Why doesn't quantifier work with character classes?

[rnaeye]

Hi! Monks,
I am trying to select lines that contain certain length of DNA string. Following code will print any DNA string that is 8 nucleotide or longer. However, I want to print DNA strings that are exactly 8 nucleotide long, such as "ATGATGAC". I thought {8} will match exactly 8 characters, but looks like I am wrong! I also tried ATGC{8,8}; did not work either.
In addition, in a separate program, I want to select DNA strings that are between 8-21 nucleotide long. Can you please give me any suggestions?
Thank you.
PS. I was able to solve this problem using "length" function without using any regex, but I would like to learn the regex solution to this problem.

#!/usr/bin/perl
use warnings;
use strict;

while (<DATA>){
    my $string = $_;
    if ($string =~ /[ATGC]{8}/) {
        print "$string";
    }
}

__DATA__
@3009W:27:32
GCTCT
+
%.8:9
@3009W:27:40
TTGGG
+
0(*2+
@3009W:31:26
AGCCT
+
5<=46
@3009W:31:35
TCAGAAAACTG
+
0.5*.--%-0-
@3009W:32:34
GGGCCTAACCTGGGAGCCCCT
+
A@.:158+,--*-%-**--%-
@3009W:34:32
CCATCATCTGGGG
+
:-:>>;;55755&
@3009W:36:21
GACTT
+
(8.7(
@3009W:40:24
ATGATCC
+
44.0,.%
@3009W:42:22
GCTTCCAGGGTCAGTTTGGGAAAC
+
:@>4;4888)1//**-%+5+25,.
@3009W:47:23
GAGCATCGA
+
%*1.0...-
@3009W:49:23
GAGTTCCATCGAAATGTACAAGCTTTACGTTTAAAAC
+
/3....0304036-22.,--(*.09*00,11),00(.
@3009W:14:90
AGCAA
+
82528
@3009W:17:84
GAAACACAC
+
05?4=:<:0
@3009W:17:95
TTTTTCTTT
+
;<<<-07<1
@3009W:19:89
CCTCTACC
+
?:>>:;83
@3009W:19:90
AAGAA
+
:4<;2
@3009W:20:74
GGTTCC
+
2&-.2.
@3009W:22:94
CATTTGGAA
+
AAAB9>8>:
@3009W:23:79
CTTACAA
+
@@9@@@@
@3009W:23:93
TCTTTTTC
+
@@@AAA/A
@3009W:24:80
GTGAGC
+
<AAA@@
@3009W:25:79
AATAT
+
?8=.0
@3009W:26:89
AGGCA
+
BB>BC
@3009W:26:99
ATCCATAT
+
/88(3979
@3009W:27:83
AGGCA
+
AA>@@

[BrowserUk]

I thought {8} will match exactly 8 characters, but looks like I am wrong!

You aren't wrong, it does match exactly 8 characters ... but if those 8 characters are at the start of a line containing more than 8 characters, it still matches exactly the first 8.

You need to anchor your regex: Ie. /^[ATGC]{8}$/. Now it will only match lines that contain exactly 8 characters.

[thezip]

Please correct me if I am wrong (or if I misunderstood your statement), but there is nothing in OP's regex that restricts a match to the beginning of a line. Hence, the eight consecutive characters may appear anywhere on a line. Right?

[BrowserUk]

Hence, the eight consecutive characters may appear anywhere on a line. Right?

Yes, but since the lines that match the character class in question contain only those characters; and a regex will always match as early as possible; that'll be at the beginning of the line in these cases.

[rnaeye]

Correct!
"BrowserUk" and "thezip" 's suggestions solved the problem. Thanks.

[thezip]

Try:

...
if ($string =~ /^[ATGC]{8}$/) {
...

Note the caret and dollar-sign. This should match sequences that are exactly eight characters long.