pragmas.pod6

=begin pod :tag<index>

=TITLE Pragmas

=SUBTITLE Special modules that define certain aspects of the behavior of the code

In Perl 6, B<pragmas> are directive used to either identify a specific
version of Perl 6 to be used or to modify the compiler's normal behavior
in some way. The C<use> keyword enables a pragma (similar to how you can
C<use> a module). To disable a pragma, use the C<no> keyword:

=begin code :solo
use v6.c;   # use 6.c language version
no worries; # don't issue compile time warnings
=end code

Following is a list of pragmas with a short description of each
pragma's purpose or a link to more details about its use.  (Note:
Pragmas marked "[NYI]" are not yet implemented, and those marked "[TBD]"
are to be defined later.)

X<|v6.x, pragma>
=head2 v6.x

This pragma states the version of the compiler that is going to be used,
and turns on its features if they are optional.

=for code :solo
use v6;   # Load latest supported version (non-PREVIEW).
          # Also, useful for producing better errors when accidentally
          # executing the program with `perl` instead of `perl6`

=for code :solo
use v6.c;         # Use the "Christmas" version of Perl 6

=for code :solo
use v6.d;         # Use the "Diwali" version of Perl 6

From 2018.11, which implemented 6.d, this pragma does not do anything.

=for code :solo
use v6.d.PREVIEW; # On 6.d-capable compilers, enables 6.d features,
                  # otherwise enables the available experimental
                  # preview features for 6.d language

Since these pragmas turn on the compiler version, they should be the
first statement in the file (preceding comments and Pod are fine).

X<|MONKEY-GUTS, pragma>
=head2 MONKEY-GUTS

This pragma is not currently part of any Perl 6 specification, but is present
in Rakudo as a synonym to C<use nqp> (see below).

X<|MONKEY-SEE-NO-EVAL, pragma>
=head2 MONKEY-SEE-NO-EVAL

L<EVAL|/routine/EVAL>

X<|MONKEY-TYPING, pragma>
=head2 MONKEY-TYPING

L<augment|/syntax/augment>

X<|MONKEY, pragma>
=head2 MONKEY

C<use MONKEY;>

Turns on all available C<MONKEY>
pragmas, currently the three above; thus, it would be equivalent to

    use MONKEY-TYPING;
    use MONKEY-SEE-NO-EVAL;
    use MONKEY-GUTS;

X<dynamic-scope, pragma>
=head2 dynamic-scope

Applies the L<is dynamic|/type/Variable#trait_is_dynamic> trait to variables
in the pragma's lexical scope. The effect can be restricted to a subset of
variables by listing their names as arguments. By default applies to I<all>
variables.

=begin code :allow<C>
# Apply C<is dynamic> only to $x, but not to $y
use dynamic-scope <$x>;

sub poke {
    say $CALLER::x;
    say $CALLER::y;
}

my $x = 23;
my $y = 34;
poke;

# OUTPUT:
# 23
# Cannot access '$y' through CALLER, because it is not declared as dynamic
=end code

This pragma is not currently part of any Perl 6 specification and was added
in Rakudo 2019.03.

X<|experimental, pragma>
=head2 experimental

Allows use of L<experimental features|/language/experimental>

X<|fatal, pragma>
=head2 fatal

A lexical pragma that makes L<Failures|/type/Failure> returned from routines
fatal. For example, prefix C<+> on a L<Str|/type/Str> coerces it to
L<Numeric|/type/Numeric>, but will return a L<Failure|/type/Failure> if the
string contains non-numeric characters. Saving that L<Failure|/type/Failure> in
a variable prevents it from being sunk, and so the first code block below
reaches the C<say $x.^name;> line and prints C<Failure> in output.

In the second block, the C<use fatal> pragma is enabled, so the C<say> line is
never reached because the L<Exception|/type/Exception> contained in the L<Failure|/type/Failure> returned from
prefix C<+> gets thrown and the C<CATCH> block gets run, printing the
C<Caught...> line. Note that both blocks are the same program and C<use fatal>
only affects the lexical block it was used in:

    {
        my $x = +"a";
        say $x.^name;
        CATCH { default { say "Caught {.^name}" } }
    } # OUTPUT: «Failure␤»

    {
        use fatal;
        my $x = +"a";
        say $x.^name;
        CATCH { default { say "Caught {.^name}" } }
    } # OUTPUT: «Caught X::Str::Numeric␤»

Inside L«C<try> blocks|/language/exceptions#index-entry-try_blocks-try», the
C<fatal> pragma is enabled by default, and you can I<disable> it with C<no
fatal>:

    try {
        my $x = +"a";
        say $x.^name;
        CATCH { default { say "Caught {.^name}" } }
    } # OUTPUT: «Caught X::Str::Numeric␤»

    try {
        no fatal;
        my $x = +"a";
        say $x.^name;
        CATCH { default { say "Caught {.^name}" } }
    } # OUTPUT: «Failure␤»

X<|internals, pragma>
=head2 internals

[NYI]

X<|invocant, pragma>
=head2 invocant

[NYI]

X<|isms, pragma>
=head2 isms

C<[2018.09 and later]>

Allow for some other language constructs that were deemed to be a trap that
warranted a warning and/or an error in normal Perl 6 programming.  Currently,
C<Perl5> and C<C++> are allowed.

=begin code :skip-test<compile-time error>
sub abs() { say "foo" }
abs;
# Unsupported use of bare "abs"; in Perl 6 please use .abs if you meant
# to call it as a method on $_, or use an explicit invocant or argument,
# or use &abs to refer to the function as a noun
=end code

In this case, providing an C<abs> sub that doesn't take any arguments, did
not make the compilation error go away.

    use isms <Perl5>;
    sub abs() { say "foo" }
    abs;   # foo

With this, the compiler will allow the offending Perl 5 construct, allowing
the code to actually be executed.

If you do not specify any language, all known language constructs are allowed.

    use isms;   # allow for Perl5 and C++ isms

X<|lib, pragma>
=head2 lib

This pragma adds subdirectories to the library search
path so that the interpreter can
L<find the modules|/language/modules#Finding_modules>.

=for code :solo
use lib <lib /opt/lib /usr/local/lib>;

This will search the directories passed in a list. Please check
L<the modules documentation|/language/modules#use> for more examples.

X<|newline, pragma>
=head2 newline

Set the value of the L<$?NL|/language/variables#Compile-time_variables> constant
in the scope it is called.  Possible values are C<:lf> (which is the default,
indicating Line Feed), C<:crlf> (indicating Carriage Return, Line Feed) and
C<:cr> (indicating Carriage Return).

X<|nqp, pragma>
=head2 nqp

Use at your own risk.

This is a Rakudo-specific pragma. With it, Rakudo provides access to the
L<nqp opcodes|https://github.com/perl6/nqp/blob/master/docs/ops.markdown>
in a top level namespace:

    use nqp;
    nqp::say("hello world");

This uses the underlying nqp C<say> opcode instead of the Perl 6 routine. This
pragma may make your code rely on a particular version of nqp, and since
that code is not part of the Perl 6 specification, it's not guaranteed to
be stable. You may find a large number of usages in the Rakudo core,
which are used to make the core functionality as fast as possible.
Future optimizations in the code generation of Rakudo may obsolete these
usages.

X<|parameters, pragma>
=head2 parameters

[NYI]

X<|precompilation, pragma>
=head2 precompilation

The default allows precompilation of source code, specifically if used in a
module.  If for whatever reason you do not want the code (of your module) to
be precompiled, you can use C<no precompilation>.  This will prevent the
entire compilation unit (usually a file) from being precompiled.

X<|soft, pragma>
=head2 soft

L<Re-dispatching|/language/functions#Re-dispatching>, L<inlining|/language/functions#index-entry-use_soft_(pragma)>

X<|strict, pragma>
=head2 strict

C<strict> is the default behavior, and requires that you declare variables
before using them. You can relax this restriction with C<no>.

=for code
no strict; $x = 42; # OK

X<|trace, pragma>
=head2 trace

When C<use trace> is activated, any line of code executing will be written to
STDERR.  You can use C<no trace> to switch off the feature, so this only happens
for certain sections of code.

X<|v6, pragma>
=head2 v6

L<Writing Tests|/language/testing#Writing_tests>

X<|variables, pragma>
=head2 variables

L<Defined Variables Pragma|/language/variables#Default_defined_variables_pragma>

X<|worries, pragma>
=head2 worries

Lexically controls whether compile-time warnings generated by the
compiler get shown. Enabled by default.

    =begin code :lang<text>
    $ perl6 -e 'say :foo<>.Pair'
    Potential difficulties:
      Pair with <> really means an empty list, not null string; use :foo('') to represent the null string,
        or :foo() to represent the empty list more accurately
      at -e:1
      ------> say :foo<>⏏.Pair
    foo => Nil

    $ perl6 -e 'no worries; say :foo<>.Pair'
    foo => Nil
    =end code

=end pod

# vim: expandtab softtabstop=4 shiftwidth=4 ft=perl6