0

I have a sample file the lines of which are delimited by "#" and it contains 3 fields.

A#B#C
D#E#F

Can I use UNIX variable substitution instead of the following to get values of each field in the lines?

cat file | while read a; do
    f1=`echo $a | cut -d# -f1`
    f2=`echo $a | cut -d# -f2`
    f3=`echo $a | cut -d# -f3`
 done
Improve this question
2
  • Your example command overwrites the variables f1,2,3 with always the last line of your input. Not sure what you want to do with the variables then ... Commented Jul 15, 2019 at 14:21
  • I will be using f1,2,3 in values for each loop iteration, so the last value is expected to be values of the last line. Commented Jul 15, 2019 at 14:23

2 Answers 2

Reset to default
1

You can use parameter expansion if that's what you mean. But it's not very readable:

while read a; do
  # comments are for the first line A#B#C
  f1f2=${a%#*}  # remove suffix #C -> A#B
  f1=${f1f2%#*} # remove suffix #B -> A
  f2=${f1f2#*#} # remove prefix A# -> B
  f3=${a##*#}   # remove longest prefix: A#B# -> C
  echo "$f1,$f2,$f3" # do something with f1,f2,f3
done < file

It's easier to set the internal field separator variable to IFS=# and split each line into three variables:

while IFS=# read f1 f2 f3; do
  echo "$f1,$f2,$f3" # do something with f1,f2,f3
done < file
Improve this answer
1
  • @ Freddy: checked the while loop, works fine. Thank you. Commented Jul 16, 2019 at 8:50
1

In ksh, you can also use arrays:

while IFS='#' read -A arr; do
    echo "${#arr[@]} items, second item: ${arr[1]}"
done < file

read -A works in Zsh too, but the array indexes are one-based. Bash's read -a works similarly to the ksh one.

Improve this answer

You must log in to answer this question.