/([0-8])[a-z](??{ $1 + 1 })/
[download]

See perlre for details.

Updated: Simplified, don't use named capture groups.

Thats brilliant.

Now I have something to work on.

MANY THANKS

Actually, the better question we haven't asked is: "Why?". Maybe a regex is not the best way to solve the problem, whatever it is. If we know why we may be able to offer a better solution.

I try to combine a coding in a two dimensional room (x,y). Explaining all the details would not help here. If a simple modification in the search string would be possible, I guess the problem could be solved quite elegant (like the e modifier). But I am not aware if this is possible. So i generated a simplified question. How to write a function, solving this is clear for me. But as usual, I try to solve it elegant and also learn something new.

Thanks ALL for your help and time !!

I would search for a number followed by a char followed by another number using a regexp, and then filter in perl:

#!/usr/bin/perl

use strict;
use warnings;
use feature 'say';

my $text='abc 1foo 1x1 ee2 3B4 n 5 6t7 q9q 5+6';
while ($text=~/(\d)([[:alpha:]])(\d)/g) {
        if ($1 + 1 == $3) {
                say "$1$2$3 is wanted because $1 + 1 = $3";
        } else {
                say "$1$2$3 matches, but is not wanted ($1 + 1 != $3)"
+;
        }
}
[download]

Output:

1x1 matches, but is not wanted (1 + 1 != 1)
3B4 is wanted because 3 + 1 = 4
6t7 is wanted because 6 + 1 = 7
[download]

Alexander

Should that include double digits? 9w10?

Should it match if the digit/number before the letter is preceded by a digit? 13s4

Should it match if the digit/number after the letter is followed by a digit? 3s44

one digit only.

so 9w10 is not valid, as 1 is not 9+1

13s4 => match would find "3s4"

3s44 => match would find "3s4"

For the 9 cases, I would simply enumerate them:

/(
 (0[a-z]1)
|(1[a-z]2)
|(2[a-z]3)
|(3[a-z]4)
|(4[a-z]5)
|(5[a-z]6)
|(6[a-z]7)
|(7[a-z]8)
|(8[a-z]9)
)/x
[download]

For further experimentation, I would write a test rig like the following to easily play with regular expressions and data and see how things work:

#!perl
use 5.020;
use Test2::V0 '-no_srand' => 1;

my @cases = (
    # case, expected result, todo text
    ['13s4', '3s4' ],
    ['3s44', '3s4' ],
    ['0a1', '0a1' ],
    ['0a2', undef ],
);

for my $case (@cases) {
    my( $line, $expected, $todo ) = @$case;
    my ($result) = $line =~ /(
                           (0[a-z]1)
                          |(1[a-z]2)
                          |(2[a-z]3)
                          |(3[a-z]4)
                          |(4[a-z]5)
                          |(5[a-z]6)
                          |(6[a-z]7)
                          |(7[a-z]8)
                          |(8[a-z]9)
                          )/x;

    my $t; if ($todo){ $t = todo( $todo ) };
    is( $result, $expected, $line );
}

done_testing;
[download]

Not elegant, but fits the spec:

/0.1|1.2|2.3|3.4|4.5|5.6|6.7|7.8|8.9|9.0/
[download]

Optimising for fewest key strokes only makes sense transmitting to Pluto or beyond

Searching for a number, followed by a char, followed by the same number is easy:

m/(\d).\1/

That also matches 111:

$ perl -E 'say "match" if "111"=~/(\d).\1/'
match
$
[download]

Alexander