`find`

find can be used both to locate files and run operations on the files it finds.

Main syntax

Without options specified, find alone will return a recursive index of all the files in the directory from which it is run.

If we pass a directory to find it will repeat the above process but specifically for that directory.

$ find i3
i3
i3/config

We can specify flags as filters (known as ‘tests’ within the program).

Filter by type: file or directory

$ find -type d # return dirs only
$ find -type f # return files only

Within a specified directory:

$ find i3 -type f

This is the most frequent use case: filter files by name with globbing.

$ find -name "config"
./.git/config
./i3/config

$ find -name "*.js"

The same, but case insensitive: iname

$ find -iname "*.JS"

As above but this time includes directory names in the match. ipath is the case-insensitive version.

$ find -path "utils*"
utils.js
utils/do-something.js

We can combine find commands by using logical operators: -and, -or, -not. For example:

$ find -not -name "*.js" -type f
./app/index.html
./app/style.css
./dist/index.html
./dist/style.c

Applied to a directory:

find . -type -f -not -path "./.git/"

Using the exec keyword we can run a program against the files that are returned from find.

In this syntax we use {} as a placeholder for the path of the file that is matched. We use ; (escaped) to indicate the end of the operation.

This script deletes the files that match the filter criteria:

$ find -name "*.js" -exec rm {} \;

This script finds all the files with the substring ‘config’ in their name and writes their file size to a file.

find -name '*config*' -exec wc -c {} \; > config-sizes