TITLE

Synopsis_3 - Perl 6 Operators

AUTHOR

Luke Palmer <luke@luqui.org>

VERSION

  Maintainer: Larry Wall <larry@wall.org>
  Date: 8 Mar 2004
  Last Modified: 26 Oct 2007
  Number: 3
  Version: 124

Overview

For a summary of the changes from Perl 5, see "Changes to Perl 5 operators".

Operator precedence

Not counting terms and terminators, Perl 6 has 23 operator precedence levels (same as Perl 5, but differently arranged). Here we list the levels from "tightest" to "loosest", along with a few examples of each level:

    Level               Examples
    =====               ========
    Terms               42 3.14 "eek" qq["foo"] $x :!verbose @$array
    Method postfix      .meth .+ .? .* .() .[] .{} .<> .«» .:: .= .^ .:
    Autoincrement       ++ --
    Exponentiation      **
    Symbolic unary      ! + - ~ ? | +^ ~^ ?^ \ ^ =
    Multiplicative      * / % +& +< +> ~& ~< ~> ?& div mod
    Additive            + - +| +^ ~| ~^ ?| ?^
    Replication         x xx
    Concatenation       ~
    Junctive and (all)  &
    Junctive or (any)   | ^
    Named unary         rand sleep abs
    Nonchaining infix   but does <=> leg cmp .. ..^ ^.. ^..^
    Chaining infix      != == < <= > >= eq ne lt le gt ge ~~ === eqv !eqv
    Tight and           &&
    Tight or            || ^^ // min max
    Conditional         ?? !! ff fff
    Item assignment     = := ::= => += -= **= xx= .=
    Loose unary         true not
    Comma operator      ,
    List infix          Z minmax X X~X X*X XeqvX
    List prefix         = : print push say die map substr ... [+] [*] any $ @
    Loose and           and andthen
    Loose or            or xor orelse
    Terminator          ; <==, ==>, <<==, ==>>, {...}, modifiers, extra ), ], }

If you don't see your favorite operator there, the following sections cover all the operators in precedence order. Basic operator descriptions are here; special topics are covered afterwards.

Term precedence

This isn't really a precedence level, but it's in here because no operator can have tighter precedence than a term. See S02 for longer descriptions of various terms.

  • Int literal

        42
  • Num literal

        3.14
  • Non-interpolating Str literal

        '$100'
  • Interpolating Str literal

        "Answer = $answer\n"
  • Generalized Str literal

        q["$100"]
        qq["$answer"]
  • Heredoc

        qq:to/END/
            Dear $recipient:
            Thanks!
            Sincerely,
            $me
            END
  • Array composer

        [1,2,3]

    Provides list context inside.

  • Hash composer

        { a => 42 }

    Inside must be a list of pairs, otherwise you must use hash() or %() instead.

  • Closure

        { ... }

    When found where a statement is expected, executes immediately. Othewise always defers evaluation of the inside scope.

  • Capture composer

        \(@a,$b,%c)

    An abstraction representing an argument list that doesn't yet know its context.

  • Sigiled variables

        $x
        @y
        %z
        $^a
        $?FILE
        @@slice
        &func
        &div:(Int, Int --> Int)
  • Sigils as contextualizer functions

        $()
        @()
        %()
        &()
        @@()
  • Regexes in quote-like notation

        /abc/
        rx:i[abc]
        s/foo/bar/
  • Transliterations

        tr/a..z/A..Z/;

    Note ranges use .. rather than -.

  • Type names

        Num
        ::Some::Package
  • Circumfixed subexpressions

        (1+2)

    Circumfixed items are treated like a term on the outside.

  • Function call

        a(1)
  • Pair composers

        :by(2)
        :!verbose
  • Signature literal

        :(Dog $self:)
  • Method call with implicit invocant

        .meth       # call on $_
        .=meth      # modify $_

    Note that this may occur only where a term is expected. Where a postfix is expected, it is a postfix. If only an infix is expected (that is, after a term with intervening whitespace), .meth is a syntax error. (The .=meth form is allowed there only because there is a special .= infix assignment operator that is equivalent in semantics to the method call form but that allows whitespace between the = and the method name.)

  • Listop (leftward)

        4,3, sort 2,1       # 4,3,1,2

    As in Perl 5, a list operator looks like a term to the expression on its left, so it binds tighter than comma on the left but looser than comma on the right--see List operator precedence below.

Method postfix precedence

All method postfixes start with a dot, though the dot is optional for subscripts. Since these are the tightest standard operator, you can often think of a series of method calls as a single term that merely expresses a complicated name.

See S12 for more discussion of single dispatch method calls.

  • Standard single-dispatch method calls

        $obj.meth
  • Variants of standard single-dispatch method call

        $obj.+meth
        $obj.?meth
        $obj.*meth

    In addition to the ordinary . method invocation, there are variants .*, .?, and .+ to control how multiple related methods of the same name are handled.

  • Class-qualified method call

        $obj.::Class::meth
        $obj.Class::meth    # same thing, assuming Class is predeclared

    As in Perl 5, tells the dispatcher which class to start searching from, not the exact method to call.

  • Mutating method call

        $obj.=meth

    The .= operator does inplace modification of the object on the left.

  • Meta-method call

        $obj.^meth

    The .^ operator calls a class metamethod; foo.^bar is short for foo.HOW.bar.

  • Method-like postcircumfixes

        $routine.()
        $array.[]
        $hash.{}
        $hash.<>
        $hash.«»

    The dotless forms of these have exactly the same precedences.

  • Dotted form of any other postfix operator

        $x.++         # postfix:<++>($x)
  • Dotted postfix form of any other prefix operator

        $x.:<++>       # prefix:<++>($x)
  • There is specifically no infix:<.> operator, so

        $foo . $bar

    will always result in a compile-time error indicating the user should use infix:<~> instead. This is to catch an error likely to be made by Perl 5 programmers learning Perl 6.

Autoincrement precedence

As in C, these operators increment or decrement the object in question either before or after the value is taken from the object, depending on whether it is put before or after. Also as in C, multiple references to a single mutating object in the same expression may result in undefined behavior unless some explicit sequencing operator is interposed. See "Sequence points".

As with all postfix operators in Perl 6, no space is allowed between a term and its postfix. See S02 for why, and for how to work around the restriction with an "unspace".

As mutating methods, all these operators dispatch to the type of the operand and return a result of the same type, but they are legal on value types only if the (immutable) value is stored in a mutable container. However, a bare undefined value (in a suitable Scalar container) is allowed to mutate itself into an Int in order to support the common idiom:

    say $x unless %seen{$x}++;

Increment of a Str (in a suitable container) works similarly to Perl 5, but is generalized slightly. A scan is made for the final alphanumeric sequence in the string that is not preceded by a '.' character. Unlike in Perl 5, this alphanumeric sequence need not be anchored to the beginning of the string, nor does it need to begin with an alphabetic character; the final sequence in the string matching <!after '.'> <rangechar>+ is incremented regardless of what comes before it.

The <rangechar> character class is defined as that subset of \w that Perl knows how to increment within a range, as defined below.

The additional matching behaviors provide two useful benefits: for its typical use of incrementing a filename, you don't have to worry about the path name or the extension:

    $file = "/tmp/pix000.jpg";
    $file++;            # /tmp/pix001.jpg, not /tmp/pix000.jph

Perhaps more to the point, if you happen to increment a string that ends with a decimal number, it's likely to do the right thing:

    $num = "123.456";
    $num++;             # 124.456, not 123.457

Character positions are incremented within their natural range for any Unicode range that is deemed to represent the digits 0..9 or that is deemed to be a complete cyclical alphabet for (one case of) a (Unicode) script. Only scripts that represent their alphabet in codepoints that form a cycle independent of other alphabets may be so used. (This specification defers to the users of such a script for determining the proper cycle of letters.) We arbitrarily define the ASCII alphabet not to intersect with other scripts that make use of characters in that range, but alphabets that intersperse ASCII letters are not allowed.

If the current character in a string position is the final character in such a range, it wraps to the first character of the range and sends a "carry" to the position left of it, and that position is then incremented in its own range. If and only if the leftmost position is exhausted in its range, an additional character of the same range is inserted to hold the carry in the same fashion as Perl 5, so incrementing '(zz99)' turns into '(aaa00)' and incrementing '(99zz)' turns into '(100aa)'.

The following Unicode ranges are some of the possible rangechar ranges. For alphabets we might have ranges like:

    A..Z        # ASCII uc
    a..z        # ASCII lc
    Α..Ω        # Greek uc
    α..ω        # Greek lc (presumably skipping U+03C2, final sigma)
    א..ת        # Hebrew
      etc.      # (XXX out of my depth here)

For digits we have ranges like:

    0..9        # ASCII
    ٠..٩        # Arabic-Indic
    ०..९        # Devangari
    ০..৯        # Bengali 
    ੦..੯        # Gurmukhi
    ૦..૯        # Gujarati
    ୦..୯        # Oriya
      etc.

Other non-script 0..9 ranges may also be incremented, such as

    ⁰..⁹        # superscripts (note, cycle includes latin-1 chars)
    ₀..₉        # subscripts
    0..9      # fullwith digits

Conjecturally, any common sequence may be treated as a cycle even if it does not represent 0..9:

    Ⅰ..Ⅻ        # clock roman numerals uc
    ⅰ..ⅻ        # clock roman numerals lc
    ①..⑳        # circled digits 1..20
    ⒜..⒵        # parenthesize lc
    ⚀..⚅        # die faces 1..6
    ❶..❿        # dingbat negative circled 1..10
      etc.

While it doesn't really make sense to "carry" such numbers when they reach the end of their cycle, treating such values as incrementable may be convenient for writing outlines and similar numbered bullet items. (Note that we can't just increment unrecognized characters, because we have to locate the string's final sequence of rangechars before knowing which portion of the string to increment. Note also that all character increments can be handled by lookup in a single table of successors since we've defined our ranges not to include overlapping cycles.)

Perl 6 also supports Str decrement with similar semantics, simply by running the cycles the other direction. However, leftmost characters are never removed, and the decrement fails when you reach a string like "aaa" or "000".

Increment and decrement on non-<Str> types are defined in terms of the .succ and .pred methods on the type of object in the Scalar container. More specifically,

    ++$var
    --$var

are equivalent to

    $var.=succ
    $var.=pred

If the type does not support these methods, the corresponding increment or decrement operation will fail. (The optimizer is allowed to assume that the ordinary increment and decrement operations on integers will not be overridden.)

Increment of a Bool (in a suitable container) turns it true. Decrement turns it false regardless of how many times it was previously incremented. This is useful if your %seen array is actually a KeySet, in which case decrement actually deletes it from the KeySet.

  • Autoincrement prefix:<++> or postfix:<++> operator

        $x++
        ++$x;
  • Autodecrement prefix:<--> or postfix:<--> operator

        $x--
        --$x

Exponentiation precedence

  • infix:<**> exponentiation operator

        $x ** 2

    If the right argument is not a non-negative integer, the result is likely to be an approximation. If the right argument is of an integer type, exponentiation is at least as accurate as repeated multiplication on the left side's type. (From which it can be deduced that Int**UInt is always exact, since Int supports arbitrary precision.) If the right argument is an integer represented in a non-integer type, the accuracy is left to implementation provided by that type; there is no requirement to recognize an integer to give it special treatment.

Symbolic unary precedence

  • prefix:<?>, boolean context

        ?$x

    Evaluates the expression as a boolean and returns True if expression is true or False otherwise. See "true" below for a low-precedence alternative.

  • prefix:<!>, boolean negation

        !$x

    Returns the opposite of what ? would. See "not" below for a low-precedence alternative.

  • prefix:<+>, numeric context

        +$x

    Unlike in Perl 5, where + is a no-op, this operator coerces to numeric context in Perl 6. (It coerces only the value, not the original variable.) The narrowest appropriate type of Int, Num, or Complex will be returned. A string containing two integers separated by a / will be returned as a Rat. Exponential notation and radix notations are recognized.

  • prefix:<->, numeric negation

        -$x

    Coerces to numeric and returns the arithmetic negation of the resulting number.

  • prefix:<~>, string context

        ~$x

    Coerces the value to a string. (It only coerces the value, not the original variable.)

  • prefix:<|>, flatten object into arglist

        | $capture

    Interpolates the contents of the Capture (or Capture-like) value into the current argument list as if they had been specified literally.

  • prefix:<+^>, numeric bitwise negation

        +^$x

    Coerces to integer and then does bitwise negation (complement) on the number.

  • prefix:<~^>, string bitwise negation

        ~^$x
    Coerces to string buffer and then does bitwise negation (complement)
    on each element.
  • prefix:<?^>, boolean bitwise negation

        ?^$x

    Coerces to boolean and then flips the bit. (Same as !.)

  • prefix:<\>, Capture constructor

        \$x
        \@x
        \%x
        \($invocant: $pos1, $pos2, :named($arg))

    Defers the contextualization of its argument or arguments till it is bound into some other context.

  • prefix:<^>, upto operator

        ^$limit

    Constructs a range of 0..^$limit or locates a metaclass as a shortcut for $limit.HOW. See "Range semantics".

  • prefix:<=>, iterate iterator

        =$iterator

    Unary = reads lines from a filehandle or filename, or iterates an iterator, or in general causes a scalar to explode its guts when it would otherwise not. How it does that is context sensitive. For instance, =$iterator is item/list sensitive and will produce one value in item context but many values in list context. (Use @$iterator to force a fetch of all the values even in item context, and $$iterator to force a fetch of a single value even in list context.) On the other hand, =$capture interpolates all parts of the capture that makes sense in the current list context, depending on what controls that list context.

Multiplicative precedence

  • infix:<*>

        $x*$y

    Multiplication, resulting in wider type of the two.

  • infix:</>

        $numerator / $denominator

    If either operand is of Num type, converts both operands to Num and does division returning Num. If the denominator is zero, returns an object representing either +Inf, NaN, or -Inf as the numerator is positive, zero, or negative. (This is construed as the best default in light of the operator's possible use within hyperoperators and junctions. Note however that these are not actually the native IEEE non-numbers; they are undefined values of the "unthrown exception" type that happen to represent the corresponding IEEE concepts, and if you subsequently try to use one of these values in a non-parallel computation, it will likely throw an exception at that point.)

    If both operands are of integer type, you still get a Num, but the Num type is allowed to do the division lazily; internally it may store a Rat until the time a value is called for. If converted to Rat directly no division ever need be done.

  • infix:<div>, generic division

        $numerator div $denominator

    Dispatches to the infix:<div> multi most appropriate to the operand types. Policy on what to do about division by zero is up to the type, but for the sake of hyperoperators and junctions those types that can represent overflow (or that can contain an unthrown exception) should try to do so rather than simply throwing an exception. (And in general, other operators that might fail should also consider their use in hyperops and junctions, and whether they can profitably benefit from a lazy exception model.)

    Use of div on two Int values results in a ratio of the Rat type. Integerizing a rat involves truncating its tail.

  • infix:<%>, modulus

        $x % $mod

    Always floor semantics using Num or Int.

  • infix:<mod>, generic modulus

        $x mod $mod

    Dispatches to the infix:<mod> multi most appropriate to the operand types.

  • infix:{'+&'}, numeric bitwise and

        $x +& $y

    Converts both arguments to integer and does a bitwise numeric AND.

  • infix:{'+<'}, numeric shift left

        $integer +< $bits
  • infix:{'+>'}, numeric shift right

        $integer +> $bits

    By default, signed types do sign extension, while unsigned types do not, but this may be enabled or disabled with a :signed or :!signed adverb.

  • infix:<~&>, buffer bitwise and

        $x ~& $y
  • infix:{'~<'}, buffer bitwise shift left

        $buf ~< $bits
  • infix:{'~>'}, buffer bitwise shift right

        $buf ~> $bits

    Sign extension is not done by default but may be enabled with a :signed adverb.

  • infix:<?&>, boolean bitwise and

        $x ?& $y

Any bit shift operator may be turned into a rotate operator with the :rotate adverb. If :rotate is specified, the concept of sign extension is meaningless, and you may not specify a :signed adverb.

Additive precedence

  • infix:<+>, numeric addition

        $x + $y

    Microeditorial: As with most of these operators, any coercion or type mismatch is actually handled by multiple dispatch. The intent is that all such variants preserve the notion of numeric addition to produce a numeric result, presumably stored in suitably "large" numeric type to hold the result. Do not overload the + operator for other purposes, such as concatenation. (And please do not overload the bitshift operators to do I/O.) In general we feel it is much better for you to make up a different operator than overload an existing operator for "off topic" uses. All of Unicode is available for this purpose.

  • infix:<->, numeric subtraction

        $x - $y
  • infix:<+|>, numeric bitwise inclusive or

        $x +| $y
  • infix:<+^> numeric bitwise exclusive or

        $x +^ $y
  • infix:<~|>, buffer bitwise inclusive or

        $x ~| $y
  • infix:<~^> buffer bitwise exclusive or

        $x ~^ $y
  • infix:<?|>, boolean bitwise inclusive or

        $x ?| $y
  • infix:<?^> boolean bitwise exclusive or

        $x ?^ $y

Replication

  • infix:<x>, string/buffer replication

        $string x $count

    Evaluates the left argument in string context, replicates the resulting string value the number of times specified by the right argument and returns the result as a single concatenated string regardless of context.

    If the count is less than 1, returns the null string. The count may not be * because Perl 6 does not support infinite strings. (At least, not yet...) Note, however, that an infinite string may be emulated with cat($string xx *).

  • infix:<xx>, list replication

        @list xx $count

    Evaluates the left argument in list context, replicates the resulting Capture value the number of times specified by the right argument and returns the result in a context dependent fashion. If the operator is being evaluated in ordinary list context, the operator returns a flattened list. In slice (@@) context, the operator converts each Capture to a separate sublist and returns the list of those sublists.

    If the count is less than 1, returns the empty list, (). If the count is *, returns an infinite list (lazily, since lists are lazy by default).

Concatenation

  • infix:<~>, string/buffer concatenation

        $x ~ $y

Junctive and (all) precedence

  • infix:<&>, all() operator

        $x & $y

Junctive or (any) precedence

  • infix:<|>, any() operator

        $x | $y
  • infix:<^>, one() operator

        $x ^ $y

Named unary precedence

  • Functions of one argument

        rand sleep abs etc.

    Note that, unlike in Perl 5, you must use the .meth forms to default to $_ in Perl 6.

Nonchaining binary precedence

  • infix:<but>

        $value but Mixin
  • infix:<does>

        $object does Mixin
  • Sort comparisons

        $num1 <=> $num2
        $str1 leg $str2
        $obj1 cmp $obj2

    These operators compare their operands using numeric, string, or eqv semantics respectively, and depending on the order return one of Order::Increase, Order::Same, or Order::Decrease (which numerify to -1, 0, or +1). See "Comparison semantics".

  • Range object constructor

        $min .. $max
        $min ^.. $max
        $min ..^ $max
        $min ^..^ $max

    Constructs Range objects, optionally excluding one or both endpoints. See "Range semantics".

Chaining binary precedence

All operators on this precedence level may be chained; see "Chained comparisons".

  • infix:<==> etc.

        == != < <= > >=

    As in Perl 5, converts to Num before comparison. != is short for !==.

  • infix:<eq> etc.

        eq ne lt le gt ge

    As in Perl 5, converts to Str before comparison. ne is short for !eq.

  • Generic ordering

        $a before $b
        $a after $b
  • Smart match

        $obj ~~ $pattern

    Perl 5's =~ becomes the "smart match" operator ~~, with an extended set of semantics. See "Smart matching" for details.

    To catch "brainos", the Perl 6 parser defines an infix:<=~> operator which always fails at compile time with a message directing the user to use ~~ or ~= (string append) instead if they meant it as a single operator, or to put a space between if they really wanted to assign a stringified value as two separate operators.

    A negated smart match is spelled !~~.

  • Container identity

        VAR($a) =:= VAR($b)

    See "Comparison semantics".

  • Value identity

        $x === $y

    For objects that are not value types, their identities are their values. (Identity is returned by the .WHICH metamethod.) The actual contents of the objects are ignored. These semantics are those used by hashes that allow objects for keys. See also "Comparison semantics".

  • Canonical equivalence

        $obj1 eqv $obj2

    Compares two objects for canonical equivalence. For value types compares the values. For object types, compares current contents according to some scheme of canonicalization. These semantics are those used by hashes that allow only values for keys (such as Perl 5 string-key hashes). See also "Comparison semantics".

  • Negated relational operators

        $num !== 42
        $str !eq "abc"
        "foo" !~~ /^ <ident> $/
        VAR($a) !=:= VAR($b)
        $a !=== $b
        $a !eqv $b

    See "Negated relational operators".

Tight and precedence

  • infix:<&&>, short-circuit and

        $condition && $whentrue

    Returns the left argument if the left argument is false, otherwise evaluates and returns the right argument. In list context forces a false return to mean (). See and below for low-precedence version.

Tight or precedence

  • infix:<||>, short-circuiting inclusive-or

        $condition || $whenfalse

    Returns the left argument if it's true, otherwise evaluates and returns the right argument. It is specifically allowed to use a list or array both as a boolean and as a list value produced if the boolean is true:

        @a = @b || @c;              # broken in Perl 5; works in Perl 6

    In list context this operator forces a false return to mean (). See or below for low-precedence version.

  • infix:<^^>, exclusive-or

        $x ^^ $y

    Returns the true argument if there is one (and only one). Returns Bool::False if both arguments are false or both arguments are true. In list context forces a false return to mean (). See xor below for low-precedence version.

  • infix:<//>, default operator

        $value // $default

    Returns the left argument if it's defined, otherwise evaluates and returns the right argument. In list context forces a false return to mean (). See orelse below for a similar but not identical low-precedence version.

  • Minimum and maximum

        $min0 min $min1
        $max0 max $max1

    Instead of deciding whether to return the left or right based on booleans or definedness, these return the minimum or maximum value. See also the minmax listop.

    Not all types can support the concept of infinity. Therefore any value of any type may be compared with +Inf or -Inf values, in which case the infinite value stands for "larger/smaller than any possible value of the type." That is,

        "foo" min +Inf              # "foo"
        "foo" min -Inf              # -Inf
        "foo" max +Inf              # +Inf
        "foo" max -Inf              # "foo"

    All orderable object types must support +Inf and -Inf values as special forms of the undefined value. It's an error, however, to attempt to store an infinite value into a native type that cannot support it:

        my int $max;
        $max max= -Inf;     # ERROR

Conditional operator precedence

  • Conditional operator

        say "My answer is: ", $maybe ?? "yes" !! "no";

    Also known as the "ternary" or "trinary" operator, but we prefer "conditional" just to stop people from fighting over the terms. The operator syntactically separates the expression into three subexpressions. It first evaluates the left part in boolean context, then based on that selects one of the other two parts to evaluate. (It never evaluates both of them.) If the conditional is true it evaluates and returns the middle part; if false, the right part. The above is therefore equivalent to:

        say "My answer is: ", do {
            if $maybe {
                "yes";
            }
            else {
                "no";
            }
        };

    It is a syntax error to use an operator in the middle part that binds looser in precedence, such as =.

        my $x;
        hmm() ?? $x = 1 !! $x = 2;        # ERROR
        hmm() ?? ($x = 1) !! ($x = 2);    # works

    Note that both sides have to be parenthesized. A partial fix is even wronger:

        hmm() ?? ($x = 1) !! $x = 2;      # parses, but WRONG

    That actually parses as:

        (
            hmm() ?? ($x = 1) !! $x
        ) = 2;

    and always assigns 2 to $x (because ($x = 1) is a valid lvalue).

    And in any case, repeating the $x forces you to declare it earlier. The best don't-repeat-yourself solution is simply:

        my $x = hmm() ?? 1 !! 2;          # much better
  • infix:<?>

    To catch likely errors by people familiar with C-derived languages (including Perl 5), a bare question mark in infix position will produce an error suggesting that the user use ?? !! instead.

  • Flipflop ranges

        start() ff end()
        start() ^ff end()
        start() ff^ end()
        start() ^ff^ end()
  • Flipflop ranges (sed style)

        start() fff end()
        start() ^fff end()
        start() fff^ end()
        start() ^fff^ end()

Item assignment precedence

  • infix:<=>

        $x = 1, $y = 2;

    With simple lvalues, = has this precedence, which is tighter than comma. (List assignments have listop precedence below.)

  • infix:<:=>, run-time binding

        $signature := $capture

    A new form of assignment is present in Perl 6, called binding, used in place of typeglob assignment. It is performed with the := operator. Instead of replacing the value in a container like normal assignment, it replaces the container itself. For instance:

        my $x = 'Just Another';
        my $y := $x;
        $y = 'Perl Hacker';

    After this, both $x and $y contain the string "Perl Hacker", since they are really just two different names for the same variable.

    There is also an identity test, =:=, which tests whether two names are bound to the same underlying variable. $x =:= $y would return true in the above example.

    The binding fails if the type of the variable being bound is sufficiently inconsistent with the type of the current declaration. Strictly speaking, any variation on

        my Any $x;
        $x := [1,2,3];

    should fail because the type being bound is not consistent with Scalar of Any, but since the Any type is not a real instantiable type but a generic (non)constraint, and Scalar of Any is sort of a double non-constraint similar to Any, we treat this situation specially as the equivalent of binding to a typeless variable.

  • infix:<::=>, compile-time binding

        $signature ::= $capture

    This does the same as := except it does it at compile time. (This implies that the expression on the right is also evaluated at compile time; it does not bind a lazy thunk.)

  • infix:{'=>'}, Pair constructor

        foo => 1, bar => "baz"

    Binary => is no longer just a "fancy comma". It now constructs a Pair object that can, among other things, be used to pass named arguments to functions. It provides item context to both sides. It does not actually do an assignment except in a notional sense; however its precedence is now equivalent to assignment, and it is also right associative. Note that, unlike in Perl 5, => binds tighter than comma.

  • Assignment operators

        += -= **= xx= .= etc.

    See "Assignment operators".

Loose unary precedence

  • prefix:<true>

        true any(@args) eq '-v' | '-V'
  • prefix:<not>

        not any(@args) eq '-v' | '-V'

Comma operator precedence

  • infix:<,>, the argument separator

        1, 2, 3, @many

    Unlike in Perl 5, comma operator never returns the last value. (In item context it returns a list instead.)

List infix precedence

List infixes all have list associativity, which means that identical infix operators work together in parallel rather than one after the other. Non-identical operators are considered non-associative and must be parenthesized for clarity.

  • infix:<Z>, the zip operator

        1,2 Z 3,4   # (1,3),(2,4)
  • infix:<minmax>, the minmax operator

        $min0, $max0 minmax $min1, $max1    # ($min0 min $min1, $max0 max $max1)

    The minmax operator is for calculating both a minimum and maximum in a single expression. Otherwise you'd have to write twice as many expressions. Instead of

        @a minmax @b

    you'd have to say something like

        ($a[0] min $b[0], $a[1] max $b[1])

    Note that there is no guarantee that the resulting minimum and maximum come from the same side. The two calculations are bundled but independent.

  • infix:<X>, the cross operator

        1,2 X 3,4          # (1,3), (1,4), (2,3), (2,4)

    In contrast to the zip operator, the X operator returns all possible lists formed by taking one element from each of its list arguments. The returned lists are ordered such that the rightmost elements vary most rapidly. If there are just two lists, for instance, it forms all pairs where one element is from the first list and the other one from the second, with the second element varying most rapidly. Hence you may say:

        <a b> X <1 2>

    and you end up with

        ('a', '1'), ('a', '2'), ('b', '1'), ('b', '2')

    This becomes a flat list in @ context and a list of arrays in @@ context:

        say @(<a b> X <1 2>)
        'a', '1', 'a', '2', 'b', '1', 'b', '2'
        say @@(<a b> X <1 2>)
        ['a', '1'], ['a', '2'], ['b', '1'], ['b', '2']

    The operator is list associative, so

        1,2 X 3,4 X 5,6

    produces

        (1,3,5),(1,3,6),(1,4,5),(1,4,6),(2,3,5),(2,3,6),(2,4,5),(2,4,6)

    On the other hand, if any of the lists is empty, you will end up with a null list.

    Only the leftmost list may usefully be an infinite list. For instance

        <a b> X 0..*

    would produce

        ('a',0), ('a',1), ('a',2), ('a',3), ('a',4), ('a',5), ...

    and you'd never get to 'b'.

  • Cross hyperoperators

        @files X~X '.' X~X @extensions
        1..10 X*X 1..10
        @x XeqvX @y
        etc.

    See "Cross operators".

Many of these operators return a list of Captures, which depending on context may or may not flatten them all out into one flat list. The default is to flatten, but see the contextualizers below.

List prefix precedence

  • infix:<=>, list assignment

        @array = 1,2,3;

    With compound targets, performs list assignment. The right side is looser than comma. You might be wondering why we've classified this as a prefix operator when its token name is infix:<=>. That's because you can view the left side as a special syntax for a prefix listop, much as if you'd said:

        @array.assign: 1,2,3

    However, the tokener classifies it as infix because it sees it when it's expecting an infix operator. Assignments in general are treated more like retroactive macros, since their meaning depends greatly on what is on the left, especially if what is on the left is a declarator of some sort. We even call some of them pseudo-assignments, but they're all a bit pseudo insofar as we have to figure out whether the left side is a list or a scalar destination.

  • infix:<:>, the invocant marker

        say $*OUT: "howdy, world"

    The colon operator turns method calls and contextualizers into list operators. It's not really a general operator; much like list assignment, it takes a special syntax on the left side and turns it into a list operator over the list on the right. See "Invocant marker".

  • Normal listops

        print push say join split substr open etc.
  • Listop forms of junctional operators

        any all one none
  • Exception generators

        fail "Division by zero"
        die System::Error(ENOSPC,"Drive $d seems to be full");
        warn "Can't open file: $!"
  • Stubby exception generators

        ...
        !!! "fill this in later, Dave"
        ??? "oops in $?CLASS"

    The ... operator is the "yada, yada, yada" list operator, which among other things is used as the body in function prototypes. It complains bitterly (by calling fail) if it is ever executed. Variant ??? calls warn, and !!! calls die. The argument is optional, but if provided, is passed onto the fail, warn, or die. Otherwise the system will make up a message for you based on the context, indicating that you tried to execute something that is stubbed out. (This message differs from what fail, warn, and die would say by default, since the latter operators typically point out bad data or programming rather than just an incomplete design.)

  • Reduce operators

        [+] [*] [<] [\+] [\*] etc.

    See "Reduction operators".

  • Sigils as contextualizer listops

        Sigil       Alpha variant
        -----       -------------
        $           item
        @           list
        @@          slice
        %           hash

    As listops, these look like terms from the left, but raise their precedence on the right sufficiently to govern list infix operators:

         $ 1,2 Z 3,4       # [\(1,3),\(2,4)]
         @ 1,2 Z 3,4       # 1,3,2,4
        @@ 1,2 Z 3,4       # [1,3],[2,4]
         % 1,2 Z 3,4       # { 1 => 3, 2 => 4 }
    
         $ 1,2 X 3,4      # [\(1,3),\(1,4),\(2,3),\(2,4)]
         @ 1,2 X 3,4      # 1,3,1,4,2,3,2,4
        @@ 1,2 X 3,4      # [1,3],[1,4],[2,3],[2,4]

    These can also influence the result of functions that returns lists of captures:

         $ map { $_, $_*2 }, ^4   # [\(0,0),\(1,2),\(2,4),\(3,6)]
         @ map { $_, $_*2 }, ^4   # 0,0,1,2,2,4,3,6
        @@ map { $_, $_*2 }, ^4   # [0,0],[1,2],[2,4],[3,6]
         % map { $_, $_*2 }, ^4   # { 0 => 0, 1 => 2, 2 => 4, 3 => 6 }
  • The item contextualizer

        item foo()

    The new name for Perl 5's scalar contextualizer. Equivalent to $(). We still call the values scalars, and talk about "scalar operators", but scalar operators are those that put their arguments into item context.

    If given a list, this function makes an Array from it. The function is agnostic about any Captures in such a list. (Use @ or @@ below to force that one way or the other).

  • The list contextualizer

        list foo()

    Forces the subsequent expression to be evaluated in list context. A list of Captures will be transformed into a flat list. Equivalent to @().

  • The slice contextualizer

        slice foo()

    Forces the subsequent expression to be evaluated in slice context. (Slices are considered to be potentially multidimensional in Perl 6.) A list of Captures will be transformed into a list of lists. Equivalent to @@().

  • The hash contextualizer

        hash foo()

    Forces the subsequent expression to be evaluated in hash context. The expression is evaluated in list context (flattening any Captures), then a hash will be created from the list, taken as a list of Pairs. (Any element in the list that is not a Pair will pretend to be a key and grab the next value in the last as its value.) Equivalent to %().

Loose and precedence

  • infix:<and>, short-circuit and

        $condition and $whentrue

    Returns the left argument if the left argument is false, otherwise evaluates and returns the right argument. In list context forces a false return to mean (). See && above for high-precedence version.

  • infix:<andthen>, proceed on success

        test1() andthen test2()

    Returns the left argument if the left argument indicates failure (that is, if the result is undefined). Otherwise it evaluates and returns the right argument.

    If the right side is a block or pointy block, the result of the left side is bound to any arguments of the block. If the right side is not a block, a block scope is assumed around the right side, and the result of the left side is implicitly bound to $_ for the scope of the right side. That is,

        test1() andthen test2()

    is equivalent to

        test1() andthen -> $_ { test2() }

    There is no corresponding high-precedence version.

Loose or precedence

  • infix:<or>, short-circuit inclusive or

        $condition or $whenfalse

    Returns the left argument if it's true, otherwise evaluates and returns the right argument. In list context forces a false return to mean (). See || above for high-precedence version.

  • infix:<xor>, exclusive or

        $x xor $y

    Returns the true argument if there is one (and only one). Returns Bool::False if both arguments are false or both arguments are true. In list context forces a false return to mean (). See ^^ above for high-precedence version.

  • infix:<orelse>, proceed on failure

        test1() orelse test2()

    Returns the left argument if the left argument indicates success (that is, if the result is defined). Otherwise it evaluates and returns the right argument.

    If the right side is a block or pointy block, the result of the left side is bound to any arguments of the block. If the right side is not a block, a block scope is assumed around the right side, and the result of the left side is implicitly bound to $! for the scope of the right side. That is,

        test1() orelse test2()

    is equivalent to

        test1() orelse -> $! { test2() }

    (The low-precedence // operator is similar, but does not set $! or treat blocks specially.)

Terminator precedence

As with terms, terminators are not really a precedence level, but looser than the loosest precedence level. They all have the effect of terminating any operator precedence parsing and returning a complete expression to the main parser. They don't care what state the operator precedence parser is in. If the parser is currently expecting a term and the final operator in the expression can't deal with a nullterm, then it's a syntax error. (Notably, the comma operator and many prefix list operators can handle a nullterm.)

  • Semicolon: ;

        $x = 1; $y = 2;

    The context determines how the expressions terminated by semicolon are interpreted. At statement level they are statements. Within a bracketing construct they are interpreted as lists of Captures, which in slice context will be treated as the multiple dimensions of a multidimensional slice. (Other contexts may have other interpretations or disallow semicolons entirely.)

  • Feed operators: <==, ==>, <<==, ==>>

        source() ==> filter() ==> sink()

    The forms with the double angle append rather than clobber the sink's todo list. The ==>> form always looks ahead for an appropriate target to append to, either the final sink in the chain, or the next filter stage with an explicit @(*) or @@(*) target. This means you can stack multiple feeds onto one filter command:

        source1() ==>>
        source2() ==>>
        source3() ==>>
        filter(@(*)) ==> sink()

    Similar semantics apply to <<== except it looks backward for an appropriate target to append to.

  • Control block: <ws>{...}

    When a block occurs after whitespace where an infix is expected, it is interpreated as a control block for a statement control construct. (If there is no whitespace, it is a subscript, and if it is where a term is expected, it's just a bare closure.) If there is no statement looking for such a block currently, it is a syntax error.

  • Statement modifiers: if, unless, while, until, for

    Statement modifiers terminate one expression and start another.

  • Any unexpected ), ], } at this level.

    Calls into the operator precedence parser may be parameterized to recognize additional terminators, but right brackets of any sort (except angles) are automatically included in the set of terminators as tokens of length one. (An infix of longer length could conceivably start with one of these characters, and would be recognized under the longest-token rule and continue the expression, but this practice is discouraged. It would be better to use Unicode for your weird operator.) Angle brackets are exempted so that they can form hyperoperators (see "Hyper operators").

  • A block-final } at the end of the line terminates the current expression. A block within an argument list terminates the argument list unless followed by the comma operator.

Changes to Perl 5 operators

Several operators have been given new names to increase clarity and better Huffman-code the language, while others have changed precedence.

  • Perl 5's ${...}, @{...}, %{...}, etc. dereferencing forms are now $(...), @(...), %(...), etc. instead. Listop-like forms use the bare sigil following by whitespace. Use of the Perl 5 curly forms will result in an error message pointing the user to the new forms.

  • -> becomes ., like the rest of the world uses. There is a pseudo postfix:{'->'} operator that produces a compile-time error reminding Perl 5 users to use dot instead. (The "pointy block" use of -> in Perl 5 requires preceding whitespace when the arrow could be confused with a postfix, that is when an infix is expected. Preceding whitespace is not required in term position.)

  • The string concatenation . becomes ~. Think of it as "stitching" the two ends of its arguments together. String append is likewise ~=.

  • The filetest operators are gone. We now use a Pair as either a pattern or a method name to get the same effect:

        if $filename ~~ :e { say "exists" }
        if $filename.:e { say "exists" }

    The 1st form actually translates to the latter form, so the object's class decides how to dispatch pair methods. It just happens that Str (filenames), IO (filehandles), and Statbuf (stat buffers) default to the expected filetest semantics, but $regex.:i might tell you whether the regex is case insensitive, for instance.

    Using the pattern form, multiple tests may be combined via junctions:

        given $handle {
            when :r & :w & :x {...}
            when :!w | :!x    {...}
            when *            {...}
        }

    When adverbial pairs are stacked into one term, it is assumed they are ANDed together, so

        when :r :w :x

    is equivalent to either of:

        when :r & :w & :x
        when all(:r,:w,:x)

    The advantage of the method form is that it can be used in places that require tighter precedence than ~~ provides:

        sort { $^a.:M <=> $^b.:M }, @files

    though that's a silly example since you could just write:

        sort { .:M }, @files

    But that demonstrates the other advantage of the method form, which is that it allows the "unary dot" syntax to test the current topic.

    Unlike in earlier versions of Perl 6, these filetests do not return stat buffers, but simple scalars of type Bool, Int, or Num.

    In general, the user need not worry about caching the stat buffer when a filename is queried. The stat buffer will automatically be reused if the same object has recently been queried, where "recently" is defined as less than a second or so. If this is a concern, an explicit stat() or lstat() may be used to return an explicit stat buffer object that will not be subject to timeout, and may be tested repeatedly just as a filename or handle can. A Statbuf object has a .file method that can be queried for its filename (if known); the .io method returns the handle (if known). If the Statbuf object doesn't know its filename but does know its IO handle, then .file attempts to return .io.file.

    Note that :s still returns the filesize, but :!s is true only if the file is of size 0.

    (Inadvertent use of the Perl 5 forms will normally result in treatment as a negated postdeclared subroutine, which is likely to produce an error message at the end of compilation.)

  • All postfix operators that do not start with a dot also have an alternate form that does. (The converse does not hold--just because you can write x().foo doesn't mean you can write x()foo. Likewise the ability to say $x.'foo' does not imply that $x'foo' will work.)

    The postfix interpretation of an operator may be overridden by use of a quoted method call, which calls the prefix form instead. So x().! is always the postfix operator, but x().'!' will always call !x(). In particular, you can say things like $array.'@'. and $fh.'=', which because of the quotes will not be confused lexically with $fh.=new.

  • Unary ~ now imposes a string (Str) context on its argument, and + imposes a numeric (Num) context (as opposed to being a no-op in Perl 5). Along the same lines, ? imposes a boolean (Bool) context, and the | unary operator imposes a function-arguments (Capture) context on its argument. Unary sigils impose the container context implied by their sigil. As with Perl 5, however, $$foo[bar] parses as ( $($foo) )[bar], so you need $($foo[bar]) to mean the other way.

  • Bitwise operators get a data type prefix: +, ~, or ?. For example, Perl 5's | becomes either +| or ~| or ?|, depending on whether the operands are to be treated as numbers, strings, or boolean values. Perl 5's left shift << becomes +< , and correspondingly with right shift. Perl 5's unary ~ (one's complement) becomes either +^ or ~^ or ?^, since a bitwise NOT is like an exclusive-or against solid ones. Note that ?^ is functionally identical to !, but conceptually coerces to boolean first and then flips the bit. Please use ! instead.

    ?| is a logical OR but differs from || in that ?| always evaluates both sides and returns a standard boolean value. That is, it's equivalent to ?$a + ?$b != 0. Another difference is that it has the precedence of an additive operator.

    ?& is a logical AND but differs from && in that ?& always evaluates both sides and returns a standard boolean value. That is, it's equivalent to ?$a * ?$b != 0. Another difference is that it has the precedence of a multiplicative operator.

    Bitwise string operators (those starting with ~) may only be applied to Buf types or similar compact integer arrays, and treat the entire chunk of memory as a single huge integer. They differ from the + operators in that the + operators would try to convert the string to a number first on the assumption that the string was an ASCII representation of a number.

  • x splits into two operators: x (which concatenates repetitions of a string to produce a single string), and xx (which creates a list of repetitions of a list or item). "foo" xx * represents an arbitrary number of copies, useful for initializing lists. The left side of an xx is evaluated only once. (To call a block repeatedly, use a map instead.)

  • The ? : conditional operator becomes ?? !!. A pseudo operator, infix:<?>, catches migratory brainos at compile time.

  • qw{ ... } gets a synonym: < ... >, and an interpolating variant, «...». For those still living without the blessings of Unicode, that can also be written: << ... >>.

  • In item context comma , now constructs a List object from its operands. You have to use a [*-1] subscript to get the last one. (Note the *. Negative subscripts no longer implicitly count from the end; in fact, the compiler may complain if you use [-1] on an object known at compile time not to have negative subscripts.)

  • The unary backslash operator captures its arguments, and returns an object representing those arguments. You can dereference this object in several ways to retrieve different parts of the arguments; see the definition of Capture in S02 for details. (No whitespace is allowed after the backslash because that would instead start an "unspace", that is, an escaped sequence of whitespace or comments. See S02 for details. However, oddly enough, because of that unspace rule, saying \\ $foo turns out to be equivalent to \$foo.)

  • The old .. flipflop operator is now done with ff operator. (.. now always produces a Range object even in item context.) The ff operator may take a caret on either end to exclude either the beginning or ending. There is also a corresponding fff operator with Perl 5's ... semantics. You may say

        /foo/ ff *

    to indicate a flipflop that never flops once flipped.

  • All comparison operators are unified at the same precedence level. See "Chained comparisons" below.

  • The list assignment operator now parses on the right like any other list operator, so you don't need parens on the right side of:

        @foo = 1,2,3;

    You do still need them on the left for

        ($a,$b,$c) = 1,2,3;

    since assignment operators are tighter than comma to their left.

  • The scalar assignment operator still parses as it did before, so

        loop ($a = 1, $b = 2; ; $a++, $b++) {...}

    still works fine. The syntactic distinction between scalar and list assignment is similar to the way Perl 5 defines it, but has to be a little different because we can no longer decide on the basis of the sigil. The following forms are parsed as "simple lvalues", and imply scalar assignment:

        $a          # simple scalar variable
        $(ANY)      # scalar dereference (including $$a)
        $::(ANY)    # symbolic scalar dereference
        ANY[SIMPLE] # single simple subscript
        ANY{SIMPLE} # single simple subscript
        ANY<x>      # single literal subscript

    Where SIMPLE is (recursively) defined as one of the forms above, plus the following forms:

        123         # single literal
        'x'         # single literal
        "$x"        # single literal
        qq/$x/      # single literal
        +TERM       # any single term coerced to numeric
        -TERM       # any single term coerced to numeric
        ~TERM       # any single term coerced to string
        ?TERM       # any single term coerced to boolean
        !TERM       # any single term coerced to boolean
        (SIMPLE)    # any simple expression in circumfix parens

    Note that circumfix parens are considered simple only when used as part of a subscript. Putting parens around the entire lvalue still implies list context as in Perl 5.

    We also include:

        OP SIMPLE   
        SIMPLE OP
        SIMPLE OP SIMPLE

    where OP includes any standard scalar operators in the five precedence levels autoincrement, exponentiation, symbolic unary, multiplicative, and additive; but these are limited to standard operators that are known to return numbers, strings, or booleans.

    Operators that imply list operations are excluded: prefix @, prefix % and infix xx, for instance. Hyper operators are also excluded, but post-assignment forms such as SIMPLE += SIMPLE are allowed.

    All other forms imply parsing as a list assignment, which may or may not result in a list assignment at run time. (See below.) However, this is exclusively a syntactic distinction, and no semantic or type information is used, since it influences subsequent parsing. In particular, even if a function is known to return a scalar value from its declaration, you must use + or ~ if you wish to force scalar parsing from within a subscript:

        @a[foo()] = bar();          # foo() and bar() called in list context
        @a[+foo()] = bar();         # foo() and bar() called in item context

    But note that the first form still works fine if foo() and bar() are item-returning functions that are not context sensitive. The difference in parsing is only an issue if bar() is followed by a comma or some such.

    For non-simple lvalues, at run time, both sides are evaluated in list context, but if the left side results in a single non-list scalar, the right side is treated as a single scalar value, as if the right side had been evaluated in list context (which is indeed the case) but coerced into item context.

    If the left side returns a list, however, then regardless of whether the list contains a single or multiple values, the right side values are assigned one by one as in any other list assignment, discarding any extra values if the right side is too long, or assigning undef if the right side is too short. To force list assignment when a subscript would return a non-list, either put parens around the entire lvalue, or use a comma within the subscript. (A semicolon in the subscript also works to indicate a multidimensional slice.)

    Assuming

        sub bar { return <a b c> }

    then we have:

        sub foo { return 1,2,3 }
        @a[foo()] = bar();          # (@a[1,2,3]) = <a b c>
    
        sub foo { return 1 }
        @a[foo()] = bar();          # @a[1] = [<a b c>]
    
        sub foo { return(1) }
        @a[foo()] = bar();          # @a[1] = [<a b c>]
    
        sub foo { return (1) }
        @a[foo()] = bar();          # (@a[1]) = <a b c>
    
        sub foo { return 1 }
        @a[foo(),] = bar();         # (@a[1]) = <a b c>
    
        sub foo { return 1 }
        (@a[foo()]) = bar();        # (@a[1]) = <a b c>

    Those are all parsed as list assignments, but we get different run-time behaviors based on the run-time type of the left side.

    In general, this will all just do what the user expects most of the time. The rest of the time item or list behavior can be forced with minimal syntax.

  • List operators are all parsed consistently. As in Perl 5, to the left a list operator look like term, while to the right it looks like an operator that is looser than comma. Unlike in Perl 5, the difference between the list operator form and the function form is consistently indicated via whitespace between the list operator and the first argument. If there is whitespace, it is always a list operator, and the next token will be taken as the first term of the list (or if there are no terms, as the expression terminator). Any infix operator occurring where a term is expected will be misinterpreted as a term:

        say + 2;    # means say(+2);

    If there is no whitespace, subsequent parsing depends on the syntactic category of the next item. Parentheses (with or without a dot) turn the list operator into a function call instead, and all the function's arguments must be passed inside the parentheses (with the sole exception of an adverbial block, which may follow the parentheses).

    Other than various forms of parentheses, all other postfixes are disallowed immediately after a list operator, even if there are no arguments. To add a postfix to an argumentless list operator you must write it as a function call with empty parentheses:

        foo.[]      # ILLEGAL
        foo++       # ILLEGAL
        foo().[]    # legal
        foo()++     # legal (if foo() is rw)

    After the parentheses any postfix operators are allowed, and apply to the result of the function call. (Also note that the postfix restriction applies only to list operators; it doesn't apply to methods. It is legal to say

        $foo.bar<a b c>

    to mean

        $foo.bar().{'a','b','c'}

    because methods never assume there are arguments unless followed by parentheses or a colon.)

    If the next item after the list operator is either an infix operator or a term, a syntax error is reported. [Conjecture: this may be relaxed in non-strict mode.]

    Examples:

        say foo + 1;                        say(foo(+1));
        say foo $x;                         say(foo($x));
        say foo$x;                          ILLEGAL, need space or parens
        say foo+1;                          ILLEGAL, need space or parens
        say foo++;                          ILLEGAL, need parens
    
        say foo($bar+1),$baz                say(foo($bar+1), $baz);
        say foo.($bar+1),$baz               say(foo($bar+1), $baz);
        say foo ($bar+1),$baz               say(foo($bar+1, $baz));
        say foo .($bar+1),$baz              say(foo($_.($bar+1), $baz));
    
        say foo[$bar+1],$baz                ILLEGAL, need foo()[]
        say foo.[$bar+1],$baz               ILLEGAL, need foo().[]
        say foo [$bar+1],$baz               say(foo([$bar+1], $baz));
        say foo .[$bar+1],$baz              say(foo($_.[$bar+1], $baz));
    
        say foo{$bar+1},$baz                ILLEGAL, need foo(){}
        say foo.{$bar+1},$baz               ILLEGAL, need foo().{}
        say foo {$bar+1},$baz               say(foo({$bar+1}, $baz));
        say foo .{$bar+1},$baz              say(foo($_.{$bar+1}, $baz));
    
        say foo<$bar+1>,$baz                ILLEGAL, need foo()<>
        say foo.<$bar+1>,$baz               ILLEGAL, need foo().<>
        say foo <$bar+1>,$baz               say(foo(<$bar+1>, $baz));
        say foo .<$bar+1>,$baz              say(foo($_.<$bar+1>, $baz));

    Note that Perl 6 is making a consistent three-way distinction between term vs postfix vs infix, and will interpret an overloaded character like < accordingly:

        any <a b c>                 any('a','b','c')        # term
        any()<a b c>                (any).{'a','b','c'}     # postfix
        any() < $x                  (any) < $x              # infix
        any<a b c>                  ILLEGAL                 # stealth postfix

    This will seem unfamiliar and "undwimmy" to Perl 5 programmers, who are used to a grammar that sloppily hardwires a few postfix operators at the price of extensibility. Perl 6 chooses instead to mandate a whitespace dependency in order to gain a completely extensible class of postfix operators.

  • A list operator's arguments are also terminated by a closure that is not followed by a comma or colon. (And a semicolon is implied if the closure is the final thing on a line. Use an "unspace" to suppress that.) This final closure may be followed by a postfix, in which case the postfix is applied to the result of the entire list operator.

  • A function predeclared as 0-ary is never considered a list operator, though it allows an optional set of empty parentheses. Unlike functions and list operators with arguments (see above), a 0-ary function does not require parentheses even if followed immediately by a postfix.

  • A non-multi sub predeclared with an arity of exactly 1 parses as a named unary in precedence. All other subs with arguments parse as list operators. (In other words, a named unary operator may be declared to take extra arguments only if they are named-only arguments.)

  • The && and || operators are smarter about list context and return () on failure in list context rather than Bool::False. The operators still short-circuit, but if either operator would return a false value, it is converted to the null list in list context so that the false results are self-deleting. (If this self-deleting behavior is not desired, put the expression into item context rather than list context.) This self-deletion is a behavior of the operators themselves, not a general property of boolean values in list context, so

        @foo = true($a||$b);

    is guaranteed to insert exactly one boolean value into @foo.

Junctive operators

|, &, and ^ are no longer bitwise operators (see "Changes to Perl 5 operators") but now serve a much higher cause: they are now the junction constructors.

A junction is a single value that is equivalent to multiple values. They thread through operations, returning another junction representing the result:

     (1|2|3) + 4;                            # 5|6|7
     (1|2) + (3&4);                          # (4|5) & (5|6)

Note how when two junctions are applied through an operator, the result is a junction representing the operator applied to each combination of values.

Junctions come with the functional variants any, all, one, and none.

This opens doors for constructions like:

     unless $roll == any(1..6) { print "Invalid roll" }

     if $roll == 1|2|3 { print "Low roll" }

Junctions work through subscripting:

    doit() if @foo[any(1,2,3)]

Junctions are specifically unordered. So if you say

    foo() | bar() | baz() == 42

it indicates to the compiler that there is no coupling between the junctional arguments. They can be evaluated in any order or in parallel. They can short-circuit as soon as any of them return 42, and not run the others at all. Or if running in parallel, the first successful thread may terminate the other threads abruptly. In general you probably want to avoid code with side effects in junctions.

Use of negative operators with syntactically recognizable junctions may produce a warning on code that works differently in English than in Perl. Instead of writing

    if $a != 1 | 2 | 3 {...}

you need to write

    if not $a == 1 | 2 | 3 {...}

However, this is only a syntactic warning, and

    if $a != $b {...}

will not complain if $b happens to contain a junction at runtime.

Junctive methods on arrays, lists, and sets work just like the corresponding list operators. However, junctive methods on a hash make a junction of only the hash's keys. Use the listop form (or an explicit .pairs) to make a junction of pairs.

Comparison semantics

  • Perl 5's comparison operators are basically unchanged, except that they can be chained because their precedence is unified.

  • Binary === tests immutable type and value correspondence: for two value types (that is, immutable types), tests whether they are the same value (eg. 1 === 1); for two mutable types (object types), checks whether they have the same identity value. (For most such types the identity is simply the reference itself.) It is not true that [1,2] === [1,2] because those are different Array objects, but it is true that @a === @a because those are the same Array object).

    Any object type may pretend to be a value type by defining a .WHICH method which returns a value type that can be recursively compared using ===, or in cases where that is impractical, by overloading === such that the comparison treats values consistently with their "eternal" identity. (Strings are defined as values this way despite also being objects.)

    Two values are never equivalent unless they are of exactly the same type. By contrast, eq always coerces to string, while == always coerces to numeric. In fact, $a eq $b really means "~$a === ~$b" and $a == $b means +$a === +$b.

    Note also that, while string-keyed hashes use eq semantics by default, object-keyed hashes use === semantics, and general value-keyed hashes use eqv semantics.

  • Binary eqv tests equality much like === does, but does so with "snapshot" semantics rather than "eternal" semantics. For top-level components of your value that are of immutable types, eqv is identical in behavior to ===. For components of your value that are mutable, however, rather than comparing object identity using ===, the eqv operator tests whether the canonical representation of both subvalues would be identical if we took a snapshot of them right now and compared those (now-immutable) snapshots using ===.

    If that's not enough flexibility, there is also an eqv() function that can be passed additional information specifying how you want canonical values to be generated before comparison. This gives eqv() the same kind of expressive power as a sort signature. (And indeed, the cmp operator from Perl 5 also has a functional analog, cmp(), that takes additional instructions on how to do 3-way comparisons of the kind that a sorting algorithm wants.) In particular, a signature passed to eqv() will be bound to the two operands in question, and then the comparison will proceed on the formal parameters according to the information contained in the signature, so you can force numeric, string, natural, or other comparisons with proper declarations of the parameter's type and traits. If the signature doesn't match the operands, eqv() reverts to standard eqv comparison. (Likewise for cmp().)

  • Binary cmp is no longer the comparison operator that forces stringification. Use the leg operator for the old Perl 5 cmp semantics. The cmp is just like the eqv above except that instead of returning Bool::False or Bool::True values it always returns Order::Increase, Order::Same, or Order::Decrease (which numerify to -1, 0, or +1).

  • The leg operator (less than, equal, or greater) is defined in terms of cmp, so $a leg $b is now defined as ~$a cmp ~$b. The sort operator still defaults to cmp rather than leg. The <=> operator's semantics are unchanged except that it returns an Order value as described above. In other words, $a <=> $b is now equivalent to +$a cmp +$b.

  • For boolean comparison operators with non-coercive cmp semantics, use the generic before and after infix operators. As ordinary infix operators these may be negated (!before and !after) as well as reduced ([before] and [after]).

  • Infix min and max may be used to select one or the other of their arguments. Reducing listop forms [min] and [max] are also available, as are the min= and max= assignment operators. By default min and max use cmp semantics. As with all cmp-based operators, this may be modified by an adverb specifying different semantics.

Range semantics

  • The .. range operator has variants with ^ on either end to indicate exclusion of that endpoint from the range. It always produces a Range object. Range objects are lazy iterators, and can be interrogated for their current .from and .to values (which change as they are iterated). The .minmax method returns both as a two-element list representing the interval. Ranges are not autoreversing: 2..1 is always a null range. Likewise, 1^..^2 produces no values when iterated, but does represent the interval from 1 to 2 excluding the endpoints when used as a pattern. To iterate a range in reverse use:

        2..1:by(-1)
        reverse 1..2

    (The reverse is preferred because it works for alphabetic ranges as well.) Note that, while .minmax normally returns (.from,.to), a negative :by causes the .minmax method returns (.to,.from) instead. You may also use .min and .max to produce the individual values of the .minmax pair, but again note that they are reversed from .from and .to when the step is negative. Since a reversed Range changes its direction, it swaps its .from and .to but not its .min and .max.

    Because Range objects are lazy, they do not automatically generate a list. One result of this is that a reversed Range object is still lazy. Another is that smart matching against a Range object smartmatches the endpoints in the domain of the object being matched, so fractional numbers are not truncated before comparison to integer ranges:

        1.5 ~~ 1^..^2  # true, equivalent to 1 < 1.5 < 2
        2.1 ~~ 1..2    # false, equivalent to 1 <= 2.1 <= 2

    If a * (see the "Whatever" type in S02) occurs on the right side of a range, it is taken to mean "positive infinity" in whatever typespace the range is operating, as inferred from the left operand. A * on the left means "negative infinity" for types that support negative values, and the first value in the typespace otherwise as inferred from the right operand. (For signed infinities the signs reverse for a negative step.) A star on both sides prevents any type from being inferred other than the Ordered role.

        0..*        # 0 .. +Inf
        'a'..*      # 'a' .. 'zzzzzzzzzzzzzzzzzzzzzzzzzzzzz...'
        *..0        # -Inf .. 0
        *..*        # "-Inf .. +Inf", really Ordered
        1.2.3..*    # Any version higher than 1.2.3.
        May..*      # May through December

    Note: infinite lists are constructed lazily. And even though *..* can't be constructed at all, it's still useful as a selector object.

    Range objects may be iterated on either end as long as it is not infinite. (Iterating an infinite end does not fail but just produces a lot of infinities.) Ordinary iteration iterates the .from value by adding the step. Either prefix:<=> or the shift function may be used to iterate the front of a range object. The pop function iterates the .to end by subtracting the step. In either case, the value returned is either the old value if the endpoint was inclusive, or the next value if the endpoint was exclusive. In the case of ranges that are not an integral multiple of the step, no check is done to see that iterating the front would produce the same list as interating from the back and reversing. So we have

        $range = 1..^42.5;
        $front = $range.shift;      # $front = 1, $range = 2..^42.5
        $back = $range.pop;      # $back = 41.5, $range = 2..^41.5

    For any kind of zip or dwimmy hyper operator, any list ending with * is assumed to be infinitely extensible by taking its final element and replicating it:

        @array, *

    is short for something like:

        @array[0..^@array], @array[*-1] xx *

    An empty Range cannot be iterated; it returns a Failure instead. An empty range still has a defined min and max, but the min is greater than the max.

  • The unary ^ operator generates a range from 0 up to one less than its argument. So ^4 is short for 0..^4 or 0..3.

        for ^4 { say $_ } # 0, 1, 2, 3

    If applied to a list, it generates a multidimensional set of subscripts.

        for ^(3,3) { ... } # (0,0)(0,1)(0,2)(1,0)(1,1)(1,2)(2,0)(2,1)(2,2)

    If applied to a type name, it indicates the metaclass instance instead, so ^Moose is short for HOW(Moose) or Moose.HOW. It still kinda means "what is this thing's domain" in an abstract sort of way.

  • Since use of Range objects in item context is usually non-sensical, a Range object used as an operand for scalar operators will generally attempt to distribute the operator to its endpoints and return another suitably modified Range instead. (Notable exceptions include infix:<~~>, which does smart matching, and prefix:<+> which returns the length of the range.) Therefore if you wish to write a slice using a length instead of an endpoint, you can say

        @foo[ start() + ^$len ]

    which is short for:

        @foo[ start() + (0..^$len) ]

    which is equivalent to something like:

        @foo[ list do { my $tmp = start(); $tmp ..^ $tmp+$len } ]

    In other words, operators of numeric and other ordered types are generally overloaded to do something sensible on Range objects. In particular, multiplicative operators not only multiply the endpoints but also the "by" of the Range object:

        (1..11:by(2)) * 5           # same as 5..55:by(10) 
        5,15,25,35,45,45,55

    Conjecture: non-linear functions might even produce non-uniform "by" values! Think of log scaling, for instance.

Chained comparisons

Perl 6 supports the natural extension to the comparison operators, allowing multiple operands:

    if 1 < $a < 100 { say "Good, you picked a number *between* 1 and 100." }

    if 3 < $roll <= 6              { print "High roll" }

    if 1 <= $roll1 == $roll2 <= 6  { print "Doubles!" }

A chain of comparisons short-circuits if the first comparison fails:

    1 > 2 > die("this is never reached");

Each argument in the chain will evaluate at most once:

    1 > $x++ > 2    # $x increments exactly once

Note: any operator beginning with < must have whitespace in front of it, or it will be interpreted as a hash subscript instead.

Smart matching

Here is the table of smart matches for standard Perl 6 (that is, the dialect of Perl in effect at the start of your compilation unit). Smart matching is generally done on the current "topic", that is, on $_. In the table below, $_ represents the left side of the ~~ operator, or the argument to a given, or to any other topicalizer. X represents the pattern to be matched against on the right side of ~~, or after a when.

The first section contains privileged syntax; if a match can be done via one of those entries, it will be. These special syntaxes are dispatched by their form rather than their type. Otherwise the rest of the table is used, and the match will be dispatched according to the normal method dispatch rules. The optimizer is allowed to assume that no additional match operators are defined after compile time, so if the pattern types are evident at compile time, the jump table can be optimized. However, the syntax of this part of the table is still somewhat privileged, insofar as the ~~ operator is one of the few operators in Perl that does not use multiple dispatch. Instead, type-based smart matches singly dispatch to an underlying method belonging to the X pattern object.

In other words, smart matches are dispatched first on the basis of the pattern's form or type (the X below), and then that pattern itself decides whether and how to pay attention to the type of the topic ($_). So the second column below is really the primary column. The Any entries in the first column indicate a pattern that either doesn't care about the type of the topic, or that picks that entry as a default because the more specific types listed above it didn't match.

    $_        X         Type of Match Implied   Match if (given $_)
    ======    =====     =====================   ===================
    Any       Code:($)  item sub truth          X($_)
    Any       Code:()   simple closure truth    X() (ignoring $_)
    Any       undef     undefined               not .defined
    Any       *         block signature match   block successfully binds to |$_
    Any       .foo      method truth            ?X       i.e. ?.foo
    Any       .foo(...) method truth            ?X       i.e. ?.foo
    Any       .(...)    sub call truth          ?X       i.e. ?.(...)
    Any       .[...]    array value slice truth ?all(X)  i.e. ?all(.[...])
    Any       .{...}    hash value slice truth  ?all(X)  i.e. ?all(.{...})
    Any       .<...>    hash value slice truth  ?all(X)  i.e. ?all(.<...>)

    Any       Bool      simple truth            X
    Any       Num       numeric equality        +$_ == X
    Any       Str       string equality         ~$_ eq X

    Any       Pair      test object             .:Xkey(Xval) (e.g. filetests)

    Set       Set       identical sets          $_ === X
    Hash      Set       hash keys same set      $_.keys === X
    Any       Set       force set comparison    Set($_) === X

    Array     Array     arrays are comparable   $_ «===» X (dwims * wildcards!)
    Set       Array     array equiv to set      $_ === Set(X)
    Any       Array     lists are comparable    @$_ «===» X

    Hash      Hash      hash keys same set      $_.keys === X.keys
    Set       Hash      hash keys same set      $_ === X.keys
    Array     Hash      hash slice existence    X.contains(any @$_)
    Regex     Hash      hash key grep           any(X.keys).match($_)
    Scalar    Hash      hash entry existence    X.contains($_)
    Any       Hash      hash slice existence    X.contains(any @$_)

    Str       Regex     string pattern match    .match(X)
    Hash      Regex     hash key "boolean grep" .any.match(X)
    Array     Regex     array "boolean grep"    .any.match(X)
    Any       Regex     pattern match           .match(X)

    Num       Range     in numeric range        X.min <= $_ <= X.max (mod ^'s)
    Str       Range     in string range         X.min le $_ le X.max (mod ^'s)
    Any       Range     in generic range        [!after] X.min,$_,X.max (etc.)

    Any       Type      type membership         $_.does(X)

    Signature Signature sig compatibility       $_ is a subset of X      ???
    Code      Signature sig compatibility       $_.sig is a subset of X  ???
    Capture   Signature parameters bindable     $_ could bind to X (doesn't!)
    Any       Signature parameters bindable     |$_ could bind to X (doesn't!)

    Signature Capture   parameters bindable     X could bind to $_

    Any       Any       scalars are identical   $_ === X

The final rule is applied only if no other pattern type claims X.

All smartmatch types are "itemized"; both ~~ and given/when provide item contexts to their arguments, and autothread any junctive matches so that the eventual dispatch to .ACCEPTS never sees anything "plural". So both $_ and X above are potentially container objects that are treated as scalars. (You may hyperize ~~ explicitly, though. In this case all smartmatching is done using the type-based dispatch to .ACCEPTS, not the form-based dispatch at the front of the table.)

The exact form of the underlying type-based method dispatch is:

    X.ACCEPTS($_)      # for ~~
    X.REJECTS($_)      # for !~~

As a single dispatch call this pays attention only to the type of X initially. The ACCEPTS method interface is defined by the Pattern role. Any class composing the Pattern role may choose to provide a single ACCEPTS method to handle everything, which corresponds to those pattern types that have only one entry with an Any on the left above. Or the class may choose to provide multiple ACCEPTS multi-methods within the class, and these will then redispatch within the class based on the type of $_. The class may also define one or more REJECTS methods; if it does not, the default REJECTS method from the Pattern role defines it in terms of a negated ACCEPTS method call. This generic method may be less efficient than a custom REJECTS method would be, however.

The smartmatch table is primarily intended to reflect forms and types that are recognized at compile time. To avoid an explosion of entries, the table assumes the following types will behave similarly:

    Actual type                 Use entries for
    ===========                 ===============
    List Seq                    Array
    KeySet KeyBag KeyHash       Hash
    Class Enum Role             Type
    Subst Grammar               Regex
    Char Cat                    Str
    Int UInt etc.               Num
    Match                       Capture
    Byte                        Str or Int
    Buf                         Str or Array of Int

(Note, however, that these mappings can be overridden by explicit definition of the appropriate ACCEPTS and REJECTS methods. If the redefinition occurs at compile time prior to analysis of the smart match then the information is also available to the optimizer.)

A Buf type containing any bytes or integers outside the ASCII range may silently promote to a Str type for pattern matching if and only if its relationship to Unicode is clearly declared or typed. This type information might come from an input filehandle, or the Buf role may be a parametric type that allows you to instantiate buffers with various known encodings. In the absence of such typing information, you may still do pattern matching against the buffer, but (apart from assuming the lowest 7 bits represent ASCII) any attempt to treat the buffer as other than a sequence integers is erroneous, and warnings may be generously issued.

Matching against a Grammar object will call the TOP method defined in the grammar. The TOP method may either be a rule itself, or may call the actual top rule automatically. How the Grammar determines the top rule is up to the grammar, but normal Perl 6 grammars will default to setting top to the first rule in the original base grammar. Derived grammars then inherit this idea of the top rule. This may be overridden in either the base grammar or a derived grammar by explicitly naming a rule TOP, or defining your own TOP method to call some other rule.

Matching against a Signature does not actually bind any variables, but only tests to see if the signature could bind. To really bind to a signature, use the * pattern to delegate binding to the when statement's block instead. Matching against * is special in that it takes its truth from whether the subsequent block is bound against the topic, so you can do ordered signature matching:

    given $capture {
        when * -> Int $a, Str $b { ... }
        when * -> Str $a, Int $b { ... }
        when * -> $a, $b         { ... }
        when *                   { ... }
    }

This can be useful when the unordered semantics of multiple dispatch are insufficient for defining the "pecking order" of code. Note that you can bind to either a bare block or a pointy block. Binding to a bare block conveniently leaves the topic in $_, so the final form above is equivalent to a default. (Placeholders parameters may also be used in the bare block form, though of course their types cannot be specified that way.)

There is no pattern matching defined for the Any pattern, so if you find yourself in the situation of wanting a reversed smartmatch test with an Any on the right, you can almost always get it by explicit call to the underlying ACCEPTS method using $_ as the pattern. For example:

    $_      X    Type of Match Wanted   What to use on the right
    ======  ===  ====================   ========================
    Code    Any  item sub truth         .ACCEPTS(X) or .(X)
    Range   Any  in range               .ACCEPTS(X)
    Type    Any  type membership        .ACCEPTS(X) or .does(X)
    Regex   Any  pattern match          .ACCEPTS(X)
    etc.

Similar tricks will allow you to bend the default matching rules for composite objects as long as you start with a dotted method on $_:

    given $somethingordered {
        when .values.'[<=]'     { say "increasing" }
        when .values.'[>=]'     { say "decreasing" }
    }

In a pinch you can define a macro to do the "reversed when":

    my macro statement_control:<ACCEPTS> () { "when .ACCEPTS: " }
    given $pattern {
        ACCEPTS $a      { ... }
        ACCEPTS $b      { ... }
        ACCEPTS $c      { ... }
    }

Various proposed-but-deprecated smartmatch behaviors may be easily (and we hope, more readably) emulated as follows:

    $_      X      Type of Match Wanted   What to use on the right
    ======  ===    ====================   ========================
    Array   Num    array element truth    .[X]
    Array   Num    array contains number  *,X,*
    Array   Str    array contains string  *,X,*
    Array   Seq    array begins with seq  X,*
    Array   Seq    array contains seq     *,X,*
    Array   Seq    array ends with seq    *,X
    Hash    Str    hash element truth     .{X}
    Hash    Str    hash key existence     .contains(X)
    Hash    Num    hash element truth     .{X}
    Hash    Num    hash key existence     .contains(X)
    Buf     Int    buffer contains int    .match(X)
    Str     Char   string contains char   .match(X)
    Str     Str    string contains string .match(X)
    Array   Scalar array contains item    .any === X
    Str     Array  array contains string  X.any
    Num     Array  array contains number  X.any
    Scalar  Array  array contains object  X.any
    Hash    Array  hash slice exists      .contains(X.all) .contains(X.any)
    Set     Set    subset relation        .contains(X)
    Set     Hash   subset relation        .contains(X)
    Any     Set    subset relation        .Set.contains(X)
    Any     Hash   subset relation        .Set.contains(X)
    Any     Set    superset relation      X.contains($_)
    Any     Hash   superset relation      X.contains($_)
    Any     Set    sets intersect         .contains(X.any)
    Set     Array  subset relation        X,*          # (conjectured)
    Array   Regex  match array as string  .Cat.match(X)  cat(@$_).match(X)

(Note that the .cat method and the Cat type coercion both take a single object, unlike the cat function which, as a list operator, takes a syntactic list (or multilist) and flattens it. All of these return a Cat object, however.)

Boolean expressions are those known to return a boolean value, such as comparisons, or the unary ? operator. They may reference $_ explicitly or implicitly. If they don't reference $_ at all, that's okay too--in that case you're just using the switch structure as a more readable alternative to a string of elsifs. Note, however, that this means you can't write:

    given $boolean {
        when True  {...}
        when False {...}
    }

because it will always choose the True case. Instead use something like:

    given $boolean {
        when .true {...}
        when .not  {...}
    }

Better, just use an if statement.

Note also that regex matching does not return a Bool, but merely a Match object that can be used as a boolean value. Use an explicit ? or true to force a Bool value if desired.

The primary use of the ~~ operator is to return a boolean value in a boolean context. However, for certain operands such as regular expressions, use of the operator within item or list context transfers the context to that operand, so that, for instance, a regular expression can return a list of matched substrings, as in Perl 5. This is done by returning an object that can return a list in list context, or that can return a boolean in a boolean context. In the case regex matching the Match object is a kind of Capture, which has these capabilities.

For the purpose of smartmatching, all Set and Bag values are considered to be of type KeyHash, that is, Hash containers where the keys represent the unique objects and the values represent the replication count of those unique keys. (Obviously, a Set can have only 0 or 1 replication because of the guarantee of uniqueness).

The Cat type allows you to have an infinitely extensible string. You can match an array or iterator by feeding it to a Cat, which is essentially a Str interface over an iterator of some sort. Then a Regex can be used against it as if it were an ordinary string. The Regex engine can ask the string if it has more characters, and the string will extend itself if possible from its underlying interator. (Note that such strings have an indefinite number of characters, so if you use .* in your pattern, or if you ask the string how many characters it has in it, or if you even print the whole string, it may be feel compelled to slurp in the rest of the string, which may or may not be expeditious.)

The cat operator takes a (potentially lazy) list and returns a Cat object. In string context this coerces each of its elements to strings lazily, and behaves as a string of indeterminate length. You can search a gather like this:

    my $lazystr := cat gather for @foo { take .bar }

    $lazystr ~~ /pattern/;

The Cat interface allows the regex to match element boundaries with the <,> assertion, and the StrPos objects returned by the match can be broken down into elements index and position within that list element. If the underlying data structure is a mutable array, changes to the array (such as by shift or pop) are tracked by the Cat so that the element numbers remain correct. Strings, arrays, lists, sequences, captures, and tree nodes can all be pattern matched by regexes or by signatures more or less interchangably. However, the structure searched is not guaranteed to maintain a .pos unless you are searching a Str or Cat.

Invocant marker

An appended : marks the invocant when using the indirect-object syntax for Perl 6 method calls. The following two statements are equivalent:

    $hacker.feed('Pizza and cola');
    feed $hacker: 'Pizza and cola';

A colon may also be used on an ordinary method call to indicate that it should be parsed as a list operator:

    $hacker.feed: 'Pizza and cola';

This colon is a separate token. A colon prefixing an adverb is not a separate token. Therefore, under the longest-token rule,

    $hacker.feed:xxx('Pizza and cola');

is tokenized as an adverb applying to the method as its "preceding operator":

    $hacker.feed :xxx('Pizza and cola');

not as an xxx sub in the argument list of .feed:

    $hacker.feed: xxx('Pizza and cola');  # wrong

If you want both meanings of colon in order to supply both an adverb and some positional arguments, you have to put the colon twice:

    $hacker.feed: :xxx('Pizza and cola'), 1,2,3;

(For similar reasons it's required to put whitespace after the colon of a label.)

Feed operators

The new operators ==> and <== are akin to UNIX pipes, but work with functions or statements that accept and return lists. Since these lists are composed of discrete objects and not liquids, we call these feed operators rather than pipes. For example,

     @result = map { floor($^x / 2) },
                 grep { /^ \d+ $/ },
                   @data;

can also now be written with rightward feeds as:

     @data ==> grep { /^ \d+ $/ }
           ==> map { floor($^x / 2) }
           ==> @result;

or with leftward feeds as:

     @result <== map { floor($^x / 2) }
             <== grep { /^ \d+ $/ }
             <== @data;

Either form more clearly indicates the flow of data. See S06 for more of the (less-than-obvious) details on these two operators.

Meta operators

Perl 6's operators have been greatly regularized, for instance, by consistently prefixing numeric, stringwise, and boolean operators with +, ~ and ? respectively to indicate whether the bitwise operation is done on a number, a string, or a single bit. But that's just a naming convention, and if you wanted to add a new bitwise ¬ operator, you'd have to add the , , and operators yourself. Similarly, the carets that exclude the endpoints on ranges are there by convention only.

In contrast to that, Perl 6 has five standard metaoperators for turning a given existing operator into a related operator that is more powerful (or at least differently powerful). These differ from a mere naming convention in that Perl automatically generates these new metaoperators from user-defined operators as well as from builtins. In fact, you're not generally supposed to define the individual metaoperations--their semantics are supposed to be self-evident by the transformation of the base operator.

Note: Spaces are never allowed between any metaoperator and the operator it's modifying, because all operators including modified ones have to be recognized by the Longest-Token Rule, which disallows spaces within a token.

Assignment operators

These are already familiar to C and Perl programmers. (Though the .= operator now means to call a mutating method on the object on the left, and ~= is string concatenation.) Most non-relational infix operators may be turned into their corresponding assignment operator by suffixing with =. The limitation is actually based on whether the left side can function both as an rvalue and an lvalue by the usual correspondence:

    A op= B;
    A = A op B;

Existing forms ending in = may not be modified with this metaoperator.

Regardless of the precedence of the base operator, the precedence of any assignment operators is forced to be the same as that of ordinary assignment.

Negated relational operators

Any infix relational operator may be transformed into its negative by prefixing with !. A couple of these have traditional shortcuts:

    Full form   Shortcut
    ---------   --------
    !==         !=
    !eq         ne

but most of them do not:

    !~~
    !<
    !>=
    !ge
    !===
    !eqv
    !=:=

To avoid visual confusion with the !! operator, you may not modify any operator already beginning with !.

The precedence of any negated operator is the same as the base operator.

Hyper operators

The Unicode characters » (\x[BB]) and « (\x[AB]) and their ASCII digraphs >> and << are used to denote a "list operation" that operates on each element of its list (or array) argument (or arguments) and returns a single list (or array) of the results. In other words, a hyper operator evaluates its arguments in item context but then distributes the operator over them as lists.

When writing a hyper operator, spaces are not allowed on the inside, that is, between any "hyper" marker and the operator it's modifying. On the outside the spacing policy is the same as the base operator. Likewise the precedence of any hyperoperator is the same as its base operator. This means that you must parenthesize your comma lists for most operators. For example:

     -« (1,2,3);                   # (-1, -2, -3)
     (1,1,2,3,5) »+« (1,2,3,5,8);  # (2,3,5,8,13)

A unary hyper operator (either prefix or postfix) has only one hyper marker, located on its argument side, while an infix operator always has one on each side to indicate there are two arguments. Unary operators always produce a list or array of exactly the same shape as their single argument. When infix operators are presented with two lists or arrays of identical shape, a result of that same shape is produced. Otherwise the result depends on how you write the hyper markers.

For an infix operator, if either argument is insufficiently dimensioned, Perl "upgrades" it, but only if you point the "sharp" end of the hypermarker at it.

     (3,8,2,9,3,8) >>->> 1;          # (2,7,1,8,2,7)
     @array »+=» 42;                 # add 42 to each element

In fact, an upgraded scalar is the only thing that will work for an unordered type such as a Bag:

     Bag(3,8,2,9,3,8) >>->> 1;       # Bag(2,7,1,8,2,7) === Bag(1,2,2,7,7,8)

In other words, pointing the small end at an argument tells the hyperoperator to "dwim" on that side. If you don't know whether one side or the other will be underdimensioned, you can dwim on both sides:

    $left «*» $right

[Note: if you are worried about Perl getting confused by something like this:

    foo «*»

then you shouldn't worry about it, because unlike previous versions, Perl 6 never guesses whether the next thing is a term or operator. In this case it is always expecting a term unless foo is predeclared to be a 0-ary sub.]

The upgrade never happens on the "blunt" end of a hyper. If you write

    $bigger «*« $smaller
    $smaller »*» $bigger

an exception is thrown, and if you write

    $foo »*« $bar

you are requiring the shapes to be identical, or an exception will be thrown. By default this dwimmery only upgrades whole dimensions, not short lists. However, any list ending with * can also be arbitrarily extended as if the last element of the list were arbitrarily replicated * times. But this happens only on the "dwimmy" side.

On the non-dwimmy side, any scalar value that does not know how to do List will be treated as a list of one element, and for infix operators must be matched by an equivalent one-element list on the other side. That is, a hyper operator is guaranteed to degenerate to the corresponding scalar operation when all its arguments are non-list arguments.

When using a unary operator, you always aim the blunt end at the single operand, because no dwimmery ever happens:

     @negatives = -« @positives;

     @positions»++;            # Increment all positions

     @positions.»++;           # Same thing, dot form
     @positions».++;           # Same thing, dot form
     @positions.».++;          # Same thing, dot form
     @positions\  .»\  .++;    # Same thing, unspace form

     @objects.».run();
     ("f","oo","bar").>>.chars;   # (1,2,3)

Note that method calls are really postfix operators, not infix, so you shouldn't put a « after the dot.

Hyper operators are defined recursively on shaped arrays, so:

    -« [[1, 2], 3]               #    [-«[1, 2], -«3]
                                 # == [[-1, -2], -3]

Likewise the dwimminess of dwimmy infixes propagates:

    [[1, 2], 3] «+» [4, [5, 6]]  #    [[1,2] «+» 4, 3 «+» [5, 6]]
                                 # == [[5, 6], [8, 9]]

More generally, a dwimmy hyper operator works recursively for any object matching the Each role even if the object itself doesn't support the operator in question:

    Bag(3,8,[2,Seq(9,3)],8) >>->> 1;         # Bag(2,7,[1,Seq(8,2)],7)
    Seq(3,8,[2,Seq(9,3)],8) >>->> (1,1,2,1); # Seq(2,7,[0,Seq(7,1)],7)

In particular, tree node types with Each semantics enable visitation:

    $tree.».foo;        # short for $tree.foo, $tree.each: { .».foo }

If not all nodes support the operation, you need a form of it that specifies the call is optional:

    $tree.».?foo;       # short for $tree.?foo, $tree.each: { .».?foo }
    $tree.».*foo;       # short for $tree.*foo, $tree.each: { .».*foo }

You are not allowed to define your own hyper operators, because they are supposed to have consistent semantics derivable entirely from the modified scalar operator. If you're looking for a mathematical vector product, this isn't where you'll find it. A hyperoperator is one of the ways that you can promise to the optimizer that your code is parallelizable. (The tree visitation above is allowed to have side effects, but it is erroneous for the meaning of those side effects to depend on the order of visitation in any way. Hyper tree visitation is not required to follow DAG semantics, at least by default.)

Even in the absence of hardware that can do parallel processing, hyperoperators may be faster than the corresponding scalar operators if they can factor out looping overhead to lower-level code, or can apply loop-unrolling optimizations, or can factor out some or all of the MMD dispatch overhead, based on the known types of the operands (and also based on the fact that hyper operators promise no interaction among the "iterations", whereas the corresponding scalar operator in a loop cannot make the same promise unless all the operations within the loop are known to be side-effect free.)

In particular, infix hyperops on two int or num arrays need only do a single MMD dispatch to find the correct function to call for all pairs, and can further bypass any type-checking or type-coercion entry points to such functions when there are known to be low-level entry points of the appropriate type. (And similarly for unary int or num ops.)

Application-wide analysis of finalizable object types may also enable such optimizations to be applied to Int, Num, and such. In the absence of that, run-time analysis of partial MMD dispatch may save some MMD searching overhead. Or particular object arrays might even keep track of their own run-time type purity and cache partial MMD dispatch tables when they know they're likely to be used in hyperops.

Beyond all that, "array of scalar" types are known at compile time not to need recursive hypers, so the operations can be vectorized aggressively.

Hypers may be applied to hashes as well as to lists. In this case "dwimminess" says whether to ignore keys that do not exist in the other hash, while "non-dwimminess" says to use all keys that are in either hash. That is,

    %foo «+» %bar;

gives you the intersection of the keys, while

    %foo »+« %bar;

gives you the union of the keys. Asymmetrical hypers are also useful; for instance, if you say:

    %outer »+» %inner;

only the %inner keys that already exist in %outer will occur in the result. Note, however, that you want

    %outer »+=« %inner;

in order to pass accumulated statistics up a tree, assuming you want %outer to have the union of keys.

Unary hash hypers and binary hypers that have only one hash operand will apply the hyper operator to just the values but return a new hash value with the same set of keys as the original hash.

Reduction operators

The fourth metaoperator in Perl 6 is the reduction operator. Any infix operator (except for non-associating operators) can be surrounded by square brackets in term position to create a list operator that reduces using that operation:

    [+] 1, 2, 3;      # 1 + 2 + 3 = 6
    my @a = (5,6);
    [*] @a;           # 5 * 6 = 30

As with all the metaoperators, space is not allowed inside. The whole thing parses as a single token.

A reduction operator has the same precedence as a list operator. In fact, a reduction operator really is a list operator, and is invoked as one. Hence, you can implement a reduction operator in one of two ways. Either you can write an explicit list operator:

    proto prefix:<[+]> (*@args) {
        my $accum = 0;
        while (@args) {
            $accum += @args.shift();
        }
        return $accum;
    }

or you can let the system autogenerate one for you based on the corresponding infix operator, probably by currying:

    # (examples, actual system may define prefix:[**] instead)
    &prefix:<[*]> ::= &reduce.assuming(&infix:<*>, 1);
    &prefix:<[**]> ::= &reducerev.assuming(&infix:<**>);

As a special form of name, the non-prefix notation, as in

    proto [foo] (*@args) {
        ...
    }

or

    &[foo] ::= ...

defines both the [foo] reduce operator and the foo infix operator. Where appropriate, use of the infix form may be optimized like this:

    # Original          # Optimized
    $a foo $b           # [foo] $a, $b
    $a foo $b foo $c    # [foo] $a, $b, $c

(Note that any infix generated with ::= will default to the precedence of +. When that is not desired you must use is equiv. See S06.)

If the reduction operator is defined separately from the infix operator, it must associate the same way as the operator used:

    [-] 4, 3, 2;      # 4-3-2 = (4-3)-2 = -1
    [**] 4, 3, 2;     # 4**3**2 = 4**(3**2) = 262144

For list-associating operators (like <), all arguments are taken together, just as if you had written it out explicitly:

    [<] 1, 3, 5;      # 1 < 3 < 5

If fewer than two arguments are given, a dispatch is still attempted with whatever arguments are given, and it is up to the receiver of that dispatch to deal with fewer than two arguments. Note that the proto list operator definition is the most general, so you are allowed to define different ways to handle the one argument case depending on type:

    multi prefix:<[foo]> (Int $x) { 42 }
    multi prefix:<[foo]> (Str $x) { fail "Can't foo a single Str" }

However, the zero argument case cannot be defined this way, since there is no type information to dispatch on. Operators that wish to specify an identity value should do so by specifying a multi variant that takes zero arguments:

    multi prefix:<[foo]> () { 0 }

Among the builtin operators, [+]() returns 0 and [*]() returns 1, for instance.

By default, if there is one argument, the built-in reduce operators return that one argument. However, this default doesn't make sense for operators like < that don't return the same type as they take, so these kinds of operators overload the single-argument case to return something more meaningful. All the comparison operators return a boolean for either 1 or 0 arguments. Negated operators return Bool::False, and all the rest return Bool::True.

You can also make a reduce operator of the comma operator. This is just the list operator form of the circumfix:<[ ]> anonymous array composer:

    [1,2,3]     # make new Array: 1,2,3
    [,] 1,2,3   # same thing

You may also reduce using the semicolon second-dimension separator:

    [[;] 1,2,3]   # equivalent to [1;2;3]

Builtin reduce operators return the following identity values:

    [**]()      # 1     (arguably nonsensical)
    [*]()       # 1
    [/]()       # fail  (reduce is nonsensical)
    [%]()       # fail  (reduce is nonsensical)
    [x]()       # fail  (reduce is nonsensical)
    [xx]()      # fail  (reduce is nonsensical)
    [+&]()      # -1    (from +^0, the 2's complement in arbitrary precision)
    [+<]()      # fail  (reduce is nonsensical)
    [+>]()      # fail  (reduce is nonsensical)
    [~&]()      # fail  (sensical but 1's length indeterminate)
    [~<]()      # fail  (reduce is nonsensical)
    [~>]()      # fail  (reduce is nonsensical)
    [+]()       # 0
    [-]()       # 0
    [~]()       # ''
    [+|]()      # 0
    [+^]()      # 0
    [~|]()      # ''    (length indeterminate but 0's default)
    [~^]()      # ''    (length indeterminate but 0's default)
    [&]()       # all()
    [|]()       # any()
    [^]()       # one()
    [!==]()     # Bool::False   (also for 1 arg)
    [==]()      # Bool::True    (also for 1 arg)
    [before]()  # Bool::True    (also for 1 arg)
    [after]()   # Bool::True    (also for 1 arg)
    [<]()       # Bool::True    (also for 1 arg)
    [<=]()      # Bool::True    (also for 1 arg)
    [>]()       # Bool::True    (also for 1 arg)
    [>=]()      # Bool::True    (also for 1 arg)
    [~~]()      # Bool::True    (also for 1 arg)
    [!~~]()     # Bool::False   (also for 1 arg)
    [eq]()      # Bool::True    (also for 1 arg)
    [!eq]()     # Bool::False   (also for 1 arg)
    [lt]()      # Bool::True    (also for 1 arg)
    [le]()      # Bool::True    (also for 1 arg)
    [gt]()      # Bool::True    (also for 1 arg)
    [ge]()      # Bool::True    (also for 1 arg)
    [=:=]()     # Bool::True    (also for 1 arg)
    [!=:=]()    # Bool::False   (also for 1 arg)
    [===]()     # Bool::True    (also for 1 arg)
    [!===]()    # Bool::False   (also for 1 arg)
    [eqv]()     # Bool::True    (also for 1 arg)
    [!eqv]()    # Bool::False   (also for 1 arg)
    [&&]()      # Bool::True
    [||]()      # Bool::False
    [^^]()      # Bool::False
    [//]()      # undef
    [min]()     # +Inf
    [max]()     # -Inf
    [=]()       # undef    (same for all assignment operators)
    [,]()       # []
    [Z]()       # []

User-defined operators may define their own identity values, but there is no explicit identity property. The value is implicit in the behavior of the 0-arg reduce, so mathematical code wishing to find the identity value for an operation can call prefix:{"[$opname]"}() to discover it.

To call some other non-infix function as a reduce operator, you may define an alias in infix form. The infix form will parse the right argument as an item even if the aliased function would have parsed it as a list:

    &infix:<dehash> ::= postcircumfix:<{ }>;
    $x = [dehash] $a,'foo','bar';  # $a<foo><bar>, not $a<foo bar>

Alternately, just define your own prefix:<[dehash]> routine.

Note that, because a reduce is a list operator, the argument list is evaluated in list context. Therefore the following would be incorrect:

    $x = [dehash] %a,'foo','bar';

You'd instead have to say one of:

    $x = [dehash] \%a,'foo','bar';
    $x = [dehash] %a<foo>,'bar';

On the plus side, this works without a star:

    @args = (\%a,'foo','bar');
    $x = [dehash] @args;

Likewise, from the fact that list context flattens inner arrays and lists, it follows that a reduced assignment does no special syntactic dwimmery, and hence only scalar assignments are supported. Therefore

    [=] $x, @y, $z, 0
    [+=] $x, @y, $z, 1

are equivalent to

    $x = @y[0] = @y[1] = @y[2] ... @y[*-1] = $z = 0
    $x += @y[0] += @y[1] += @y[2] ... @y[*-1] += $z += 1

rather than

    $x = @y = $z = 0;
    $x += @y += $z += 1;

(And, in fact, the latter are already easy to express anyway, and more obviously nonsensical.)

Most reduce operators return a simple scalar value, and hence do not care whether they are evaluated in item or list context. However, as with other list operators and functions, a reduce operator may return a list that will automatically be interpolated into list context, so you may use it on infix operators that operate over lists as well as scalars:

    my ($min, $max) = [minmax] @minmaxpairs;

A variant of the reduction metaoperator is pretty much guaranteed to produce a list; to lazily generate all intermediate results along with the final result, you can backslash the operator:

    say [\+] 1..*  #  (1, 3, 6, 10, 15, ...)

The visual picture of a triangle is not accidental. To produce a triangular list of lists, you can use a "triangular comma":

    [\,] 1..5
    [1],
    [1,2],
    [1,2,3],
    [1,2,3,4],
    [1,2,3,4,5]

If there is ambiguity between a triangular reduce and an infix operator beginning with backslash, the infix operator is chosen, and an extra backslash indicates the corresponding triangular reduce. As a consequence, defining an infix operator beginning with backslash, infix:<\x> say, will make it impossible to write certain triangular reduction operators, since [\x] would mean the normal reduction of infix:<\x> operator, not the triangular reduction of infix:<x>. This is deemed to be an insignificant problem.

Cross operators

The final metaoperator is the cross metaoperator. It is formed syntactically by placing an infix operator between two X characters. It applies the modified operator across all groupings of its list arguments as returned by the ordinary infix:<X> operator. All cross operators are of list infix precedence, and are list associative.

The string concatenating form is:

    <a b> X~X <1 2>           #  'a1', 'a2', 'b1', 'b2'

The X~X operator desugars to something like:

    [~]«( <a b> X,X <1 2> )  #  'a1', 'a2', 'b1', 'b2'

The list concatenating form, X,X, when used like this:

    <a b> X,X 1,2 X,X <x y>

produces

    ('a', 1, 'x'),
    ('a', 1, 'y'),
    ('a', 2, 'x'),
    ('a', 2, 'y'),
    ('b', 1, 'x'),
    ('b', 1, 'y'),
    ('b', 2, 'x'),
    ('b', 2, 'y')

The list form is common enough to have a shortcut, the ordinary infix X operator described earlier.

For the general form, any existing, non-mutating infix operator may be used.

    1,2 X*X 3,4               # 3,4,6,8

(Note that <== and ==> are considered mutating, as well as all assignment operators.)

If the underlying operator is non-associating, so is the metaoperator:

    @a XcmpX @b XcmpX @c       # ILLEGAL
    @a XeqX @b XeqX @c         # ok

In fact, though the X operators are all list associative syntactically, the underlying operator is always applied with its own associativity, just as the corresponding reduce operator would do.

Note that only the first term of an X operator may reasonably be an infinite list.

Multidimensional lists should be handled properly.

Nesting of metaoperators

In order to match operators by the longest-token rule, the compiler pregenerates various metaforms based on existing operators. Unfortunately, with nesting metaoperators there are an infinite number of metaforms, so we arbitrarily say that no metacircumfix form is pregenerated that uses the same grammatical category more than once. Therefore forms like [+=] and »!===« and X*X= are generated, but not forms like »X*X« or X«*»X or <<+=>>=. You do get [X*X], though, because reduction is prefix_circumfix_meta_operator while cross operators are infix_circumfix_meta_operator.

This use-each-category-once limitation is not a great hardship since you can define your own infix operators. Suppose you say

    &infix:<xp> ::= &infix:<X*X>;

After this you can use XxpX, [xp], and [«xp»=]« as if xp were a built-in. Not that any of those necessarily make sense...

The compiler is not actually required to pregenerate the metaform tokens as long as it can guarantee the same semantics, that is, that it follows the longest-token rule across all syntax categories active at that spot in the parse. This could be achieved by use of a DFA parser (or exhaustive NFA matcher) to guarantee longest match of the generatable forms, for instance, followed by a check to make sure it is not trumped by an even longer "hardwired" token. Suppose the user were to define, say, infix:<XxX!#@$> or statement_modifier:<XxXfor>; those hardwired forms must take precedence over the XxX operator even if the metaform DFA only knows how to recognize the XxX part.

Note that the maximum possible token length is bounded by the sum of the maximum lengths of each metacategory, so an enumerated solution is always possible, though perhaps not practical. It's also the case that no metaform is allowed to contain whitespace, so a solution that starts at the length of the current "word" and works down is also possible.

Declarators

The list of variable declarators has expanded from my and our to include:

    my $foo             # ordinary lexically scoped variable
    our $foo            # lexically scoped alias to package variable
    has $foo            # object attribute
    state $foo          # persistent lexical (cloned with closures)
    constant $foo       # lexically scoped compile-time constant

Variable declarators such as my now take a signature as their argument. (The syntax of function signatures is described more fully in S06.)

The parentheses around the signature may be omitted for a simple declaration that declares a single variable, along with its associated type, traits and the initializer:

    constant $foo = 123;    # okay: initializes $foo to 123
    constant ($foo = 123);  # same thing (with explicit parens)
    constant :($foo = 123); # same thing (full Signature form)
    constant ($foo) = 123;  # wrong: constants cannot be assigned to

List-context assignment is supported for simple declarations:

    constant @foo = 1,2,3;      # okay: initializes @foo to (1,2,3)
    constant (@foo = 1,2,3);    # wrong: 2 and 3 are not variable names

When parentheses are omitted, you may use an infix assignment operator instead of = as the initializer. In that case, the left hand side of the infix operator will be the variable's prototype object:

    constant Dog $fido .= new;      # okay: a constant Dog object
    constant Dog $fido = Dog.new;   # same thing
    constant Dog $fido = $fido.new; # wrong: invalid self-reference
    constant (Dog $fido .= new);    # wrong: cannot use .= with parens

Parentheses must always be used when declaring multiple parameters:

    my $a;                  # okay
    my ($b, $c);            # okay
    my ($b = 1, $c = 2);    # okay - "my" intializers assign at runtime
    my $b, $c;              # wrong: "Use of undeclared variable: $c"

Types occurring between the declarator and the signature are distributed into each variable:

    my Dog ($b, $c);
    my (Dog $b, Dog $c);    # same thing

[XXX the following probably belongs in S06.] The syntax for constructing a Signature object when the parser isn't already expecting one is:

    :(Dog $a, *@c)

This might be used like this:

    my $sig = :(Dog $a, *@c);

Signatures are expected after declarators such as my, sub, method, rule, etc. In such declarators the colon may be omitted. But it's also legal to use it:

    my :($b, $c);               # okay
    sub foo :($a,$b) {...}      # okay

The -> "pointy block" token also introduces a signature, but in this case you must omit both the colon and the parens. For instance, if you're defining the "loop variable" of a loop block:

    for @dogpound -> Dog $fido { ... }

If a signature is assigned to (whether declared or colon form), the signature is converted to a list of lvalue variables and the ordinary rules of assignment apply, except that the evaluation of the right side and the assignment happens at time determined by the declarator. (With my this is always when an ordinary assignment would happen.) If the signature is too complicated to convert to an assignment, a compile-time error occurs. Assignment to a signature makes the same item/list distinction as ordinary assignment, so

    my $a = foo();      # foo in item context
    my ($a) = foo();    # foo in list context

If a signature is bound to an argument list, then the binding of the arguments proceeds as if the signature were the formal parameters for a function, except that, unlike in a function call, the parameters are bound rw by default rather than readonly. See Binding above.

Note that temp and let are not variable declarators, because their effects only take place at runtime. Therefore, they take an ordinary lvalue object as their argument. See S04 for more details.

There are a number of other declarators that are not variable declarators. These include both type declarators:

    package Foo
    module Foo
    class Foo
    role Foo
    subset Foo

and code declarators:

    sub foo
    method foo
    submethod foo
    multi foo
    proto foo
    macro foo
    quote qX
    regex foo
    rule foo
    token foo

These all have their uses and are explained in subsequent Synopses.

Argument List Interpolating

Perl 5 forced interpolation of a function's argument list by use of the & prefix. That option is no longer available in Perl 6, so instead the | prefix operator serves as an interpolator, by casting its operand to a Capture object and inserting the capture's parts into the current argument list. This operator can be used to interpolate an Array or Hash into the current call, as positional and named arguments respectively.

Note that the resulting arguments still must comply with the subroutine's signature, but the presence of | defers that test until run time for that argument (and for any subsequent arguments):

    my @args := [item @foo, @bar];
    push |@args;

is equivalent to:

    push @foo: @bar;

which is just another way to write:

    @foo.push(@bar);

Unlike [,], | does not flatten its argument, but instead directly converts its argument into a Capture:

    my @args = \$x, 1, 2, 3;
    say [,] @args;  # say($x: 1, 2, 3);
    say |@args;     # say(\$x, 1, 2, 3);

Because of this, |%args always produces named arguments, and |@args always produces positional arguments.

In list context, a Scalar holding an Array object does not flatten. Hence

    $bar = @bar;
    @foo.push($bar);

merely pushes a single Array object onto @foo. You can explicitly flatten it in one of these ways:

    @foo.push(@$bar);
    @foo.push($bar[]);
    @foo.push(|$bar);

Those three forms work because the slurpy array in push's signature flattens the Array object into a list argument.

Note that the first two forms also allow you to specify list context on assignment:

    @$bar = 1,2,3;
    $bar[] = 1,2,3;

For long lvalue expressions, the second form can keep the "arrayness" of the lvalue close to the assignment operator:

    $foo.bar.baz.bletch.whatever.attr[] = 1,2,3;

Otherwise you'd have to write:

    @($foo.bar.baz.bletch.whatever.attr) = 1,2,3;

and remember the @ at the front until you get to the =.

The empty [] and .[] postfix operators are interpreted as a zero-dimensional slice returning the entire array, not as a one-dimensional null slice returning no elements. Likewise for {} and .{} on hashes, as well as the <>, .<>, «», and .«» constant and interpolating slice subscripting forms.

The | operator interpolates lazily for Array and Range objects. To get an immediate interpolation like Perl 5 does, add the eager list operator:

    func(|(1..Inf));       # works fine
    func(|eager 1..Inf);   # never terminates (well, actually...)

To interpolate a function's return value, you can say:

    push |func();

Within such an argument list, function return values are automatically exploded into their various parts, as if you'd said:

    my $capture = \(func());
    push $$capture: @$capture, %$capture;

or some such. The | then handles the various zones appropriately depending on the context. An invocant only makes sense as the first argument to the outer function call. An invocant inserted anywhere else just becomes a positional argument at the front of its list, as if its colon changed back to a comma.

If you already have a capture variable, you can interpolate all of its bits at once using the prefix:<|> operator:

    my (|$capture) := func();
    push |$capture;

Traversing arrays in parallel

In order to support parallel iteration over multiple arrays, Perl 6 has a zip function that builds a list of Seq objects from the elements of two or more arrays. In ordinary list context this behaves as a list of Captures and automatically flattens.

    for zip(@names; @codes) -> $name, $zip {
        print "Name: $name;   Zip code: $zip\n";
    }

zip has an infix synonym, the Z operator.

In an explicitly multidimensional list context, however, the sequences turn into subarrays, and each element would then have to be unpacked by the signature:

    for @@(zip(@names; @codes)) -> [$name, $zip] {
        print "Name: $name;   Zip code: $zip\n";
    }

By default the zip function reads to the end of the shortest list, but a short list may always be extended arbitrarily by putting * after the final value, which replicates the final value as many times as necessary. If instead of supplying a default value for short lists, you just wish to skip missing entries, use roundrobin instead:

    for roundrobin(@queue1; @queue2; @queue3) -> $next {
        ...
    }

To read arrays serially rather than in parallel, use list(@x;@y). This wins a "useless use of list award" in this case since you could always just write (@x,@y) to mean the same thing. But sometimes it's nice to be explicit about that:

    @foo := [[1,2,3],[4,5,6]]; say list([;] @foo); # 1,2,3,4,5,6

Minimal whitespace DWIMmery

Whitespace is no longer allowed before the opening bracket of an array or hash subscript, or the opening parenthesis of an argument list. That is:

    @deadbeef[$x]         # okay
    @a       [$b]         # WRONG
    %monsters{'cookie'}   # okay
    %people  {'john'}     # WRONG
    saymewant('cookie')   # okay
    mewant   ('cookie')   # WRONG

One of the several useful side-effects of this restriction is that parentheses are no longer required around the condition of control constructs:

    if $value eq $target {
        print "Bullseye!";
    }
    while $i < 10 { $i++ }

It is, however, still possible to align subscripts and other postfix operators by explicitly using the unspace syntax (see S02):

     %monsters.{'cookie'} = Monster.new;
     %beatles\.{'ringo'}  = Beatle.new;
     %people\ .{'john'}   = Person.new;
     %cats\   .{'fluffy'} = Cat.new;

Whitespace is in general required between any keyword and any opening bracket that is not introducing a subscript or function arguments. Any keyword followed directly by parentheses will be taken as a function call instead.

    if $a == 1 { say "yes" }            # preferred syntax
    if ($a == 1) { say "yes" }          # P5-ish if construct
    if($a,$b,$c)                        # if function call

It is possible for if() to also invoke a macro call, but if so, it's a prefix:<if> macro rather than a statement_control:<if> macro.

Sequence points

Certain operators are guaranteed to provide sequence points. Sequence points are guaranteed whenever some thunk (a lazy chunk of code) is conditionally evaluated based on the result of some other evaluation, so the short-circuit and conditional operators all provide sequence points.

Certain other operators guarantee the absence of sequence points, including junctional operators, hyperoperators, and feed operators. These operators promise the compiler that you consider the bits of code not to be dependent on each other so that they can operate in parallel if they like.

A large number of operators (such as +) are stuck in the middle, and may exhibit sequential behavior today, but might not tomorrow. A program that relies on either sequential or parallel behavior for one of these operators is erroneous. As we get more feedback from people writing parallelizing optimizers, we reserve the right to classify various of the unclassified operators into one of the two specified sets. (We need to give these three sets of operators good names.)