You can let the regular expression do the work of finding where the word is and printing it's positions. Doing a global match in an empty while loop, it uses the @- array which records the start positions of "last match", see also @+ for end positions. It also uses a regular expression code block (?{...}) to print the position once the match has been found.

use strict;
use warnings;

use re q{eval};

my $word = q{test};
my $rxWord = qr{\b$word\b};

# This is the one the does the work.
#
my $rxWordPos = qr{\b($word)\b(?{print $-[0], q{ }})};

while (<DATA>)
{
    next unless m{$rxWord};
    print qq{Match found on line $., column };

    while (m{$rxWordPos}g) {;}

    print qq{\n};
}

__END__
This is a test from tester okay
nothing
message test center test
test in proress ... test one test two
a tester in this line
[download]

Here's the output

Match found on line 1, column 10 
Match found on line 3, column 8 20 
Match found on line 4, column 0 20 29
[download]

I hope this is of use.

Cheers,

JohnGG

Update: Simpler version eliminating the regular expression code block.

use strict;
use warnings;

my $word = q{test};
my $rxWord = qr{\b($word)\b};

while (<DATA>)
{
    next unless m{$rxWord};
    print qq{Match found on line $., column };
    while (m{$rxWord}g)
    {
        print qq{$-[0] };
    }
    print qq{\n};
}

__END__
This is a test from tester okay
nothing
message test center test
test in proress ... test one test two
a tester in this line
[download]