Need help in understanding perl tr command with /d

Question

I came across the following Perl example on the web.

#!/usr/bin/perl 

$string = 'the cat sat on the mat.';
$string =~ tr/a-z/b/d;

print "$string\n";

result:

b b   b.

Can someone please explain how ?

Answer 1

/d denotes delete.

It's quite unusual to do a tr like that because it's confusing.

tr/a-z//d

would delete all 'a-z' characters.

tr/a-z/b/ 

would transliterate all a-z characters to b.

What's happening here though is - because your transliteration doesn't map an equal number of character on each side - anything that doesn't map is deleted.

So what you're effectively doing is:

tr/b-z//d;
tr/a/b/;

E.g. transliterating all the as to bs and then deleting anything else (except spaces and dots).

To illustrate:

use strict;
use warnings;
my $string = 'the cat sat on the mat.';
$string =~ tr/the/xyz/d;

print "$string\n";

Warns:

Useless use of /d modifier in transliteration operator at line 5.

and prints:

xyz cax sax on xyz max.

If you change that to:

#!/usr/bin/perl 
use strict;
use warnings;
my $string = 'the cat sat on the mat.';
$string =~ tr/the/xy/d;

print "$string\n";

You get instead:

xy cax sax on xy max.

And thus: t -> x and h -> y. e just gets deleted.

Answer 2

d is used to delete found but not replaced characters.

To remove the characters which are not in the matching list can be done by appending d to the end of the tr operator.

#!/usr/bin/perl 

my $string = 'my name is serenesat';
$string =~ tr/a-z/bcd/d;
print "$string\n";

Prints:

 b  b

Not matching characters in the string is removed, and only the matching character replaced (a replaced with b).