Is there a way to use the Locate database to perform a copy in Unix Shell?

From a Unix shell.

I want to find a file specifically using locate <filename> and then I want to copy the files that were located using cp. I've tried something like:

Locate -name "*.doc" -exec cp {} /path/to/copy/to

If you want to perform actions with files-path found by the locate database, the locate command has no -exec option to execute a command on located files (-exec exist for the find command).

If working with the Locate database's locate command:

1. You probably want to locate from the base name rather than the whole path; so use -b '*.doc'.
2. You want the pattern inside single quotes so it is not expanded by the current shell interpreter, but only by the locate command.
3. You want the located files to actually exist so use the -e switch.
4. You want a null delimited list of entries, so it is safe to iterate paths, even if it contains non-printable special characters, spaces, newlines, tabs…
5. Since locate has no direct mean to execute a command with arguments, you will pipe the output stream into xargs to do it instead.

locate -0eb '*.doc' |
  xargs -0 \
    sh -c 'cp -- "$@" /path/to/copy/to/' _

Explanations:

• locate -0eb '*.doc': Using the locate database; locate files that still exist right-now having the base name matching the *.doc pattern, and output the list as a null delimited list of paths.
• | xargs -0: Pipe the null delimited list of paths to xargs to pass the entries as argument to the following command.
• sh -c: Execute the inline shell script that follow.
• cp -- "$@" /path/to/copy/to/': The inline shell script that copy all received paths as arguments array $@ into /path/to/copy/to/.

Finally note some key differences between locate and find:

• locate uses a database to reference files within the system. There is a programmed cron job, to update the Locate database (usually once a day, calling the /etc/cron.daily/mlocate script for example). locate is fast because of that, but it also cannot find recent files until it has updated its database.
• find directly accesses the filesystem. It traverses directories recursively and identify files/paths endpoint types. It also has more options and capabilities like passing paths as arguments and execute commands.

For further reading, see: SuperUser.com: What is the difference between 'locate' and 'find' in Linux?