Re^3: What's so bad about &function(...)?

If you don't specify arguments, it passes any existing @_.

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

yes_ampersand("yes\n");
no_ampersand("no\n");

sub foo {
    print $_[0];
}

sub yes_ampersand {
    print $_[0];
    &foo;   
}

sub no_ampersand {
    print $_[0];
    foo;    
}
[download]

Here's the output:

yes
yes
no
Use of uninitialized value in print at ampersands.plx line 9.
[download]

--
Marvin Humphrey
Rectangular Research ― http://www.rectangular.com

Comment on Re^3: What's so bad about &function(...)? Select or Download Code

Replies are listed 'Best First'.
Re^4: What's so bad about &function(...)? by japhy (Canon) on Dec 07, 2005 at 20:31 UTC
I made it clear in my original node that I'm not referring to a bare `&foo`, but rather `&foo(...)`. Jeff `japhy` Pinyan, P.L., P.M., P.O.D, X.S.: Perl, regex, and `perl` hacker How can we ever be the sold short or the cheated, we who for every service have long ago been overpaid? ~~ Meister Eckhart	[reply] [d/l] [select]
Re^5: What's so bad about &function(...)? by creamygoodness (Curate) on Dec 07, 2005 at 22:02 UTC
I should have elaborated. Here's "what's so bad about &function(...)": The ampersand has a very subtle quirk which arises when it is used without parentheses, as illustrated in my previous post. The directive "don't use the ampersand" is easy to understand and follow. If you get "bareword" errors, they generally cause catastrophic failure and are thus straightforward to troubleshoot, even for beginners. The use of parentheses to group arguments to a subroutine is not enforced by Perl. Usage is all over the place. The directive "go ahead use the ampersand, but only with parentheses" is more difficult for a beginner to adhere to, because in the vast majority of cases there is no penalty for omitting the parentheses. When they do finally hit a problem because they forgot the parens, it's not going to be a nice easy bareword crash -- it's going to be a silent killer. IMO, it's not reasonable to argue that there's no problem with `&function(...)` while punting on `&function`, and since there's a problem with `&function`, there's a problem with `&function(...)`. Feel free to disagree. :) -- Marvin Humphrey Rectangular Research ― http://www.rectangular.com	[reply] [d/l] [select]
Re^6: What's so bad about &function(...)? by Perl Mouse (Chaplain) on Dec 08, 2005 at 17:18 UTC
That sounds to me as very illogical. You are afraid that, for the case of a function taking optional arguments, a beginner who uses & in front of the function name, might forget to use parenthesis? (Note that if the function requires arguments, omitting the parenthesis in combination of using & is a syntax error, and if the function doesn't use arguments, passing in some extras isn't going to hurt) Sounds too rare of a problem to get all dogmatic about. `Perl --((8:>*`	[reply]


Perl Monk, Perl Meditation
	PerlMonks