Bash: Assign each line in file to array; if array[n] is null then replace value-CodePudding

I'm stuck on a bash problem. The script should read each line in data.txt and assign it to an array before checking if each array value is empty. If the array value is empty, the text "unknown" should replace the empty value. Line 2 would otherwise have the text $Author.

data.txt

07/22/2022, 00:00:00

NASA Reveals Webb Telescope's first images of unseen universe
"We are elated to celebrate this extraordinary day with the world," said Greg Robinson, Webb program director at NASA Headquarters.
webb, telescopes, images, reveals, galaxies, star, unseen, nasa, universe, nebula, webbs, gas, view, space

bashscript.sh

# assign each line in data.txt to array
readarray data < data.txt

# for each value in array
for i in "${data[@]}"; do

    # check if value is null
    if [ -z "$data[i]" ]; then

        # if value is null, assign text "unknown" to value
        data[i]="unknown"

    fi

done


Author=${data[1]}

echo "author: $Author"

ouput:

author:

expected output:

author: unknown

Thanks for reading/helping....

CodePudding user response：

The main problem with your script is that a string consisting of a new-line character isn't considered a null string. The check for the empty line should have been if [ "${data[i]}" = $'\n' ]. Also you should have used a C-style for loop. Below is a corrected version of your script:

#!/bin/bash

readarray data < data.txt

# for each value in array
for ((i = 0; i < ${#data[@]};   i)); do
    # check if value is null
    if [  "${data[i]}" = $'\n' ]; then

        # if value is null, assign text "unknown" to value
        data[i]="unknown"

    fi

done

Author=${data[1]}

echo "author: $Author"

Or, as suggested by glenn jackman, use the readarry with the -t flag and check for the empty string as [ -z "${data[i]}" ]

CodePudding user response：

In bash, try to avoid explicit iteration over the data. Such an approach is not only slow, but also prone to errors, especially for beginners. Instead, use tool designed for the job, which will do the looping for you.

sed 's/^$/unknown/' puts the string unknown into all empty lines.

To retrieve a single line, you can use head | tail or sed again.

Your entire script can be boiled down to

sed -n '2{s/^$/unknown/;s/^/author: /;p;q;}' data.txt