How can I regexp capture the string between 2 specific sets of double underscores?

David Gauthier 0
Viewed 64

I want to regexp capture the string between 2 specific sets of double underscores. The string that get captured may itself have single underscore occurrences in it. Here's the test Perl script I've been working with:

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

my $str = "DFD_20220913_121409_strix1a0__z1_erx_adm__CL1695331__RTL_Dfdsg4__regression__df_umc_nbio_hubs_gfx__220913_150718";
(my $grp) = $str =~ /CL\d+\_\_(\w+)\_\_/;
print "grp = $grp\n";

exit;

This returns...

grp = RTL_Dfdsg4__regression__df_umc_nbio_hubs_gfx

I want...

grp = RTL_Dfdsg4

As you can see, I know something about where the first set of double underscores exists (after the CL\d+). But for some reason, the regexp reads past the next occurrence of the double underscores until it hits the last set.

perl regex
2 Answers

You need to use the non-greedy quantifier, ?.

(my $grp) = $str =~ /CL\d+__(\w+?)__/;

I removed the unnecessary backslashes from before the underscores.

Andy Lester0

Note that using the non-greedy modifier is fragile and can easily work differently than intended. This is the robust alternative:

my ( $grp ) = $str =~ /
   CL \d+
   __ 
   ( [^\W_]+ (?: _ [^\W_]+ )* )     # `[^\W_]` is `\w` minus `_`
   __
/x;
ikegami0
Related