perl match single occurence pattern in string - regex

Question

I have a list of names and I want to look for names containing two given letters asigned using variables.
$one = "A";
$two = "O";
Please note that I want those letters to be present anywhere in the checked names, so that I can get outputs like this:
Jason
Damon
Amo
Noma
Boam
...
But each letter must only be present once per name, meaning that this wouldn't work.
Alamo
I've tried this bit of code but it doesn't work.
foreach my $name (#list) {
if ($name =~ /$one/) {
if ($name =~ /$two/) {
print $name;
}}
else {next}; }

Answer 1

How about this?
for my $name (#list) {
my $ones = () = $name =~ /$one/gi;
my $twos = () = $name =~ /$two/gi;
if ($ones == 1 && $twos == 1) {
print $name;
}
}

Answer 2

#!/usr/bin/env perl
#
# test.pl is the name of this script
use warnings;
use strict;
my %char = map {$_ => 1} grep {/[a-z]/} map {lc($_)} split //, join '', #ARGV;
my #chars = sort keys %char; # the different characters appearing in the command line arguments
while (my $line = <STDIN>)
{
grep {$_ <=> 1} map {scalar(() = $line =~ /$_/ig )} #chars
or print $line;
}
Now:
echo hello world | test.pl fw will print nothing (w occurs exactly once in hello world, but f does not)
echo hello world | test.pl hw will print a line consisting of hello world (both h and w occur exactly once).

Answer 3

One way to get it all into a single regex is to use an expression within the regex pattern to search for the other letter (a or o) based on which one was found first:
#!/usr/bin/env perl
use 5.010; use strict; use warnings;
while(<DATA>){
chomp;
say if m/^
[^ao]* # anything but a or o
([ao]) # an 'a' or 'o'
[^ao]* # anything but a or o
(??{($1 and lc($1) eq 'a') ? 'o' : 'a'}) # the other 'a' or 'o'
[^ao]* $/xi; # anything but a or o
}
__DATA__
Abdi
Chelal
Jucal
Husham
Gallio
Pileser
Tekoa
Kenaz
Raphah
See the perlre section on Extended Expressions for more info.

Answer 4

This is my solution. You don't make it clear whether there will always be just two single-character strings to match but I have assumed that there may be more
Unfortunately the classical way of counting occurrences of a character -- tr/// -- doesn't interpolate variables into its searchlist and doesn't have a case-independent modifier /i. But the pattern-match operator m// does, so that is what I have used
I thoroughly dislike the so-called goatse operator, but there isn't a neater way that I know of that allows you to count the number of times a global regex pattern matches
I could have used a grep for the inner loop, but I went for a regular for loop and a next with a label as I believe it's more readable this way
use strict;
use warnings;
use v5.10.1;
use autodie;
my #list = do {
open my $fh, '<', 'names.txt';
<$fh>;
};
chomp #list;
my ($one, $two) = qw/ A O /;
NAME:
for my $name ( #list ) {
for ( $one, $two) {
my $count = () = $name =~ /$_/gi;
next NAME unless $count == 1;
}
say $name;
}
output
Gallio
Tekoa
Achbor
Clopas
This is the input that I used
Abdi
Chelal
Jucal
Husham
Gallio
Pileser
Tekoa
Kenaz
Raphah
Tiras
Jehudi
Bildad
Shemidah
Meshillemoth
Tabeel
Achbor
Jesus
Osee
Elnaam
Rephah
Asaiah
Er
Clopas
Penuel
Shema
Marsena
Jaare
Joseph
Shamariah
Levi
Aphses