`read`¶

The read builtin utility is used to parse input into a variable.

One of its most common uses is to get input from a user via stdin, but it can be used in other ways as well.

Using `read`¶

Getting User Input¶

Typically when prompting for user input, you'd use read.

#!/bin/bash

declare MY_VAR

read -r -p "Enter input: " MY_VAR

printf "You entered: %s\n" "$MY_VAR"

read -r -p "Enter input: " MY_VAR:
- -r: Do not expand escape sequences (e.g., \n)
- -p "Enter input: ": Show this prompt to the user (Enter input:).
- MY_VAR: The variable name to save the input into.

This is a very simple way to interactively get input from a user and save it into a variable.

Reading Programmatically with `read`¶

The read builtin can also be used to parse lines into an array.

To do so, you direct an input stream to read via stdin, either via pipe (|) or input redirection (<).

#!/bin/bash

declare MY_FILE='./file.txt'
declare -a LINES
IFS=$'\n' read -r -d '' -a LINES < "$MY_FILE"

printf "File: %s\n" "${LINES[@]}"

IFS=$'\n': Set the internal field separator to a newline so elements are read into the array line-by-line.
-d '': This tells read to read until end of file (EOF), rather than end of line (EOL).
-a LINES: Specify the array to read into.

Reading Multiple Values at a Time¶

We can use read to split up lines into different variables.

read -r name age <<< "paul 50"
echo "Name: $name, Age: $age"
# output: 
# Name: paul, Age: 50

This is useful if you want to read in multiple variables from lines that are separated into columns (without using awk or cut).

For instance, if we wanted to utilize this to loop over lines with two columns:

tmpfile="$(mktemp)"
cat << EOF > "$tmpfile"
paul 50
jack 22
anna 30
tara 45
EOF

while read -r name age; do
    echo "Name: $name, Age: $age"
done < "$tmpfile"

This separates the elements based on the IFS value, which includes whitespace by default.

If we set IFS to null, it would not work as intended.

IFS= read -r name age <<< "paul 50" # BAD
echo "Name: $name, Age: $age"
# Output:
# Name: paul 50, Age:

If we want to be specific, and ensure that the two columns are split up into their respective variables, we can set the IFS to a space just for the read command.

IFS=' ' read -r name age <<< "paul 50"

This will guarantee that the line gets split on spaces.

If we also wanted to include tabs to account for different types of whitespace, we could add that to IFS as well.

IFS=$' \t' read -r name age <<< "paul 50"

ANSI-C Quoting

When setting the IFS to characters like tab (\t), use ANSI-C quoting ($'...') to make sure it's interpreted correctly, as a literal tab character instead of the literal \t characters.

More about ANSI-C Quoting here.

read¶

Using read¶