Loop controls: next, last, continue, break

In Perl there are 3 loop control keywords. The two commonly used are next and last and there is a third which is rarely used called redo.

In most of the other languages the respective keywords are continue and break.

next of Perl is the same as the continue in other languages and the last if Perl is the same as the break of other languages.

redo probably does not have its counterpart.

next will jump to the condition and if it is true Perl will execute the next iteration.

last will quite the loop.

Take a look at this example. This script will count the number of non-empty rows in a file until it encounters __END__ or the end of file.

examples/loop_control.pl

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

my $filename = shift or die "Usage: $0 FILENAME\n";
main($filename);

sub main {
    my ($file) = @_;

    open my $fh, '<', $file or die;

    my $counter = 0;
    my $empty = 0;
    my $nonempty = 0;
    while (my $line = <$fh>) {
        $counter++;
        chomp $line;
        if ($line eq "__END__") {
            last;     # end processing file
        }

        if ($line eq "") {
            $empty++;
            next;     # don't process empty rows, go to next line
        }

        print "Process '$line'\n";
        $nonempty++;
                                        ### next jumps here ###
    }
                                        ### last jumps here ###
    print "Number of empty rows $empty and non empty rows: $nonempty out of a total of $counter\n"
}

Given this input file:

examples/loop_control.txt

First line

Line after empty line

Another text line
__END__
Some text here
that won't be printed

the output will be this:

Process 'First line'
Process 'Line after empty line'
Process 'Another text line'
Number of non empty rows: 3 out of a total of 6

next

The next keyword is usually used when we during our iteration we have an condition inside the loop and if that condition is matched (or faild to match) then we would like to end the current iteration and go to the next one. For example if we are looking for a string that starts with "DNA:" and then has a DNA in it with some complex rule, we might first check if the current row starts with "DNA:". If it does not then there is no point in further investigating the current string and we can go to the next iteration.

last

On the other hand if we found the DNA we were looking for we might stop the whole search if we are only interested in the first such DNA. In this case calling last will break us out from the current loop saving a lot of time by not iterating over the rest of the values.

In this example we were looking for a DNA with a 3-character repetition.

examples/find_dna.pl

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

my $file = shift or die "Usage: $0 FILENAME\n";

open my $fh, '<', $file or die;

my $interesting_dna;
while (my $line = <$fh>) {
    chomp $line;

    my ($dna) = $line =~ /^DNA:(.*)/;
    if (not $dna) {
        next; #skip rows without DNA: at the beginning
    }
    print "Checking $dna\n";

    if ($dna =~ /([ACTG]{3}).*\1/) {
        $interesting_dna = $dna;
        last; # got to the end of the loop, skip rest of the file
    }
}

if ($interesting_dna) {
    print "First interesting DNA: $interesting_dna\n";
}

This is our very limited input file:

examples/find_dna.txt

human:
DNA:ACTGGTATTA

mouse: 
DNA:ACTGTAACTG

elephant:
DNA:AACCTTGGAA
 

The resulting output looks like this:

Checking DNA:ACTGGTATTA
Checking DNA:ACTGTAACTG
First interesting DNA: ACTGTAACTG

It skipped all the lines that did not start with DNA: and once it found a match (in the 2nd DNA) it stopped looking skipping all the other lines.

redo

Calling redo will execute the current iteration again without checking the condition. I don't recall using it so let's not get into it now.

Comments

I was expecting something on the "continue" keyword, which is actually useful in some cases. There's a good example on the "eof" doc page (https://perldoc.pl/function....

I find "continue" useful when parsing files, and I need to deal with line numbers. Also, when there's some work that needs to be done between iterations, but "next" or "last" is used a lot, "continue" can be more obvious and maintainable.

"redo" is also useful sometimes, when external state changes, or IO is needed.

When I wrote this I was more focused on the idea that people new to perl would want to use the "continue" keyword they know from other languages and I wanted to direct them to the "next" keyword.