Command Substitution

Command substitution runs a command and uses its output as a value: $(command).

printf '%s\n' a b c d > items.txt

count=$(wc -l < items.txt)
echo "There are $count items"

first=$(head -n 1 items.txt)
echo "The first is $first"

echo "Uppercase: $(echo "shout" | tr a-z A-Z)"

There are 4 items
The first is a
Uppercase: SHOUT

• The output replaces the $(...), with trailing newlines removed.
• It works inside double quotes, and can be nested.
• wc -l < file (rather than wc -l file) prints just the number, without the file name.

Everyday uses

# illustrative
today=$(date +%Y-%m-%d)
backup="backup-$today.tar.gz"
branch=$(git rev-parse --abbrev-ref HEAD)
echo "Creating $backup on branch $branch"

Creating backup-2025-04-22.tar.gz on branch main

Backticks

Older scripts use backticks: count=`wc -l < items.txt`. It means the same thing, but $(...) is easier to read and nests cleanly — prefer it.

Common mistakes

• Forgetting quotes around the result: ls $(find ...) breaks on names with spaces. Quote it — "$(command)" — when using it as a single value.
• Using command substitution to read big outputs into a variable when a pipe would do.