CODEWITHSUNDEEP
linuxscopeenvironment-variablesecho
santhosh kumar
santhosh kumar

Reputation: 21

Scope of Linux Environment Variables

Can someone explain the difference in scope of environment variables in the following 4 examples?

❯❯❯ b=1 echo $b

❯❯❯ b=1; echo $b
1
❯❯❯ 
❯❯❯ cat gen.sh
#!/bin/bash
echo $a
❯❯❯ 
❯❯❯ a=1 ./gen.sh
1
❯❯❯ a=1; ./gen.sh

Upvotes: 0

Views: 761

Answers (1)

linuxfan says Reinstate Monica
linuxfan says Reinstate Monica

Reputation: 4704

  1. Expansion of $parameters is done before executing a command, using the current environment.

  2. Assigning a parameter does not automatically export it to child processes (commands), but only on subsequent invocations of commands (in their command line, as per rule #1).

  3. But, a command in the form "a=b command" has an assignment that is exported to the command, but only temporarily - only for that single command invocation.

Given the three rules above:

❯❯❯ b=1 echo $b

gets first translated to: b=1 echo ""
... and prints nothing.

❯❯❯ b=1; echo $b

works, because there are two distinct commands, as if they were put on two different lines. The first is expanded and sets a variable; then the second is expanded.

#!/bin/bash
echo $a

(this script prints the $a found in the environment).

❯❯❯ a=1 ./gen.sh

This works because a=1 is exported, so gen.sh founds it in the environment.

❯❯❯ a=1; ./gen.sh

This does not work because a=1 is set but not exported to child processes; so gen.sh starts but does not find $a in the environment. The following instead would work:

❯❯❯ export a=1; ./gen.sh

Upvotes: 2

Related Questions