/server.log | cut -d " " -f 4 priya sam sam priya priya
Several fields
You can ask for several fields, separated by commas. Fields two, and three, give the time and the level. Cut is perfect for CSV files too, where columns are separated by commas.
head -n 4 logs/server.log | cut -d " " -f 2,3 09:00:23 ERROR 09:00:29 INFO 09:00:55 INFO 09:01:00 INFO
Who uses the site most?
Now put it all together. Cut out the user field. Sort the names. Count them with uniq dash C. Then sort the counts in reverse number order. Sam is our most active user. Five small commands, one real answer.
cut -d " " -f 4 logs/server.log | sort | uniq -c | sort -rn 18 sam 16 alex 15 priya 11 jo
Summary
Cut picks columns with dash D and dash F. And cut, sort, uniq dash C, sort dash RN is a famous pipeline for counting things. Next, a few more text tools.
Key takeaways- cut -d ' ' -f 4
- -f 2,3 for several
- cut | sort | uniq -c | sort -rn