Usage

$ export KEY=VALUE

Or set a value first and export it later.

$ KEY=VALUE
$ export KEY

Export multiple values:

$ export KEY1=VALUE1 KEY2=VALUE2

See exported variables:

$ export
COLORFGBG='15;0'
COLORTERM=truecolor
HOME=/Users/mcurrin
HOMEBREW_NO_AUTO_UPDATE=1
...

That includes some system-set variables and some variables that I exported in my configs like ~/.bashrc.

Examples

Here the NAME variable is not set, so we default to using World.

$ NAME=developer
$ bash -c 'echo "Hello, ${NAME}."
Hello, .

In the flow, we export a new variable.

$ export NAME=developer

$ bash -c 'echo "Subshell - hello, ${NAME}."'
Hello, developer.

$ echo "Outer shell - hello, $NAME."
Outer shell - hello, developer.

Or export an existing variable.

$ NAME=developer
$ export NAME
$ bash -c 'echo "Subshell - hello, ${NAME}."'
Subshell - hello, developer.

Why do I need export

Here is when you don’t need export. If the variable is used in the same process.

$ NAME=developer
$ echo "Hello, $NAME!"
Hello, developer!

But if you run a subprocess, then that won’t be able to access the environment variables a level higher.

For example if you have Node or Python script which looks for NAME on the environment variables. Or a shell script which uses $NAME.

Here we in a Bash subshell, passing commands inline with -c flag. Or we could have used a separate script and called it as bash script.sh.

$ NAME=developer
$ bash -c 'echo "Hello, $NAME."'
Hello, .

Now we use export to make the variable accessible in the subshell.

$ export NAME=developer
$ bash -c 'echo "Hello, $NAME."'
Hello, developer.

Functions

Note: If you are in ZSH, then first run bash -l to make sure that you define and export the function in Bash, before using it in a Bash subshell.

foo () {
  echo "Foo $1"
}

foo bar
# Foo
echo $(foo bar)
# Foo bar

Export and use it in a subshell:

$ export -f foo
$ bash -c foo bar
Foo

Example use case:

$ export -f foo
$ find . -name README.md -exec bash -c 'foo $0' '{}' \;
Foo ./docs/README.md
Foo ./README.md
Foo ./vendor/bundle/ruby/2.7.0/gems/sass-3.7.4/README.md

Export config file

See dotenv file cheatsheet for more info.

If you had values to set by hand, you could do this:

$ export FOO=bar FIZZ=buzz

But what those in are a file?

  • .env 
      FOO=bar
  FIZZ=buzz

Then you can do this, which reads the contents of the .env shell file and passes the contents to export to evaluate it.

$ export $(xargs< .env)

Using < file.txt is a more efficient form of cat file.txt. See Useless use of cat award.

Check the export variables to see that it worked.

$ export | egrep '(FOO)|(FIZZ)'
FIZZ=buzz
FOO=bar

Breaking it down:

$ echo $(< .env) 
FOO=bar FIZZ=buzz

$ xargs < .env
FOO=bar FIZZ=buzz

Evaluate:

$ export $(xargs< .env)

I find it works without xargs sometimes but it is safer with, perhaps for multi-line values?

Warning - make sure to not have any comments in your file, as xargs makes the output all on one line and you’ll lose anything after the comment.