Fun. Some edge cases got missed several times. Strings under 12 characters and a first word of over 12 characters were the most common, I think. Though what to do when the first word is over 12 characters is not clear from the problem statement. Strict interpretation of the problem statement would end up with an empty string for that case. More likely, the first 12 characters of the first word is appropriate.

But I think you can still do this and keep it pretty simple:

my( $first )= $message =~ /^\s*(.{1,12}\s|.{0,12})/;
[download]

Many ways to do this. For such a tiny, simple operation, I could envision a half-dozen tests in the UT suite off the top of my head. (:

Note that I don't collapse internal spaces, which is a nice touch in Transient's solution.

This will extract anything up to 12 characters (greedy) that is followed by whitespace. It does what you described, but needs work to be of more general use.

#! /usr/bin/perl

use strict;
use warnings;

my $string = 'this is a test message';
my ($chunk) = $string =~ /(.{1,12})\s/;

print $chunk, "\n";
[download]

(?:\s|$) works better than \s.

An other alternative is \b, which will chop after a word, but before punctuation.

Depends on what a word is. If a word is \w+, this variant might be of use: ((\s|\w){0,11})(\s|$)

#!/usr/bin/perl

my $max_str_len = 12;

my $string = "This is a test message";

$string =~ tr/ / /s;
$string =~ s/^\s*//;
$string =~ /(.{0,$max_str_len})(\s|$)/;

my $trunc_string = $1;

my @words = split /\s+/, $trunc_string;

print "Word: ", $_, "\n" foreach @words;
[download]

This code is also a bit more readable, in my opinion.

use warnings;
use strict;

use Text::Wrap qw(wrap);

my $phrase = 'This is a test message with lots of words in it. Tintina
+bulations!';

$Text::Wrap::columns = 13;
my $wrapped = wrap('','',$phrase);

print $wrapped;
[download]

If you decide to use Text::ExtractWords, note that there is an error in the documentation: Text::ExtractWords minwordlen and maxwordlen, as described by Bob9000 in his reply.

HTH,




Update: If anyone is using Text::ExtractWords for their project, you should know wfsp has more problems to report: Unpredictable output from Text::ExtractWords.

planetscape

Thanks for the link, i've had a look at the module though and i'm not sure it does what I need it to... i'm just installing it now and will have a go and see if I can get something to work.

Learning without thought is labor lost; thought without learning is perilous. - Confucius
WebChalkboard.com | For the love of art...

my $string = 'this is a test message';
my ($chunk) = $string =~ /(.{1,12})\b/;
[download]

$msg = "This is a test message";
@lines = $msg =~ /(?>\s*)(.{0,11}\S)(?!\S)/g;
[download]

No more than 12 characters: ^.{0,12}

Words: ^(\w+(\s|$))*

Both must be true:

HOW?

Is there maybe some way to do this with a perl6 rule, or Parse::Recdescent?