camelize/decamelize in Perl without String::CamelCase.

camelcase.pl

#!/usr/bin/env perl
use strict;
use warnings;
use utf8;

use Test::More;


sub camelize {
    my ($s) = @_;
    $s =~ s{(\w+)}{
        ($a = lc $1) =~ s<(^[a-z]|_[a-z])><
            ($b = uc $1) =~ s/^_//;
            $b;
        >eg;
        $a;
    }eg;
    $s;
}

sub decamelize {
    my ($s) = @_;
    $s =~ s{(\w+)}{
        ($a = $1) =~ s<(^[A-Z]|(?![a-z])[A-Z])><
            "_" . lc $1
        >eg;
        substr $a, 1;
    }eg;
    $s;
}




is camelize('snake_case'), 'SnakeCase';
is decamelize('CamelCase'), 'camel_case';
is decamelize(camelize('snake_case')), 'snake_case';
is camelize(decamelize('CamelCase')), 'CamelCase';

done_testing
__END__

=head1 NAME

    perl.pl - NO DESCRIPTION YET.


=head1 SYNOPSIS


=head1 OPTIONS


=head1 AUTHOR

tyru <tyru.exe@gmail.com>

I think there might be an off-by-one error on line 27, where the substring index should start at 0.

I used this snippet for converting a grammar wth mixed case parser tokens to ANTLR4 tokens (only lowercase allowed with underscores and digits) so I added somes tests:

is decamelize('lexer'), 'lexer';
is decamelize('LexerIATA'), 'lexer_iata';

For that I needed to change the decamelize function to

sub decamelize {
my ($s) = @;
$s =~ s{([a-zA-Z][a-zA-Z0-9]*([a-z0-9]|\b))}{
my $replaced = 0;
($a = $1) =~ s<(^[A-Z]+|(?![a-z])[A-Z]+)><
$replaced = 1;
"_" . lc $1;
>eg;
($replaced ? substr $a, 1 : $a);
}eg;
$s;
}