# wf2.sh: Crude word frequency analysis on a text file. #+ but using it is not quite as educational.Įxample 16-8. # The 'killall' command has the same effect as this script, # Can grep '$PROCESS_NAME' be replaced by pidof '$PROCESS_NAME'? # 2&>/dev/null suppresses unwanted error messages.

# The curly brackets are the placeholder for the replacement. # -i is the 'replace strings' option to xargs. # ctime = last status change time (via 'chmod' or otherwise)įind '$DIR' -type f -atime +5 -exec rm 2&>/dev/null # mtime = last modification time of the target file

# Same as above, but modified *exactly* one day ago. #+ that were modified within the last day (current_day - 1). # Lists all files in /home/bozo/projects directory tree # Removes all core dump files from user's home directory. Substitutes the full path name of the selected file for

Note that you can also use another file for a reference timestamp instead of explicitly specifying the timestamp, in which case the ‘t’ from the argument can be skipped.Literally, without interpreting it as a special character).īash$ find ~/ -name '*.txt' /home/bozo/.kde/share/apps/karm/karmdata.txt We have learned how to use the '-newerXY' the argument of command find to search for files based on timestamp. Verify with ‘stat’ if the timestamps of both files are indeed greater than the specified time. Let’s try this on a folder with few files and only two files having accessed, modified and status changed after a specified time. Here, the format of the timestamp should be: ‘YYYY-MM-DD HH:MM:SS’. To generalize, the command to be run looks like this: find -newert ‘’ creation time of a file is not maintained in Unix based file systems Thus in the argument '-newerXY', we can have X as a, c, or m, respectively for last access, last status change, and last modification times, and Y as t, so that we can have the next argument to be a timestamp string. Now, to find files based on the timestamp, we use the argument '-newerXY' of find from the man page.

Bash find files by date