get the first 5 characters from each line in shell script

Question

Here is my sample.txt file it contains following

31113    70:54:D2 - a-31003
31114    70:54:D2 - b-31304
31111    4C:72:B9 - c-31303
31112    4C:72:B9 - d-31302

I have to write the shell script in that I am passing first 5 characters (eg 31113) as input id to other script. For this I have tried this

#!/bin/sh
filename='sample.txt'
filelines=`cat $filename`
while read -r line
do
  id= cut -c-5 $line
  echo $id
  #code for passing id to other script file as parameter
done < "$filename"

but it is not working this gives me error as

cut: 31113: No such file or directory
cut: 70:54:D2 No such file or directory
31114
31111
31112
: No such file or directory

How can I do this?

Answer 1

If you want to use cut this way, you need to use redirection <<< (a here string) like:

var=$(cut -c-5 <<< "$line")

Note the use of var=$(command) expression instead of id= cut -c-5 $line. This is the way to save the command into a variable.

Also, use /bin/bash instead of /bin/sh to have it working.

---

Full code that is working to me:

#!/bin/bash

filename='sample.txt'
while read -r line
do
  id=$(cut -c-5 <<< "$line")
  echo $id
  #code for passing id to other script file as parameter
done < "$filename"

Answer 2

Well, its a one-liner cut -c-5 sample.txt. Example:

$ cut -c-5 sample.txt 
31113
31114
31111
31112

From there-on, you can pipe it to any other script or command:

$ cut -c-5 sample.txt | while read line; do echo Hello $line; done
Hello 31113
Hello 31114
Hello 31111
Hello 31112

Answer 3

Rather than piping echo into cut, just pipe the output of cut directly to the while loop:

cut -c 1-5 sample.txt |
while read -r id; do
  echo $id
  #code for passing id to other script file as parameter
done

Answer 4

Maybe you need this, awk can recognize white space automatically.

awk '{print $1}' sample.txt

Answer 5

Please check the following simple example:

while read line; do id=$(echo $line | head -c5); echo $id; done < file

where head -c5 is the right command to get the first 5 characters from the string.

Answer 6

If you trying to fetch the first column from file try awk:

#!/bin/sh
filename='sample.txt'

while read -r line
do
  id=$(echo $line | awk '{print $1}')
  echo $id
  #code for passing id to other script file as parameter
done < "$filename"

Answer 7

Somewhat simpler than the top answer:

#!/bin/bash
filename='sample.txt'
while read -r line; do
  id=${line:0:5}
  echo $id
  #code for passing id to other script file as parameter
done < "$filename"

Answer 8

Use sed with a capture group that matches the first 5 characters and returns only that group:

sed -E 's/(.{0,5}).*/\1/' sample.txt

(.{0,5}) greedily matches any character up to 5 times, and creates a capture group.

.* matches the rest of the line, because we want to replace the entire line, not just the capture group.

\1 is a backreference which refers to the first capture group.

So we are capturing the group of 5 characters we want, and then replacing the entire matched line with only that capture group.