TL;DR

A pull request for String::Util.

I recently posted about String::Util, a nice idea for collecting common functions that were partially made available through the builtin module.

Well, namespace. Well, way to access functions built in the interpreter. Whatever.

Thanks to Mutant Rob for reminding me about builtin::trim by the way!

As I was skimming through the docs of the module, I wondered how startswith, endswith and contains were implemented. Surely there was index involved, right? Sure it was:

sub contains {
	my ($str, $substr) = @_;

	if (!defined($str)) {
		return undef;
	}

	if (!$substr) {
		$substr = $str;
		$str    = $_;
	}

	my $ret = index($str, $substr, 0) != -1;

	return $ret;
}

Uh-oh.

I still have multiple scars from this particular bite, so I have a sort of reflex when I see this:

if (!$substr) {

As everything with a soul, Perl has its own ways of being useful which can sometimes get in the way. Like, for example, not having a proper boolean underlying type, and resorting to figuring it out in a do what I mean way:

Let’s recap what Perl thinks is false:

• Empty lists (also when expanding an array/hash).
• Empty strings.
• undef (this was easy)
• 0 (as a number)
• '0' (the string)

I was not there, but I guess that the fift one got some suspicious looks back in the time, much like the Parallel postulate… yes… this makes sense… but it’s a bit close to the border….

I take it as a consequence of the seamless switch between integers/numbers/strings that makes Perl so friendly; sometimes, though, it might mess things up, like in this case:

say 'yes' if endswith('total is 0', '0');

or, I daresay, the following case too:

say 'yes' if endswith('total is whatever', '');

So… instead of just writing about it here, it was time for a pull request! I hope it gets considered, evaluated, and eventually accepted (as-is or with modifications, I might have deviated from the style in the rest of the module…)

Until then… Stay safe!