How to get number in range 10-20 in grep

Question

I need from this file extract line that starts with a number in the range 10-20 and I have tried use grep "[10-20]" tmp_file.txt, but from a file that has this format

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.aa
12.bbb
13.cccc
14.ddddd
15.eeeeee
16.fffffff
17.gggggggg
18.hhhhhhhhh
19.iiiiiiiiii
20.jjjjjjjjjjj
21.

it returned everything and marked every number that contains either 1, 0, 10, 2, 0, 20 or 21 :/

See https://stackoverflow.com/questions/47330069/grep-range-of-numbers — The fourth bird, Feb 24 '22 at 21:23
is the file guaranteed to be sorted numerically by the first column? — markp-fuso, Feb 24 '22 at 21:27
grep is for finding text patterns, not for determining numeric values. — Andy Lester, Feb 24 '22 at 22:41

score 4 · Answer 1 · answered Feb 24 '22 at 21:15

4

With an extended regular expression (-E):

grep -E '^(1[0-9]|20)\.' file

Output:

10.aa
12.bbb
13.cccc
14.ddddd
15.eeeeee
16.fffffff
17.gggggggg
18.hhhhhhhhh
19.iiiiiiiiii
20.jjjjjjjjjjj

See: The Stack Overflow Regular Expressions FAQ

answered Feb 24 '22 at 21:15

Cyrus

84,225
14
89
153

ufopilot · Accepted Answer · 2022-02-24T21:32:15.153

1

An other one with awk

awk '/^10\./,/^20\./' tmp_file.txt

awk '/^10\./,/^13\./' tmp_file.txt
10.aa
12.bbb
13.cccc

edited Feb 24 '22 at 21:32

answered Feb 24 '22 at 21:26

ufopilot

3,269
2
10
12

score 1 · Answer 3 · answered Feb 24 '22 at 22:34

Try

grep -w -e '^1[[:digit:]]' -e '^20' tmp_file.txt

-w forces matches of whole words. That prevents matching lines like 100.... It's not POSIX, but it's supported by every grep that most people will encounter these days. Use grep -e '^1[[:digit:]]\.' -e '^20\.' ... if you are concerned about portability.
The -e option can be used multiple times to specify multiple patterns.
[[:digit:]] may be more reliable than [0-9]. See In grep command, can I change [:digit:] to [0-9]?.

score 1 · Answer 4 · answered Feb 24 '22 at 22:59

1

Assuming the file might not be sorted and using numeric comparison

awk -F. '$1 >= 10 && $1 <= 20' < file.txt

answered Feb 24 '22 at 22:59

Diego Torres Milano

65,697
9
111
134

score 0 · Answer 5 · answered Feb 24 '22 at 22:47

grep is not the tool for this, because grep finds text patterns, but does not understand numeric values. Making patterns that match the 11 values from 10-20 is like stirring a can of paint with a screwdriver. You can do it, but it's not the right tool for the job.

A much clearer way to do this is with Perl:

$ perl -n -e'print if /^(\d+)/ && $1 >= 10 && $1 <= 20' foo.txt

This says to print a line of the file if the beginning of the line ^ matches one or more digits \d+ and if the numeric value of what was matched $1 is between the values of 10 and 20.

How to get number in range 10-20 in grep

5 Answers5