Extract filename and extension from a string in Bash

The Problem

Given a file path string in Bash, how can I reliably extract the filename and extension, even in cases where the name contains multiple .s? For example, I would like to extract the name “requirements.updated” and extension “txt” in separate variables from this string:

/home/user/requirements.updated.txt

The Solution

We can achieve this in Bash using the basename command and parameter expansion.

First, extract the filename from the path using basename:

filepath="/home/user/requirements.updated.txt"
filename_with_ext=$(basename "$filepath")

Next, use parameter expansion to separate the filename and extension:

filename="${filename_with_ext%.*}"
extension="${filename_with_ext##*.}"

In the first line, the ${filename_with_ext%.*} parameter expansion will remove the shortest match of .* from the end of $filename_with_ext, removing only the final . and text after it.

In the second line, the ${filename##*.} parameter expansion will remove the longest match of *. from the beginning of $filename_with_ext, leaving us with just the text after the final ..

Our full Bash script will look like this:

#!/bin/bash

filepath="/home/user/requirements.updated.txt"
filename_with_ext=$(basename "$filepath")
filename="${filename_with_ext%.*}"
extension="${filename_with_ext##*.}"

echo "Path: $filepath"
echo "Filename with extension: $filename_with_ext"
echo "Filename without extension: $filename"
echo "File extension: $extension"

When run, this script will produce the output below:

Path: /home/user/requirements.updated.txt
Filename with extension: requirements.updated.txt
Filename without extension: requirements.updated
File extension: txt