How to use find command to delete files matching a pattern?

Solution 1:

I generally find that using the -exec option for find to be easier and less confusing. Try this:

find . -name vmware-*.log -exec rm -i {} \;

Everything after -exec is taken to be a command to run for each result, up to the ;, which is escaped here so that it will be passed to find. The {} is replaced with the filename that find would normally print.

Once you've verified it does what you want, you can remove the -i.

Solution 2:

If you have GNU find you can use the -delete option:

find . -name "vmware-*.log" -delete

To use xargs and avoid the problem with spaces in filenames:

find . -name vmware-*.log -print0 | xargs -0 rm

However, your log files shouldn't have spaces in their names. Word processing documents and MP3 files are likely to have them, but you should be able to control the names of your log files.

---

Solution 3:

You can tell find to delimit the output list with NULLs, and xargs to receive its input list the same:

$ ls -l "file 1" "file 2"
-rw-r--r-- 1 james james 0 Oct 19 13:28 file 1
-rw-r--r-- 1 james james 0 Oct 19 13:28 file 2

$ find . -name 'file *' -print0 | xargs -0 ls -l
-rw-r--r-- 1 james james 0 Oct 19 13:28 ./file 1
-rw-r--r-- 1 james james 0 Oct 19 13:28 ./file 2

$ find . -name 'file *' -print0 | xargs -0 rm -v
removed `./file 2'
removed `./file 1'

Also, make sure you escape the *, either with a backslash, or by containing the vmware-*.log in single quotes, otherwise your shell may try to expand it before passing it off to find.

---

Solution 4:

Dont't forget the find's -delete option. It remove the file without error with special characters...