Find is one of the most useful commands in the entire Unix operating system, yet its slightly unusual syntax keeps many from enjoying its utility. The find command has a three-part syntax:
find <where to start finding> <what to find> <what to do once it's found>
An example:
find /home/support/marty -size +1000k -print
This will start looking in my home directory for files larger than 1 MB (1000 kB) and will print them to standard output once it finds them. Had I specified "-1000k" instead of "+1000k" it would look for files smaller than 1 MB.
Make certain that special shell characters are escaped:
find . -name \*tex -print
The possibilities are endless; You can tell find whether or not to cross filesystems, NFS mount points, or whether or not to follow symbolic links. You can specify the "what to find" section as a boolean string, and can also operate on the files it finds with the -exec option:
find . -type f \( -mtime -90 -a -size +1000k \) -exec rm {} \;
This finds plain files that haven't been accessed in any way (not even
read) for at least ninety days, and (-a) whose
size is at least 1MB. It then removes these files.
The curly braces {} equates to the output of
find. The above example might read "remove whatever find finds.
The escaped semiclon \; is a long story. It's needed when doing an
-exec.
The above has barely scratched the surface of what may be done with find. This page's purpose is twofold: