how to restrict length of string present in a line using linux

Question

I have data of the following form:

num1    This is a string
num2    This is another string

I want to limit length of all strings which are after the first tab..such that length(string)<4. Therefore, the output which I get is:

num1    This is a string
num2    This is another

I can do this using python. But I am trying to find a linux equivalent in order to achieve the same.

You say "length(string)<4", but I don't see anything in your output that's consistent with that; for example, `"This is another"` is 15 characters long. — Keith Thompson, Nov 08 '13 at 22:34
This is almost the same as http://stackoverflow.com/q/19804806/3165552 — isaias-b, Feb 14 '16 at 14:23

score 33 · Answer 1 · edited Mar 22 '19 at 15:39

33

In bash, you can use the following to limit the string, in this case, from index 0 to index 17.

$ var="this is a another string"

$ echo ${var:0:17}

this is a another

edited Mar 22 '19 at 15:39

TrebledJ

answered Nov 08 '13 at 22:23

jramirez

Yes you are right...i have a very big file..and i want to automate the procedure...instead of doing the same 1 line at a time manually – Jannat Arora Nov 08 '13 at 22:29

Gilles Quénot · Accepted Answer · 2013-11-08T22:38:14.960

19

Using awk, by columns :

$ awk '{print $1, $2, $3, $4}' file

or with sed :

sed -r 's@^(\S+\s+\S+\s+\S+\s+\S+).*@\1@' file

or by length using cut :

$ cut -c 1-23 file

edited Nov 08 '13 at 22:38

answered Nov 08 '13 at 22:28

Gilles Quénot

Yes you are right...i have a very big file(1 TB)..and i want to automate the procedure...instead of doing the same 1 line at a time manually – Jannat Arora Nov 08 '13 at 22:29
is there some way of iteraing through each line of file..along with cut – Jannat Arora Nov 08 '13 at 22:33
That's the default behaviour of `cut` if you provide a file as argument. – Gilles Quénot Nov 08 '13 at 22:34

score 0 · Answer 3 · answered Nov 09 '13 at 00:24

If you'd like to truncate strings on word boundaries, you could use fold with the -s option:

awk -F"\t" '{
    printf "%s\t", $1; system(sprintf("fold -sw 17 <<< \"%s\" | sed q", $2))
}'

The drawback is fold and sed need to be called for each line (sed q is the same as tail -n1).

3 Answers3