Last updated

Linux and Unix pwd command tutorial with examples

Tutorial on using pwd, a UNIX and Linux command for printing the name of the current working directory. Examples of printing the current working directory, avoiding symlinks and how to get the current working directory in shell scripts.

Estimated reading time: 3 minutes

Table of contents

Terminal showing pwd man page

What is the pwd command in UNIX?

The pwd command is a command line utility for printing the current working directory. It will print the full system path of the current working directory to standard output. By default the pwd command ignores symlinks, although the full physical path of a current directory can be shown with an option. The pwd command is normally a shell builtin meaning it is part of the code that runs the shell rather than an external executable.

How to print the current working directory

To print the current working directory run the pwd command. The full path of the current working directory will be printed to standard output.

pwd
/home/george

To avoid symlinks pass the -P option. This will cause pwd to show the physical location rather than a symlink. To demonstrate this the following example shows a folder that is symlinked to another location. Without the -P option the symlink is ignored. In the following example the folder zsh is a symlink.

ls -la | grep -E '.zsh$'    
lrwxrwxrwx    1 george users    25 Sep 15  2013 .zsh -> /home/george/dotfiles/zsh

Running pwd within the zsh folder does not show the symlink.

cd .zsh
pwd
/home/george/.zsh

Running pwd with the -P shows the physical location.

cd .zsh
pwd -P  
/home/george/dotfiles/zsh

pwd is normally a shell builtin

In most shells pwd is a shell builtin. This means the command is present in the shell rather than calling an external program. This means that the code will run significantly faster than calling an external executable.

which pwd
pwd: shell builtin command

Whilst most shells have pwd as a shell builtin the command also exists on systems as an executable. On my own system the executable is located at /bin/pwd.

/bin/pwd
/home/george/dotfiles/zsh

How to reference pwd in shell scripts

In most shells the $PWD variable is available and is set each time a user or in script changes directory. As such this variable can be referenced to show the current working directory.

echo $PWD
/home/george

Although $PWD is ubiquitous it is specific to the shell running the command so some prefer to reference the pwd command and store it in a variable. This makes scripts more portable as they will call the system executable if the shell builtin or the $PWD variable does not exist.

CWD=$(pwd)
echo $CWD
/home/george

Althought the $PWD variable is present is almost all shells it is possible it may not be there. If you are really paranoid use CWD=$(pwd) but in most cases $PWD will work.

Further reading

Have an update or suggestion for this article? You can edit it here and send me a pull request.

Tags

Recent Posts