Oh Shit, Perl

info

You can access to this page from ohshitperl.com, ohshit.foo/perl or ohshit.bar/perl.

Operators

Perl has a lot of operators, and some of them are a bit weird. Here are some of them:

==

These are all true:

1 == 1
1 == 1.0
1 == "1"
1 == "1.0"
1 == "a1.0aa\n"
"foo" == "bar"; # because why not?

This happens because == is for numbers only, and strings without any numbers in them would be casted to 0, and so in Perl's world "foo" == "bar". For strings eq should be used instead.

+

+ is for numbers which is why string + int returns int, even if the string isn't completely made of numbers:

print "123aaa" + 2; # this will return 125

This is used to convert strings to int as well.

q and qq

You can change ' and " using q and qq operator:

my $var = "Hello World from \"$_\"";
my $easy_var = qq\Hello World from "$_"\;
my $weird_syntax = qq qHello World from \qq;

<=>

The <=> (which could be a method, like CompareTo in C#) returns -1, 0, or 1 based on the comparison result. It's used to compare numbers, and it's used in sort method.

my @sorted = sort { $a <=> $b } @unsorted;

=> (Fat Comma) and ,=> (Winking Fat Comma)

=>, called fat command has similar functionality to , (comma), with additional functionality of automatically quoting the left side. It's a syntactic sugar used for hashes:

my %hash = (a => 1, b => 2, c => 3);
my @array = (a => b => c => d); # there's nothing special about =>, it's just a comma with left side quoting

and if you don't want the left side to be quoted, you can use ,=> (winking fat comma).

=~

=~ doesn't have a consistent behavior. It returns a list of matches when used with m/.../ to match the left side with the regex on the right side, but it's a binding operator when used with s/.../.../ to replace the left side with the right side.

Also it can be used inside a while loop to iterate over matches:

my $str = "hello world this is a test for regex in perl";
while ($str =~ /(\w+)/g) {
  print "Word is $1, ends at position ", pos $str, "\n";
}

Precedence

• Operator Precedence and Associativity
• Why do Perl's logical operators &&, ||, ! take precedence over and, or, and not?

Secret Operators

perlsecret - Perl secret operators and constants

Arrays, Hashes, and Lists

• Lists aren't data structures, they're just a way to group values together. They would be stored in memory either as an array or a hash when assigned to a variable.

• You can have a scalar, an array, and a hash with the same name.

  my $var = "scalar";
  my @var = ("array");
  my %var = ("hash" => "value");
  print "\$var = $var\n\@var = @var\n\%var = %var\n";

• To get values out of arrays and hashes, you need to indicate the expected output type with $ (scalar), @ (array), or % (hash) sigils prefixes, and [] for arrays or {} for hashes:

  my @array = ("a", "b", "c", "d");
  my %hash = ("a" => 1, "b" => 2, "c" => 3);
  print "array[0] = $array[0]\n";
  print "hash{a} = $hash{a}\n";
  print qq/hash out of array = @hash{"a", "b"}\n/;
  print qq/array out of hash = %array{"a", "b"}\n/;

  This can get very confusing if you reuse variable names for different types:

  my @var = ("a", "b", "c", "d");
  my %var = ("a" => 1, "b" => 2, "c" => 3);
  my $var = "scalar";
  
  print "\$var{a} = $var{a}\n"; # prints 1, returns scalar from hash
  print "\$var[0] = $var[0]\n"; # prints a, returns scalar from array
  print qq/\@var{"a", "b"} = @var{a, b}\n/; # prints 1 2, returns array (because of @) from hash (because of {})
  print qq/\%var[0,1] = %var{"a", "b"}\n/; # prints 0 => a, 1 => b, returns hash (because of %) from array (because of [])

• List of lists is not a 2D array, it'll be flattened into a 1D array:

  my @array = (1, 2, 3, (4, 5, 6), 7, 8, 9);
  print "@array\n"; # prints 1 2 3 4 5 6 7 8 9

• To create a 2D array, you need to use references:

  my @array = (1, 2, 3, [4, 5, 6], 7, 8, 9);
  print "@array\n"; # prints 1 2 3 ARRAY(0x1f9f8a0) 7 8 9
  print "@{$array[3]}\n"; # prints 4 5 6

• You can use lists to set values of multiple variables at once:

  my ($a, $b, $c) = (1, 2, 3);
  print "$a $b $c\n"; # prints 1 2 3

• When initializing multiple arrays using a list, the leftmost array is greedy and takes all the values:

  my (@array1, $array2) = (1, 2, 3, (4, 5, 6), 7, 8, 9);
  print "@array1\n"; # prints 1 2 3 4 5 6 7 8 9
  print "$array2\n"; # prints nothing

Flow Control

Next, Last, Redo, and Continue

last is the same as break in C# and next is the same as continue in C#. Additionally, redo is used to restart the loop without reevaluating the condition and continue is like finally in exception handling, for the next statement.

Take the example from perldoc:

while (EXPR) {
    ### redo always comes here
    do_something;
} continue {
    ### next always comes here
    do_something_else;
    # then back the top to re-check EXPR
}
### last always comes here

For and Foreach

for and foreach are the same and can be used interchangeably.

Boolean

Perl doesn't have a boolean type, it uses falsey and truthy values instead. The falsey values are undef, 0, "", "0", and empty list, array, and hash. Everything else is truthy.

print undef ? "true" : "false"; # prints false
print 0 ? "true" : "false"; # prints false
print "" ? "true" : "false"; # prints false
print "0" ? "true" : "false"; # prints false
print "false" ? "true" : "false"; # prints true
print "0.0" ? "true" : "false"; # prints true
print () ? "true" : "false"; # prints false
my @array = ();
print @array ? "true" : "false"; # prints false
my %hash = ();
print %hash ? "true" : "false"; # prints false

Subroutines

• There's no proper way to define subroutine arguments and return. Everything that's passed to the subroutine is stored in @_ array, and you need to manually check the number of arguments and their types.

• You can call subroutines without parenthesis.

  sub test {
    print $_[0] . " " . $_[1] . "\n";
  }
  
  test "hello", "world";

• There's no proper way to define subroutine return types as well. See wantarray for more oh shit moments.

• Variables passed to the subroutine are passed by reference and if you modify values of the @_ array, original values would be modified as well. It's a good practice to always unpack @_:

  sub test1 {
    my $arg1 = $_[0];
  }
  
  sub test2 {
    my $arg1 = shift; # Because it's Perl, and @_ would be the unannounced argument for shift and pop: https://perldoc.perl.org/perlvar#@_
  }
  
  sub test3 {
    my ($arg1, $arg2) = @_;
  }

• You can expect a hash as the subroutine's input and create a hash from @_ on the subroutine. This approach has the benefit of copying the values, having named parameters, and being able to set default values for parameters.

  sub test {
    my %args = (
      first_arg = "default value",
      second_arg = undef,
      @_ # use @_ as a hash and overwrite the hash values if they exist on @_
    )
  }

  But this approach doesn't support references. test({ second_arg = "world" }); doesn't work because in this case $_[0] would be a hash reference, and would not be unpacked into the hash. To support this case, you need to dereference the hash reference:

  sub test {
    my %args = (
      first_arg = "default value",
      second_arg = undef,
      ref($_[0]) ? %{$_[0]} : @_
    )
  }

  But note that this approach is not perfect and is error prone. For example, test(first_arg=>"hello", { second_arg = "world" }); would not unpack second_arg because ref($_[0]) would be false and @_ would be used instead of dereferencing the hash reference.

  To keep it simple, just use @_ and unpack it.

• You can use the experimental signatures feature to define subroutine arguments.

• You can omit return <something>; and use <something> instead.

  sub test {
    return "hello";
  }
  
  sub test2 {
    "hello"
  }

• You can omit $_ in some operators and functions to write shorter code:

  print for @array;
  grep { m/<some regex>/ } @array;

  This can also be used for bash one-liners:

  perl -ne 'print if m/<some regex>/' <file>

wantarray

wantarray depends on the context of the function call, meaning what you're doing with the return value of the function. If you're not doing anything with the return value, it's undef. If you're assigning it to a scalar, it's false. If you're assigning it to a list, it's true.

This function creates a lot of confusion because there's no guarantee that the function is using wantarray in its implementation, and how it's using it. Therefore it's recommended to avoid using it.

Classes

• Other than scalar, array, and hash, there are no other data structures in Perl.
• Object oriented programming is not built into the language, and you need to use modules to implement it. An Object is Simply a Data Structure.
  • There's no reserved name for constructors, and you can use any name you want. new is just a convention.

More Resources

• Raku: A Language for Gremlins
  • YC discussion