Shell Script: Is there a difference between "local foo" and "local foo="?

Question

I found following code in /etc/init.d/functions on CentOS.

status() {
    local base pid lock_file= pid_file=
    ...

4 variables are declared. Two of them are not initialized, base and pid. But rest of them are initialized with empty value, lock_file and pid_file.

I tested following code and found no differences.

local a b=

echo "a is $a, length is ${#a}"
echo "b is $b, length is ${#b}"

Is there any differences between them?

Answer 1

Yes, there is a difference. Consider the following function:

x() {
    local a b=

    echo ${a-X}
    echo ${b-X}
}

Calling this function in bash-4.x results in this output:

$ x
X

$

The ${parameter−word} parameter expansion expands to the expansion of word (in this case X) if the parameter is unset, or to the parameter value if it is set.

From the example output, it is obvious that local a leaves the variable a unset, while local b= explicitly sets it to the empty (null) string.

EDIT:

On the other hand, on bash-3.x you get this:

$ x


$

A call to set within the function verifies that local a in bash-3.x initializes that variable to the empty string. This, however, seems to have been a bug. From the bash changelog:

This document details the changes between this version, bash-4.0-beta, and the previous version, bash-4.0-alpha.

...

e. Fixed a bug that caused local variables to be created with the empty string for a value rather than no value.