If you spot a problem with this page, click here to create a Bugzilla issue.

Quickly fork, edit online, and submit a pull request for this page. Requires a signed-in GitHub account. This works well for small changes. If you'd like to make larger changes you may want to consider using a local clone.

std.typecons

This module implements a variety of type constructors, i.e., templates that allow construction of new, useful general-purpose types.


Category	Symbols
Tuple	isTuple Tuple tuple reverse
Flags	BitFlags isBitFlagEnum Flag No Yes
Memory allocation	SafeRefCounted safeRefCounted RefCountedAutoInitialize scoped Unique
Code generation	AutoImplement BlackHole generateAssertTrap generateEmptyFunction WhiteHole
Nullable	Nullable nullable NullableRef nullableRef
Proxies	Proxy rebindable Rebindable ReplaceType unwrap wrap
Types	alignForSize Ternary Typedef TypedefType UnqualRef

License:

Boost License 1.0.

Source std/typecons.d

Authors:

Andrei Alexandrescu, Bartosz Milewski, Don Clugston, Shin Fujishiro, Kenji Hara

Examples:

Value tuples

alias Coord = Tuple!(int, "x", int, "y", int, "z");
Coord c;
c[1] = 1;       // access by index
c.z = 1;        // access by given name
writeln(c); // Coord(0, 1, 1)

// names can be omitted, types can be mixed
alias DictEntry = Tuple!(string, int);
auto dict = DictEntry("seven", 7);

// element types can be inferred
writeln(tuple(2, 3, 4)[1]); // 3
// type inference works with names too
auto tup = tuple!("x", "y", "z")(2, 3, 4);
writeln(tup.y); // 3

Examples:

Rebindable references to const and immutable objects

class Widget
{
    void foo() const @safe {}
}
const w1 = new Widget, w2 = new Widget;
w1.foo();
// w1 = w2 would not work; can't rebind const object

auto r = Rebindable!(const Widget)(w1);
// invoke method as if r were a Widget object
r.foo();
// rebind r to refer to another object
r = w2;

struct Unique(T);

Encapsulates unique ownership of a resource.

When a Unique!T goes out of scope it will call destroy on the resource T that it manages, unless it is transferred. One important consequence of destroy is that it will call the destructor of the resource T. GC-managed references are not guaranteed to be valid during a destructor call, but other members of T, such as file handles or pointers to malloc memory, will still be valid during the destructor call. This allows the resource T to deallocate or clean up any non-GC resources.

If it is desirable to persist a Unique!T outside of its original scope, then it can be transferred. The transfer can be explicit, by calling release, or implicit, when returning Unique from a function. The resource T can be a polymorphic class object or instance of an interface, in which case Unique behaves polymorphically too.

If T is a value type, then Unique!T will be implemented as a reference to a T.

Examples:

struct S
{
    int i;
    this(int i){this.i = i;}
}
Unique!S produce()
{
    // Construct a unique instance of S on the heap
    Unique!S ut = new S(5);
    // Implicit transfer of ownership
    return ut;
}
// Borrow a unique resource by ref
void increment(ref Unique!S ur)
{
    ur.i++;
}
void consume(Unique!S u2)
{
    writeln(u2.i); // 6
    // Resource automatically deleted here
}
Unique!S u1;
assert(u1.isEmpty);
u1 = produce();
writeln(u1.i); // 5
increment(u1);
writeln(u1.i); // 6
//consume(u1); // Error: u1 is not copyable
// Transfer ownership of the resource
consume(u1.release);
assert(u1.isEmpty);

alias RefT = T;

Represents a reference to T. Resolves to T* if T is a value type.

Unique!T create(A...)(auto ref A args) if (__traits(compiles, new T(args)));

Allows safe construction of Unique. It creates the resource and guarantees unique ownership of it (unless T publishes aliases of this).

Note Nested structs/classes cannot be created.

Parameters:

A args

Arguments to pass to T's constructor.

static class C {}
auto u = Unique!(C).create();

this(RefT p);

Constructor that takes an rvalue. It will ensure uniqueness, as long as the rvalue isn't just a view on an lvalue (e.g., a cast). Typical usage:

Unique!Foo f = new Foo;

this(ref RefT p);

Constructor that takes an lvalue. It nulls its source. The nulling will ensure uniqueness as long as there are no previous aliases to the source.

this(U)(Unique!U u) if (is(u.RefT : RefT));

Constructor that takes a Unique of a type that is convertible to our type.

Typically used to transfer a Unique rvalue of derived type to a Unique of base type.

Example

class C : Object {}

Unique!C uc = new C;
Unique!Object uo = uc.release;

void opAssign(U)(Unique!U u) if (is(u.RefT : RefT));

Transfer ownership from a Unique of a type that is convertible to our type.

const @property bool isEmpty();

Returns whether the resource exists.

Unique release();

Transfer ownership to a Unique rvalue. Nullifies the current contents. Same as calling std.algorithm.move on it.

struct Tuple(Specs...) if (distinctFieldNames!Specs);

Tuple of values, for example Tuple!(int, string) is a record that stores an int and a string. Tuple can be used to bundle values together, notably when returning multiple values from a function. If obj is a Tuple, the individual members are accessible with the syntax obj[0] for the first field, obj[1] for the second, and so on.

See Also:

tuple.

Parameters:

Specs A list of types (and optionally, member names) that the Tuple contains.

Examples:

Tuple!(int, int) point;
// assign coordinates
point[0] = 5;
point[1] = 6;
// read coordinates
auto x = point[0];
auto y = point[1];

Examples:

Tuple members can be named. It is legal to mix named and unnamed members. The method above is still applicable to all fields.

alias Entry = Tuple!(int, "index", string, "value");
Entry e;
e.index = 4;
e.value = "Hello";
writeln(e[1]); // "Hello"
writeln(e[0]); // 4

Examples:

A Tuple with named fields is a distinct type from a Tuple with unnamed fields, i.e. each naming imparts a separate type for the Tuple. Two Tuples differing in naming only are still distinct, even though they might have the same structure.

Tuple!(int, "x", int, "y") point1;
Tuple!(int, int) point2;
assert(!is(typeof(point1) == typeof(point2)));

Examples:

Use tuples as ranges

import std.algorithm.iteration : sum;
import std.range : only;
auto t = tuple(1, 2);
writeln(t.expand.only.sum); // 3

Examples:

Concatenate tuples

import std.meta : AliasSeq;
auto t = tuple(1, "2") ~ tuple(ushort(42), true);
static assert(is(t.Types == AliasSeq!(int, string, ushort, bool)));
writeln(t[1]); // "2"
writeln(t[2]); // 42
writeln(t[3]); // true

alias Types = staticMap!(extractType, fieldSpecs);

The types of the Tuple's components.

alias fieldNames = staticMap!(extractName, fieldSpecs);

The names of the Tuple's components. Unnamed fields have empty names.

Examples:

import std.meta : AliasSeq;
alias Fields = Tuple!(int, "id", string, float);
static assert(Fields.fieldNames == AliasSeq!("id", "", ""));

Types expand;

Use t.expand for a Tuple t to expand it into its components. The result of expand acts as if the Tuple's components were listed as a list of values. (Ordinarily, a Tuple acts as a single value.)

Examples:

auto t1 = tuple(1, " hello ", 'a');
writeln(t1.toString()); // `Tuple!(int, string, char)(1, " hello ", 'a')`

void takeSeveralTypes(int n, string s, bool b)
{
    assert(n == 4 && s == "test" && b == false);
}

auto t2 = tuple(4, "test", false);
//t.expand acting as a list of values
takeSeveralTypes(t2.expand);

this(Types values);

Constructor taking one value for each field.

Parameters:

Types values A list of values that are either the same types as those given by the Types field of this Tuple, or can implicitly convert to those types. They must be in the same order as they appear in Types.

Examples:

alias ISD = Tuple!(int, string, double);
auto tup = ISD(1, "test", 3.2);
writeln(tup.toString()); // `Tuple!(int, string, double)(1, "test", 3.2)`

this(U, size_t n)(U[n] values) if (n == Types.length && allSatisfy!(isBuildableFrom!U, Types));

Constructor taking a compatible array.

Parameters:

U[n] values A compatible static array to build the Tuple from. Array slices are not supported.

Examples:

int[2] ints;
Tuple!(int, int) t = ints;

this(U)(U another) if (areBuildCompatibleTuples!(typeof(this), U) && (noMemberHasCopyCtor!(typeof(this)) || !is(Unqual!U == Unqual!(typeof(this)))));

Constructor taking a compatible Tuple. Two Tuples are compatible iff they are both of the same length, and, for each type T on the left-hand side, the corresponding type U on the right-hand side can implicitly convert to T.

Parameters:

U another A compatible Tuple to build from. Its type must be compatible with the target Tuple's type.

Examples:

alias IntVec = Tuple!(int, int, int);
alias DubVec = Tuple!(double, double, double);

IntVec iv = tuple(1, 1, 1);

//Ok, int can implicitly convert to double
DubVec dv = iv;
//Error: double cannot implicitly convert to int
//IntVec iv2 = dv;

bool opEquals(R)(R rhs) if (areCompatibleTuples!(typeof(this), R, "==")); const bool opEquals(R)(R rhs) if (areCompatibleTuples!(typeof(this), R, "==")); bool opEquals(R...)(auto ref R rhs) if (R.length > 1 && areCompatibleTuples!(typeof(this), Tuple!R, "=="));

Comparison for equality. Two Tuples are considered equal iff they fulfill the following criteria:

Each Tuple is the same length.
For each type T on the left-hand side and each type U on the right-hand side, values of type T can be compared with values of type U.
For each value v1 on the left-hand side and each value v2 on the right-hand side, the expression v1 == v2 is true.

Parameters:

R rhs The Tuple to compare against. It must meeting the criteria for comparison between Tuples.

Returns:

true if both Tuples are equal, otherwise false.

Examples:

Tuple!(int, string) t1 = tuple(1, "test");
Tuple!(double, string) t2 =  tuple(1.0, "test");
//Ok, int can be compared with double and
//both have a value of 1
writeln(t1); // t2

auto opCmp(R)(R rhs) if (areCompatibleTuples!(typeof(this), R, "<")); const auto opCmp(R)(R rhs) if (areCompatibleTuples!(typeof(this), R, "<"));

Comparison for ordering.

Parameters:

R rhs The Tuple to compare against. It must meet the criteria for comparison between Tuples.

Returns:

For any values v1 contained by the left-hand side tuple and any values v2 contained by the right-hand side:

0 if v1 == v2 for all members or the following value for the first position were the mentioned criteria is not satisfied:

NaN, in case one of the operands is a NaN.
A negative number if the expression v1 < v2 is true.
A positive number if the expression v1 > v2 is true.

Examples:

The first v1 for which v1 > v2 is true determines the result. This could lead to unexpected behaviour.

auto tup1 = tuple(1, 1, 1);
auto tup2 = tuple(1, 100, 100);
assert(tup1 < tup2);

//Only the first result matters for comparison
tup1[0] = 2;
assert(tup1 > tup2);

auto opBinary(string op, T)(auto ref T t) if (op == "~" && !(is(T : U[], U) && isTuple!U)); auto opBinaryRight(string op, T)(auto ref T t) if (op == "~" && !(is(T : U[], U) && isTuple!U));

Concatenate Tuples. Tuple concatenation is only allowed if all named fields are distinct (no named field of this tuple occurs in t and no named field of t occurs in this tuple).

Parameters:

T t The Tuple to concatenate with

Returns:

A concatenation of this tuple and t

ref Tuple opAssign(R)(auto ref R rhs) if (areCompatibleTuples!(typeof(this), R, "="));

Assignment from another Tuple.

Parameters:

R rhs The source Tuple to assign from. Each element of the source Tuple must be implicitly assignable to each respective element of the target Tuple.

inout ref auto rename(names...)() return if (names.length == 0 || allSatisfy!(isSomeString, typeof(names)));

Renames the elements of a Tuple.

rename uses the passed names and returns a new Tuple using these names, with the content unchanged. If fewer names are passed than there are members of the Tuple then those trailing members are unchanged. An empty string will remove the name for that member. It is an compile-time error to pass more names than there are members of the Tuple.

Examples:

auto t0 = tuple(4, "hello");

auto t0Named = t0.rename!("val", "tag");
writeln(t0Named.val); // 4
writeln(t0Named.tag); // "hello"

Tuple!(float, "dat", size_t[2], "pos") t1;
t1.pos = [2, 1];
auto t1Named = t1.rename!"height";
t1Named.height = 3.4f;
writeln(t1Named.height); // 3.4f
writeln(t1Named.pos); // [2, 1]
t1Named.rename!"altitude".altitude = 5;
writeln(t1Named.height); // 5

Tuple!(int, "a", int, int, "c") t2;
t2 = tuple(3,4,5);
auto t2Named = t2.rename!("", "b");
// "a" no longer has a name
static assert(!__traits(hasMember, typeof(t2Named), "a"));
writeln(t2Named[0]); // 3
writeln(t2Named.b); // 4
writeln(t2Named.c); // 5

// not allowed to specify more names than the tuple has members
static assert(!__traits(compiles, t2.rename!("a","b","c","d")));

// use it in a range pipeline
import std.range : iota, zip;
import std.algorithm.iteration : map, sum;
auto res = zip(iota(1, 4), iota(10, 13))
    .map!(t => t.rename!("a", "b"))
    .map!(t => t.a * t.b)
    .sum;
writeln(res); // 68

const tup = Tuple!(int, "a", int, "b")(2, 3);
const renamed = tup.rename!("c", "d");
writeln(renamed.c + renamed.d); // 5

inout ref auto rename(alias translate)() if (is(typeof(translate) : V[K], V, K) && isSomeString!V && (isSomeString!K || is(K : size_t)));

Overload of rename that takes an associative array translate as a template parameter, where the keys are either the names or indices of the members to be changed and the new names are the corresponding values. Every key in translate must be the name of a member of the tuple. The same rules for empty strings apply as for the variadic template overload of rename.

Examples:

//replacing names by their current name

Tuple!(float, "dat", size_t[2], "pos") t1;
t1.pos = [2, 1];
auto t1Named = t1.rename!(["dat": "height"]);
t1Named.height = 3.4;
writeln(t1Named.pos); // [2, 1]
t1Named.rename!(["height": "altitude"]).altitude = 5;
writeln(t1Named.height); // 5

Tuple!(int, "a", int, "b") t2;
t2 = tuple(3, 4);
auto t2Named = t2.rename!(["a": "b", "b": "c"]);
writeln(t2Named.b); // 3
writeln(t2Named.c); // 4

const t3 = Tuple!(int, "a", int, "b")(3, 4);
const t3Named = t3.rename!(["a": "b", "b": "c"]);
writeln(t3Named.b); // 3
writeln(t3Named.c); // 4

Examples:

//replace names by their position

Tuple!(float, "dat", size_t[2], "pos") t1;
t1.pos = [2, 1];
auto t1Named = t1.rename!([0: "height"]);
t1Named.height = 3.4;
writeln(t1Named.pos); // [2, 1]
t1Named.rename!([0: "altitude"]).altitude = 5;
writeln(t1Named.height); // 5

Tuple!(int, "a", int, "b", int, "c") t2;
t2 = tuple(3, 4, 5);
auto t2Named = t2.rename!([0: "c", 2: "a"]);
writeln(t2Named.a); // 5
writeln(t2Named.b); // 4
writeln(t2Named.c); // 3

inout @property ref @trusted inout(Tuple!(sliceSpecs!(from, to))) slice(size_t from, size_t to)() if (from <= to && (to <= Types.length));

Takes a slice by-reference of this Tuple.

Parameters:

from	A size_t designating the starting position of the slice.
to	A size_t designating the ending position (exclusive) of the slice.

Returns:

A new Tuple that is a slice from [from, to) of the original. It has the same types and values as the range [from, to) in the original.

Examples:

Tuple!(int, string, float, double) a;
a[1] = "abc";
a[2] = 4.5;
auto s = a.slice!(1, 3);
static assert(is(typeof(s) == Tuple!(string, float)));
assert(s[0] == "abc" && s[1] == 4.5);

// https://issues.dlang.org/show_bug.cgi?id=15645
Tuple!(int, short, bool, double) b;
static assert(!__traits(compiles, b.slice!(2, 4)));

const nothrow @safe size_t toHash();

Creates a hash of this Tuple.

Returns:

A size_t representing the hash of this Tuple.

const string toString()();

Converts to string.

Returns:

The string representation of this Tuple.

const void toString(DG)(scope DG sink); const void toString(DG, Char)(scope DG sink, ref scope const FormatSpec!Char fmt);

Formats Tuple with either %s, %(inner%) or %(inner%|sep%).

Formats supported by Tuple
Format	Description
%s	Format like Tuple!(types)(elements formatted with %s each).
%(inner%)	The format inner is applied the expanded Tuple, so it may contain as many formats as the Tuple has fields.
%(inner%\|sep%)	The format inner is one format, that is applied on all fields of the Tuple. The inner format must be compatible to all of them.

Parameters:

DG `sink`	A char accepting delegate
FormatSpec!Char `fmt`	A std.format.FormatSpec

Examples:

import std.format : format;

Tuple!(int, double)[3] tupList = [ tuple(1, 1.0), tuple(2, 4.0), tuple(3, 9.0) ];

// Default format
writeln(format("%s", tuple("a", 1))); // `Tuple!(string, int)("a", 1)`

// One Format for each individual component
writeln(format("%(%#x v %.4f w %#x%)", tuple(1, 1.0, 10))); // `0x1 v 1.0000 w 0xa`
writeln(format("%#x v %.4f w %#x", tuple(1, 1.0, 10).expand)); // `0x1 v 1.0000 w 0xa`

// One Format for all components
// `>abc< & >1< & >2.3< & >[4, 5]<`
writeln(format("%(>%s<%| & %)", tuple("abc", 1, 2.3, [4, 5])));

// Array of Tuples
writeln(format("%(%(f(%d) = %.1f%);  %)", tupList)); // `f(1) = 1.0;  f(2) = 4.0;  f(3) = 9.0`

Examples:

import std.exception : assertThrown;
import std.format : format, FormatException;

// Error: %( %) missing.
assertThrown!FormatException(
    format("%d, %f", tuple(1, 2.0)) == `1, 2.0`
);

// Error: %( %| %) missing.
assertThrown!FormatException(
    format("%d", tuple(1, 2)) == `1, 2`
);

// Error: %d inadequate for double
assertThrown!FormatException(
    format("%(%d%|, %)", tuple(1, 2.0)) == `1, 2.0`
);

auto reverse(T)(T t) if (isTuple!T);

Creates a copy of a Tuple with its fields in reverse order.

Parameters:

T t The Tuple to copy.

Returns:

A new Tuple.

Examples:

auto tup = tuple(1, "2");
writeln(tup.reverse); // tuple("2", 1)

template tuple(Names...)

Constructs a Tuple object instantiated and initialized according to the given arguments.

Parameters:

Names

An optional list of strings naming each successive field of the Tuple or a list of types that the elements are being casted to. For a list of names, each name matches up with the corresponding field given by Args. A name does not have to be provided for every field, but as the names must proceed in order, it is not possible to skip one field and name the next after it. For a list of types, there must be exactly as many types as parameters.

Examples:

auto value = tuple(5, 6.7, "hello");
writeln(value[0]); // 5
writeln(value[1]); // 6.7
writeln(value[2]); // "hello"

// Field names can be provided.
auto entry = tuple!("index", "value")(4, "Hello");
writeln(entry.index); // 4
writeln(entry.value); // "Hello"

auto tuple(Args...)(Args args);

Parameters:

Args args Values to initialize the Tuple with. The Tuple's type will be inferred from the types of the values given.

Returns:

A new Tuple with its type inferred from the arguments given.

enum auto isTuple(T);

Returns true if and only if T is an instance of std.typecons.Tuple.

Parameters:

T	The type to check.

Returns:

true if T is a Tuple type, false otherwise.

Examples:

static assert(isTuple!(Tuple!()));
static assert(isTuple!(Tuple!(int)));
static assert(isTuple!(Tuple!(int, real, string)));
static assert(isTuple!(Tuple!(int, "x", real, "y")));
static assert(isTuple!(Tuple!(int, Tuple!(real), string)));

template Rebindable(T) if (is(T == class) || is(T == interface) || isDynamicArray!T || isAssociativeArray!T) struct Rebindable(T) if (!is(T == class) && !is(T == interface) && !isDynamicArray!T && !isAssociativeArray!T);

Rebindable!(T) is a simple, efficient wrapper that behaves just like an object of type T, except that you can reassign it to refer to another object. For completeness, Rebindable!(T) aliases itself away to T if T is a non-const object type.

You may want to use Rebindable when you want to have mutable storage referring to const objects, for example an array of references that must be sorted in place. Rebindable does not break the soundness of D's type system and does not incur any of the risks usually associated with cast.

Parameters:

Any type.

Examples:

Regular const object references cannot be reassigned.

class Widget { int x; int y() @safe const { return x; } }
const a = new Widget;
// Fine
a.y();
// error! can't modify const a
// a.x = 5;
// error! can't modify const a
// a = new Widget;

Examples:

However, Rebindable!(Widget) does allow reassignment, while otherwise behaving exactly like a const Widget.

class Widget { int x; int y() const @safe { return x; } }
auto a = Rebindable!(const Widget)(new Widget);
// Fine
a.y();
// error! can't modify const a
// a.x = 5;
// Fine
a = new Widget;

Examples:

Using Rebindable in a generic algorithm:

import std.range.primitives : front, popFront;

// simple version of std.algorithm.searching.maxElement
typeof(R.init.front) maxElement(R)(R r)
{
    auto max = rebindable(r.front);
    r.popFront;
    foreach (e; r)
        if (e > max)
            max = e; // Rebindable allows const-correct reassignment
    return max;
}
struct S
{
    char[] arr;
    alias arr this; // for comparison
}
// can't convert to mutable
const S cs;
static assert(!__traits(compiles, { S s = cs; }));

alias CS = const S;
CS[] arr = [CS("harp"), CS("apple"), CS("pot")];
CS ms = maxElement(arr);
writeln(ms.arr); // "pot"

Examples:

static struct S
{
    int* ptr;
}
S s = S(new int);

const cs = s;
// Can't assign s.ptr to cs.ptr
static assert(!__traits(compiles, {s = cs;}));

Rebindable!(const S) rs = s;
assert(rs.ptr is s.ptr);
// rs.ptr is const
static assert(!__traits(compiles, {rs.ptr = null;}));

// Can't assign s.ptr to rs.ptr
static assert(!__traits(compiles, {s = rs;}));

const S cs2 = rs;
// Rebind rs
rs = cs2;
rs = S();
assert(rs.ptr is null);

Rebindable!T rebindable(T)(T obj) if (is(T == class) || is(T == interface) || isDynamicArray!T || isAssociativeArray!T); Rebindable!T rebindable(T)(T value) if (!is(T == class) && !is(T == interface) && !isDynamicArray!T && !isAssociativeArray!T && !is(T : Rebindable!U, U));

Convenience function for creating a Rebindable using automatic type inference.

Parameters:

T obj A reference to a value to initialize the Rebindable with.

Returns:

A newly constructed Rebindable initialized with the given reference.

Examples:

class C
{
    int payload;
    this(int p) { payload = p; }
}
const c = new C(1);

auto c2 = c.rebindable;
writeln(c2.payload); // 1
// passing Rebindable to rebindable
c2 = c2.rebindable;

c2 = new C(2);
writeln(c2.payload); // 2

const c3 = c2.get;
writeln(c3.payload); // 2

Examples:

immutable struct S
{
    int[] array;
}
auto s1 = [3].idup.rebindable;
s1 = [4].idup.rebindable;
writeln(s1); // [4]

Rebindable!T rebindable(T)(Rebindable!T obj);

This function simply returns the Rebindable object passed in. It's useful in generic programming cases when a given object may be either a regular class or a Rebindable.

Parameters:

Rebindable!T obj An instance of Rebindable!T.

Returns:

obj without any modification.

Examples:

class C
{
    int payload;
    this(int p) { payload = p; }
}
const c = new C(1);

auto c2 = c.rebindable;
writeln(c2.payload); // 1
// passing Rebindable to rebindable
c2 = c2.rebindable;
writeln(c2.payload); // 1

template UnqualRef(T) if (is(T == class) || is(T == interface))

Similar to Rebindable!(T) but strips all qualifiers from the reference as opposed to just constness / immutability. Primary intended use case is with shared (having thread-local reference to shared class data)

Parameters:

T	A class or interface type.

Examples:

class Data {}

static shared(Data) a;
static UnqualRef!(shared Data) b;

import core.thread;

auto thread = new core.thread.Thread({
    a = new shared Data();
    b = new shared Data();
});

thread.start();
thread.join();

assert(a !is null);
assert(b is null);

string alignForSize(E...)(const char[][] names...);

Order the provided members to minimize size while preserving alignment. Alignment is not always optimal for 80-bit reals, nor for structs declared as align(1).

Parameters:

E	A list of the types to be aligned, representing fields of an aggregate such as a struct or class.
char[][] `names`	The names of the fields that are to be aligned.

Returns:

A string to be mixed in to an aggregate, such as a struct or class.

Examples:

struct Banner {
    mixin(alignForSize!(byte[6], double)(["name", "height"]));
}

struct Nullable(T); auto nullable(T)(T t);

Defines a value paired with a distinctive "null" state that denotes the absence of a value. If default constructed, a Nullable!T object starts in the null state. Assigning it renders it non-null. Calling nullify can nullify it again.

Practically Nullable!T stores a T and a bool.

See also: apply, an alternative way to use the payload.

Examples:

struct CustomerRecord
{
    string name;
    string address;
    int customerNum;
}

Nullable!CustomerRecord getByName(string name)
{
    //A bunch of hairy stuff

    return Nullable!CustomerRecord.init;
}

auto queryResult = getByName("Doe, John");
if (!queryResult.isNull)
{
    //Process Mr. Doe's customer record
    auto address = queryResult.get.address;
    auto customerNum = queryResult.get.customerNum;

    //Do some things with this customer's info
}
else
{
    //Add the customer to the database
}

Examples:

import std.exception : assertThrown;

auto a = 42.nullable;
assert(!a.isNull);
writeln(a.get); // 42

a.nullify();
assert(a.isNull);
assertThrown!Throwable(a.get);

Examples:

import std.algorithm.iteration : each, joiner;
Nullable!int a = 42;
Nullable!int b;
// Add each value to an array
int[] arr;
a.each!((n) => arr ~= n);
writeln(arr); // [42]
b.each!((n) => arr ~= n);
writeln(arr); // [42]
// Take first value from an array of Nullables
Nullable!int[] c = new Nullable!int[](10);
c[7] = Nullable!int(42);
writeln(c.joiner.front); // 42

inout this(inout T value);

Constructor initializing this with value.

Parameters:

T value The value to initialize this Nullable with.

bool opEquals(this This, Rhs)(auto ref Rhs rhs) if (!is(CommonType!(This, Rhs) == void)); bool opEquals(this This, Rhs)(auto ref Rhs rhs) if (is(CommonType!(This, Rhs) == void) && is(typeof(this.get == rhs)));

If they are both null, then they are equal. If one is null and the other is not, then they are not equal. If they are both non-null, then they are equal if their values are equal.

Examples:

Nullable!int empty;
Nullable!int a = 42;
Nullable!int b = 42;
Nullable!int c = 27;

writeln(empty); // empty
writeln(empty); // Nullable!int.init
assert(empty != a);
assert(empty != b);
assert(empty != c);

writeln(a); // b
assert(a != c);

assert(empty != 42);
writeln(a); // 42
assert(c != 42);

string toString(); const string toString(); void toString(W)(ref W writer, ref scope const FormatSpec!char fmt) if (isOutputRange!(W, char)); const void toString(W)(ref W writer, ref scope const FormatSpec!char fmt) if (isOutputRange!(W, char));

Gives the string "Nullable.null" if isNull is true. Otherwise, the result is equivalent to calling std.format.formattedWrite on the underlying value.

Parameters:

W `writer`	A char accepting output range
FormatSpec!char `fmt`	A std.format.FormatSpec which is used to represent the value if this Nullable is not null

Returns:

A string if writer and fmt are not set; void otherwise.

const pure nothrow @property @safe bool isNull();

Check if this is in the null state.

Returns:

true iff this is in the null state, otherwise false.

Examples:

Nullable!int ni;
assert(ni.isNull);

ni = 0;
assert(!ni.isNull);

void nullify()();

Forces this to the null state.

Examples:

Nullable!int ni = 0;
assert(!ni.isNull);

ni.nullify();
assert(ni.isNull);

ref Nullable opAssign()(T value) return;

Assigns value to the internally-held state. If the assignment succeeds, this becomes non-null.

Parameters:

T value A value of type T to assign to this Nullable.

Examples:

If this Nullable wraps a type that already has a null value (such as a pointer), then assigning the null value to this Nullable is no different than assigning any other value of type T, and the resulting code will look very strange. It is strongly recommended that this be avoided by instead using the version of Nullable that takes an additional nullValue template argument.

//Passes
Nullable!(int*) npi;
assert(npi.isNull);

//Passes?!
npi = null;
assert(!npi.isNull);

inout pure nothrow @property ref @safe inout(T) get(); inout @property inout(T) get()(inout(T) fallback); inout @property auto get(U)(inout(U) fallback);

Gets the value if not null. If this is in the null state, and the optional parameter fallback was provided, it will be returned. Without fallback, calling get with a null state is invalid.

When the fallback type is different from the Nullable type, get(T) returns the common type.

Parameters:

inout(T) fallback the value to return in case the Nullable is null.

Returns:

The value held internally by this Nullable.

alias empty = isNull; alias popFront = nullify; alias popBack = nullify; inout pure nothrow @property ref @safe inout(T) front(); alias back = front; inout @property inout(typeof(this)) save(); inout inout(typeof(this)) opIndex(size_t[2] dim); const size_t[2] opSlice(size_t dim : 0)(size_t from, size_t to); const pure nothrow @property @safe size_t length(); template opDollar(size_t dim : 0) inout pure nothrow ref @safe inout(T) opIndex(size_t index);

Range interface functions.

auto opSlice(this This)();

Converts Nullable to a range. Works even when the contained type is immutable.

struct Nullable(T, T nullValue); auto nullable(alias nullValue, T)(T t) if (is(typeof(nullValue) == T));

Just like Nullable!T, except that the null state is defined as a particular value. For example, Nullable!(uint, uint.max) is an uint that sets aside the value uint.max to denote a null state. Nullable!(T, nullValue) is more storage-efficient than Nullable!T because it does not need to store an extra bool.

Parameters:

T	The wrapped type for which Nullable provides a null value.
nullValue	The null value which denotes the null state of this `Nullable`. Must be of type T.

Examples:

Nullable!(size_t, size_t.max) indexOf(string[] haystack, string needle)
{
    //Find the needle, returning -1 if not found

    return Nullable!(size_t, size_t.max).init;
}

void sendLunchInvite(string name)
{
}

//It's safer than C...
auto coworkers = ["Jane", "Jim", "Marry", "Fred"];
auto pos = indexOf(coworkers, "Bob");
if (!pos.isNull)
{
    //Send Bob an invitation to lunch
    sendLunchInvite(coworkers[pos]);
}
else
{
    //Bob not found; report the error
}

//And there's no overhead
static assert(Nullable!(size_t, size_t.max).sizeof == size_t.sizeof);

Examples:

import std.exception : assertThrown;

Nullable!(int, int.min) a;
assert(a.isNull);
assertThrown!Throwable(a.get);
a = 5;
assert(!a.isNull);
writeln(a); // 5
static assert(a.sizeof == int.sizeof);

Examples:

auto a = nullable!(int.min)(8);
writeln(a); // 8
a.nullify();
assert(a.isNull);

this(T value);

Constructor initializing this with value.

Parameters:

T value The value to initialize this Nullable with.

const @property bool isNull();

Check if this is in the null state.

Returns:

true iff this is in the null state, otherwise false.

Examples:

Nullable!(int, -1) ni;
//Initialized to "null" state
assert(ni.isNull);

ni = 0;
assert(!ni.isNull);

void nullify()();

Forces this to the null state.

Examples:

Nullable!(int, -1) ni = 0;
assert(!ni.isNull);

ni = -1;
assert(ni.isNull);

void opAssign()(T value);

Assigns value to the internally-held state. If the assignment succeeds, this becomes non-null. No null checks are made. Note that the assignment may leave this in the null state.

Parameters:

T value A value of type T to assign to this Nullable. If it is nullvalue, then the internal state of this Nullable will be set to null.

Examples:

If this Nullable wraps a type that already has a null value (such as a pointer), and that null value is not given for nullValue, then assigning the null value to this Nullable is no different than assigning any other value of type T, and the resulting code will look very strange. It is strongly recommended that this be avoided by using T's "built in" null value for nullValue.

//Passes
enum nullVal = cast(int*) 0xCAFEBABE;
Nullable!(int*, nullVal) npi;
assert(npi.isNull);

//Passes?!
npi = null;
assert(!npi.isNull);

inout @property ref inout(T) get();

Gets the value. this must not be in the null state. This function is also called for the implicit conversion to T.

Preconditions isNull must be false.

Returns:

The value held internally by this Nullable.

Examples:

import std.exception : assertThrown, assertNotThrown;

Nullable!(int, -1) ni;
//`get` is implicitly called. Will throw
//an error in non-release mode
assertThrown!Throwable(ni == 0);

ni = 0;
assertNotThrown!Throwable(ni == 0);

template apply(alias fun)

Unpacks the content of a Nullable, performs an operation and packs it again. Does nothing if isNull.

When called on a Nullable, apply will unpack the value contained in the Nullable, pass it to the function you provide and wrap the result in another Nullable (if necessary). If the Nullable is null, apply will return null itself.

Parameters:

T t	a Nullable
fun	a function operating on the content of the nullable

Returns:

fun(t.get).nullable if !t.isNull, else Nullable.init.

See Also:

AutoImplement, generateEmptyFunction

Examples:

import std.math.traits : isNaN;

static abstract class C
{
    int m_value;
    this(int v) { m_value = v; }
    int value() @property { return m_value; }

    abstract real realValue() @property;
    abstract void doSomething();
}

auto c = new BlackHole!C(42);
writeln(c.value); // 42

// Returns real.init which is NaN
assert(c.realValue.isNaN);
// Abstract functions are implemented as do-nothing
c.doSomething();

template WhiteHole(Base)

WhiteHole!Base is a subclass of Base which automatically implements all abstract member functions as functions that always fail. These functions simply throw an Error and never return. Whitehole is useful for trapping the use of class member functions that haven't been implemented.

The name came from Class::WhiteHole Perl module by Michael G Schwern.

Parameters:

Base	A non-final class for `WhiteHole` to inherit from.

See Also:

AutoImplement, generateAssertTrap

Examples:

import std.exception : assertThrown;

static class C
{
    abstract void notYetImplemented();
}

auto c = new WhiteHole!C;
assertThrown!NotImplementedError(c.notYetImplemented()); // throws an Error

class AutoImplement(Base, alias how, alias what = isAbstractFunction) if (!is(how == class)): Base; class AutoImplement(Interface, BaseClass, alias how, alias what = isAbstractFunction) if (is(Interface == interface) && is(BaseClass == class)): BaseClass, Interface;

AutoImplement automatically implements (by default) all abstract member functions in the class or interface Base in specified way.

The second version of AutoImplement automatically implements Interface, while deriving from BaseClass.

Parameters:

how

template which specifies how functions will be implemented/overridden.

Two arguments are passed to how: the type Base and an alias to an implemented function. Then how must return an implemented function body as a string.

The generated function body can use these keywords:

a0, a1, …: arguments passed to the function;
args: a tuple of the arguments;
self: an alias to the function itself;
parent: an alias to the overridden function (if any).

You may want to use templated property functions (instead of Implicit Template Properties) to generate complex functions:

// Prints log messages for each call to overridden functions.
string generateLogger(C, alias fun)() @property
{
    import std.traits;
    enum qname = C.stringof ~ "." ~ __traits(identifier, fun);
    string stmt;

    stmt ~= q{ struct Importer { import std.stdio; } };
    stmt ~= `Importer.writeln("Log: ` ~ qname ~ `(", args, ")");`;
    static if (!__traits(isAbstractFunction, fun))
    {
        static if (is(ReturnType!fun == void))
            stmt ~= q{ parent(args); };
        else
            stmt ~= q{
                auto r = parent(args);
                Importer.writeln("--> ", r);
                return r;
            };
    }
    return stmt;
}

what

template which determines what functions should be implemented/overridden.

An argument is passed to what: an alias to a non-final member function in Base. Then what must return a boolean value. Return true to indicate that the passed function should be implemented/overridden.

// Sees if fun returns something.
enum bool hasValue(alias fun) = !is(ReturnType!(fun) == void);

Note Generated code is inserted in the scope of std.typecons module. Thus, any useful functions outside std.typecons cannot be used in the generated code. To workaround this problem, you may import necessary things in a local struct, as done in the generateLogger() template in the above example.

Bugs:

Variadic arguments to constructors are not forwarded to super.
Deep interface inheritance causes compile error with messages like "Error: function std.typecons.AutoImplement!(Foo).AutoImplement.bar does not override any function". [Bugzilla 2525]
The parent keyword is actually a delegate to the super class' corresponding member function. [Bugzilla 2540]
Using alias template parameter in how and/or what may cause strange compile error. Use template tuple parameter instead to workaround this problem. [Bugzilla 4217]

Examples:

interface PackageSupplier
{
    int foo();
    int bar();
}

static abstract class AbstractFallbackPackageSupplier : PackageSupplier
{
    protected PackageSupplier default_, fallback;

    this(PackageSupplier default_, PackageSupplier fallback)
    {
        this.default_ = default_;
        this.fallback = fallback;
    }

    abstract int foo();
    abstract int bar();
}

template fallback(T, alias func)
{
    import std.format : format;
    // for all implemented methods:
    // - try default first
    // - only on a failure run & return fallback
    enum fallback = q{
        try
        {
            return default_.%1$s(args);
        }
        catch (Exception)
        {
            return fallback.%1$s(args);
        }
    }.format(__traits(identifier, func));
}

// combines two classes and use the second one as fallback
alias FallbackPackageSupplier = AutoImplement!(AbstractFallbackPackageSupplier, fallback);

class FailingPackageSupplier : PackageSupplier
{
    int foo(){ throw new Exception("failure"); }
    int bar(){ return 2;}
}

class BackupPackageSupplier : PackageSupplier
{
    int foo(){ return -1; }
    int bar(){ return -1;}
}

auto registry = new FallbackPackageSupplier(new FailingPackageSupplier(), new BackupPackageSupplier());

writeln(registry.foo()); // -1
writeln(registry.bar()); // 2

template generateEmptyFunction(C, func...) enum string generateAssertTrap(C, func...);

Predefined how-policies for AutoImplement. These templates are also used by BlackHole and WhiteHole, respectively.

Examples:

alias BlackHole(Base) = AutoImplement!(Base, generateEmptyFunction);

interface I
{
    int foo();
    string bar();
}

auto i = new BlackHole!I();
// generateEmptyFunction returns the default value of the return type without doing anything
writeln(i.foo); // 0
assert(i.bar is null);

Examples:

import std.exception : assertThrown;

alias WhiteHole(Base) = AutoImplement!(Base, generateAssertTrap);

interface I
{
    int foo();
    string bar();
}

auto i = new WhiteHole!I();
// generateAssertTrap throws an exception for every unimplemented function of the interface
assertThrown!NotImplementedError(i.foo);
assertThrown!NotImplementedError(i.bar);

template wrap(Targets...) if (Targets.length >= 1 && allSatisfy!(isMutable, Targets)) template wrap(Targets...) if (Targets.length >= 1 && !allSatisfy!(isMutable, Targets)) template unwrap(Target) if (isMutable!Target) template unwrap(Target) if (!isMutable!Target)

Supports structural based typesafe conversion.

If Source has structural conformance with the interface Targets, wrap creates an internal wrapper class which inherits Targets and wraps the src object, then returns it.

unwrap can be used to extract objects which have been wrapped by wrap.

Examples:

interface Quack
{
    int quack();
    @property int height();
}
interface Flyer
{
    @property int height();
}
class Duck : Quack
{
    int quack() { return 1; }
    @property int height() { return 10; }
}
class Human
{
    int quack() { return 2; }
    @property int height() { return 20; }
}

Duck d1 = new Duck();
Human h1 = new Human();

interface Refleshable
{
    int reflesh();
}

// does not have structural conformance
static assert(!__traits(compiles, d1.wrap!Refleshable));
static assert(!__traits(compiles, h1.wrap!Refleshable));

// strict upcast
Quack qd = d1.wrap!Quack;
assert(qd is d1);
assert(qd.quack() == 1);    // calls Duck.quack
// strict downcast
Duck d2 = qd.unwrap!Duck;
assert(d2 is d1);

// structural upcast
Quack qh = h1.wrap!Quack;
assert(qh.quack() == 2);    // calls Human.quack
// structural downcast
Human h2 = qh.unwrap!Human;
assert(h2 is h1);

// structural upcast (two steps)
Quack qx = h1.wrap!Quack;   // Human -> Quack
Flyer fx = qx.wrap!Flyer;   // Quack -> Flyer
assert(fx.height == 20);    // calls Human.height
// structural downcast (two steps)
Quack qy = fx.unwrap!Quack; // Flyer -> Quack
Human hy = qy.unwrap!Human; // Quack -> Human
assert(hy is h1);
// structural downcast (one step)
Human hz = fx.unwrap!Human; // Flyer -> Human
assert(hz is h1);

Examples:

import std.traits : FunctionAttribute, functionAttributes;
interface A { int run(); }
interface B { int stop(); @property int status(); }
class X
{
    int run() { return 1; }
    int stop() { return 2; }
    @property int status() { return 3; }
}

auto x = new X();
auto ab = x.wrap!(A, B);
A a = ab;
B b = ab;
writeln(a.run()); // 1
writeln(b.stop()); // 2
writeln(b.status); // 3
static assert(functionAttributes!(typeof(ab).status) & FunctionAttribute.property);

enum RefCountedAutoInitialize: int;

Options regarding auto-initialization of a SafeRefCounted object (see the definition of SafeRefCounted below).

Examples:

import core.exception : AssertError;
import std.exception : assertThrown;

struct Foo
{
    int a = 42;
}

SafeRefCounted!(Foo, RefCountedAutoInitialize.yes) rcAuto;
SafeRefCounted!(Foo, RefCountedAutoInitialize.no) rcNoAuto;

writeln(rcAuto.refCountedPayload.a); // 42

assertThrown!AssertError(rcNoAuto.refCountedPayload);
rcNoAuto.refCountedStore.ensureInitialized;
writeln(rcNoAuto.refCountedPayload.a); // 42

no: Do not auto-initialize the object
yes: Auto-initialize the object

struct SafeRefCounted(T, RefCountedAutoInitialize autoInit = RefCountedAutoInitialize.yes) if (!is(T == class) && !is(T == interface));

Defines a reference-counted object containing a T value as payload.

An instance of SafeRefCounted is a reference to a structure, which is referred to as the store, or storage implementation struct in this documentation. The store contains a reference count and the T payload. SafeRefCounted uses malloc to allocate the store. As instances of SafeRefCounted are copied or go out of scope, they will automatically increment or decrement the reference count. When the reference count goes down to zero, SafeRefCounted will call destroy against the payload and call free to deallocate the store. If the T payload contains any references to GC-allocated memory, then SafeRefCounted will add it to the GC memory that is scanned for pointers, and remove it from GC scanning before free is called on the store.

One important consequence of destroy is that it will call the destructor of the T payload. GC-managed references are not guaranteed to be valid during a destructor call, but other members of T, such as file handles or pointers to malloc memory, will still be valid during the destructor call. This allows the T to deallocate or clean up any non-GC resources immediately after the reference count has reached zero.

Without -preview=dip1000, SafeRefCounted is unsafe and should be used with care. No references to the payload should be escaped outside the SafeRefCounted object.

With -preview=dip1000, SafeRefCounted is safe if it's payload is accessed only with the borrow function. Scope semantics can also prevent accidental escaping of refCountedPayload, but it's still up to the user to not destroy the last counted reference while the payload is in use. Due to that, refCountedPayload remains accessible only in @system code.

The autoInit option makes the object ensure the store is automatically initialized. Leaving autoInit == RefCountedAutoInitialize.yes (the default option) is convenient but has the cost of a test whenever the payload is accessed. If autoInit == RefCountedAutoInitialize.no, user code must call either refCountedStore.isInitialized or refCountedStore.ensureInitialized before attempting to access the payload. Not doing so results in null pointer dereference.

If T.this() is annotated with @disable then autoInit must be RefCountedAutoInitialize.no in order to compile.

See Also:

RefCounted

Examples:

// A pair of an `int` and a `size_t` - the latter being the
// reference count - will be dynamically allocated
auto rc1 = SafeRefCounted!int(5);
writeln(rc1); // 5
// No more allocation, add just one extra reference count
auto rc2 = rc1;
// Reference semantics
rc2 = 42;
writeln(rc1); // 42
// the pair will be freed when rc1 and rc2 go out of scope

struct RefCountedStore;

SafeRefCounted storage implementation.

const pure nothrow @nogc @property @safe bool isInitialized();: Returns true if and only if the underlying store has been allocated and initialized.
const pure nothrow @nogc @property @safe size_t refCount();: Returns underlying reference count if it is allocated and initialized (a positive integer), and 0 otherwise.
pure nothrow @safe void ensureInitialized()();: Makes sure the payload was properly initialized. Such a call is typically inserted before using the payload.
This function is unavailable if T.this() is annotated with @disable.

inout nothrow @property ref @safe inout(RefCountedStore) refCountedStore();

Returns storage implementation struct.

this(A...)(auto ref A args) if (A.length > 0); this(return scope T val);

Constructor that initializes the payload.

Postcondition refCountedStore.isInitialized

void opAssign(typeof(this) rhs); void opAssign(T rhs);

Assignment operators.

Note You may not assign a new payload to an uninitialized SafeRefCounted, if auto initialization is off. Assigning another counted reference is still okay.

@property ref @system T refCountedPayload() return; inout pure nothrow @nogc @property ref @system inout(T) refCountedPayload() return;

Returns a reference to the payload. If (autoInit == RefCountedAutoInitialize.yes), calls refCountedStore.ensureInitialized. Otherwise, just issues assert(refCountedStore.isInitialized). Used with alias refCountedPayload this;, so callers can just use the SafeRefCounted object as a T.

The first overload exists only if autoInit == RefCountedAutoInitialize.yes. So if autoInit == RefCountedAutoInitialize.no or called for a constant or immutable object, then refCountedPayload will also be qualified as nothrow (but will still assert if not initialized).

template borrow(alias fun)

Borrows the payload of SafeRefCounted for use in fun. Inferred as @safe if fun is @safe and does not escape a reference to the payload. The reference count will be incremented for the duration of the operation, so destroying the last reference will not leave dangling references in fun.

Parameters:

fun	A callable accepting the payload either by value or by reference.
RC refCount	The counted reference to the payload.

Returns:

The return value of fun, if any. ref in the return value will be forwarded.

Issues For yet unknown reason, code that uses this function with UFCS syntax will not be inferred as @safe. It will still compile if the code is explicitly marked @safe and nothing in fun prevents that.

Examples:

This example can be marked @safe with -preview=dip1000.

auto rcInt = safeRefCounted(5);
writeln(rcInt.borrow!(theInt => theInt)); // 5
auto sameInt = rcInt;
writeln(sameInt.borrow!"a"); // 5

// using `ref` in the function
auto arr = [0, 1, 2, 3, 4, 5, 6];
sameInt.borrow!(ref (x) => arr[x]) = 10;
writeln(arr); // [0, 1, 2, 3, 4, 10, 6]

// modifying the payload via an alias
sameInt.borrow!"a*=2";
writeln(rcInt.borrow!"a"); // 10

SafeRefCounted!(T, RefCountedAutoInitialize.no) safeRefCounted(T)(T val);

Initializes a SafeRefCounted with val. The template parameter T of SafeRefCounted is inferred from val. This function can be used to move non-copyable values to the heap. It also disables the autoInit option of SafeRefCounted.

Parameters:

T val The value to be reference counted

Returns:

An initialized SafeRefCounted containing val.

See Also:

refCounted C++'s make_shared

Examples:

static struct File
{
    static size_t nDestroyed;
    string name;
    @disable this(this); // not copyable
    ~this() { name = null; ++nDestroyed; }
}

auto file = File("name");
writeln(file.name); // "name"
// file cannot be copied and has unique ownership
static assert(!__traits(compiles, {auto file2 = file;}));

writeln(File.nDestroyed); // 0

// make the file ref counted to share ownership
// Note:
//   We write a compound statement (brace-delimited scope) in which all `SafeRefCounted!File` handles are created and deleted.
//   This allows us to see (after the scope) what happens after all handles have been destroyed.
{
    // We move the content of `file` to a separate (and heap-allocated) `File` object,
    // managed-and-accessed via one-or-multiple (initially: one) `SafeRefCounted!File` objects ("handles").
    // This "moving":
    //   (1) invokes `file`'s destructor (=> `File.nDestroyed` is incremented from 0 to 1 and `file.name` becomes `null`);
    //   (2) overwrites `file` with `File.init` (=> `file.name` becomes `null`).
    // It appears that writing `name = null;` in the destructor is redundant,
    // but please note that (2) is only performed if `File` defines a destructor (or post-blit operator),
    // and in the absence of the `nDestroyed` instrumentation there would have been no reason to define a destructor.
    import std.algorithm.mutation : move;
    auto rcFile = safeRefCounted(move(file));
    writeln(rcFile.name); // "name"
    writeln(File.nDestroyed); // 1
    writeln(file.name); // null

    // We create another `SafeRefCounted!File` handle to the same separate `File` object.
    // While any of the handles is still alive, the `File` object is kept alive (=> `File.nDestroyed` is not modified).
    auto rcFile2 = rcFile;
    writeln(rcFile.refCountedStore.refCount); // 2
    writeln(File.nDestroyed); // 1
}
// The separate `File` object is deleted when the last `SafeRefCounted!File` handle is destroyed
// (i.e. at the closing brace of the compound statement above, which destroys both handles: `rcFile` and `rcFile2`)
// (=> `File.nDestroyed` is incremented again, from 1 to 2):
writeln(File.nDestroyed); // 2

template Proxy(alias a)

Creates a proxy for the value a that will forward all operations while disabling implicit conversions. The aliased item a must be an lvalue. This is useful for creating a new type from the "base" type (though this is not a subtype-supertype relationship; the new type is not related to the old type in any way, by design).

The new type supports all operations that the underlying type does, including all operators such as +, --, <, [], etc.

Parameters:

a	The value to act as a proxy for all operations. It must be an lvalue.

Examples:

struct MyInt
{
    private int value;
    mixin Proxy!value;

    this(int n){ value = n; }
}

MyInt n = 10;

// Enable operations that original type has.
++n;
writeln(n); // 11
writeln(n * 2); // 22

void func(int n) { }

// Disable implicit conversions to original type.
//int x = n;
//func(n);

Examples:

The proxied value must be an lvalue.

struct NewIntType
{
    //Won't work; the literal '1'
    //is an rvalue, not an lvalue
    //mixin Proxy!1;

    //Okay, n is an lvalue
    int n;
    mixin Proxy!n;

    this(int n) { this.n = n; }
}

NewIntType nit = 0;
nit++;
writeln(nit); // 1


struct NewObjectType
{
    Object obj;
    //Ok, obj is an lvalue
    mixin Proxy!obj;

    this (Object o) { obj = o; }
}

NewObjectType not = new Object();
assert(__traits(compiles, not.toHash()));

Examples:

There is one exception to the fact that the new type is not related to the old type. Pseudo-member functions are usable with the new type; they will be forwarded on to the proxied value.

import std.math.traits : isInfinity;

float f = 1.0;
assert(!f.isInfinity);

struct NewFloat
{
    float _;
    mixin Proxy!_;

    this(float f) { _ = f; }
}

NewFloat nf = 1.0f;
assert(!nf.isInfinity);

struct Typedef(T, T init = T.init, string cookie = null);

Typedef allows the creation of a unique type which is based on an existing type. Unlike the alias feature, Typedef ensures the two types are not considered as equals.

Parameters:

init	Optional initial value for the new type.
cookie	Optional, used to create multiple unique types which are based on the same origin type T

Note If a library routine cannot handle the Typedef type, you can use the TypedefType template to extract the type which the Typedef wraps.

Examples:

alias MyInt = Typedef!int;
MyInt foo = 10;
foo++;
writeln(foo); // 11

Examples:

custom initialization values

alias MyIntInit = Typedef!(int, 42);
static assert(is(TypedefType!MyIntInit == int));
static assert(MyIntInit() == 42);

Examples:

Typedef creates a new type

alias MyInt = Typedef!int;
static void takeInt(int) {}
static void takeMyInt(MyInt) {}

int i;
takeInt(i);    // ok
static assert(!__traits(compiles, takeMyInt(i)));

MyInt myInt;
static assert(!__traits(compiles, takeInt(myInt)));
takeMyInt(myInt);  // ok

Examples:

Use the optional cookie argument to create different types of the same base type

alias TypeInt1 = Typedef!int;
alias TypeInt2 = Typedef!int;

// The two Typedefs are the same type.
static assert(is(TypeInt1 == TypeInt2));

alias MoneyEuros = Typedef!(float, float.init, "euros");
alias MoneyDollars = Typedef!(float, float.init, "dollars");

// The two Typedefs are _not_ the same type.
static assert(!is(MoneyEuros == MoneyDollars));

string toString(this T)(); void toString(this T, W)(ref W writer, ref scope const FormatSpec!char fmt) if (isOutputRange!(W, char));

Convert wrapped value to a human readable string

Examples:

import std.conv : to;

int i = 123;
auto td = Typedef!int(i);
writeln(i.to!string); // td.to!string

template TypedefType(T)

Get the underlying type which a Typedef wraps. If T is not a Typedef it will alias itself to T.

Examples:

import std.conv : to;

alias MyInt = Typedef!int;
static assert(is(TypedefType!MyInt == int));

/// Instantiating with a non-Typedef will return that type
static assert(is(TypedefType!int == int));

string num = "5";

// extract the needed type
MyInt myInt = MyInt( num.to!(TypedefType!MyInt) );
writeln(myInt); // 5

// cast to the underlying type to get the value that's being wrapped
int x = cast(TypedefType!MyInt) myInt;

alias MyIntInit = Typedef!(int, 42);
static assert(is(TypedefType!MyIntInit == int));
static assert(MyIntInit() == 42);

template scoped(T) if (is(T == class))

Allocates a class object right inside the current scope, therefore avoiding the overhead of new. This facility is unsafe; it is the responsibility of the user to not escape a reference to the object outside the scope.

The class destructor will be called when the result of scoped() is itself destroyed.

Scoped class instances can be embedded in a parent class or struct, just like a child struct instance. Scoped member variables must have type typeof(scoped!Class(args)), and be initialized with a call to scoped. See below for an example.

Note It's illegal to move a class instance even if you are sure there are no pointers to it. As such, it is illegal to move a scoped object.

Examples:

class A
{
    int x;
    this()     {x = 0;}
    this(int i){x = i;}
    ~this()    {}
}

// Standard usage, constructing A on the stack
auto a1 = scoped!A();
a1.x = 42;

// Result of `scoped` call implicitly converts to a class reference
A aRef = a1;
writeln(aRef.x); // 42

// Scoped destruction
{
    auto a2 = scoped!A(1);
    writeln(a2.x); // 1
    aRef = a2;
    // a2 is destroyed here, calling A's destructor
}
// aRef is now an invalid reference

// Here the temporary scoped A is immediately destroyed.
// This means the reference is then invalid.
version (Bug)
{
    // Wrong, should use `auto`
    A invalid = scoped!A();
}

// Restrictions
version (Bug)
{
    import std.algorithm.mutation : move;
    auto invalid = a1.move; // illegal, scoped objects can't be moved
}
static assert(!is(typeof({
    auto e1 = a1; // illegal, scoped objects can't be copied
    assert([a1][0].x == 42); // ditto
})));
static assert(!is(typeof({
    alias ScopedObject = typeof(a1);
    auto e2 = ScopedObject();  // illegal, must be built via scoped!A
    auto e3 = ScopedObject(1); // ditto
})));

// Use with alias
alias makeScopedA = scoped!A;
auto a3 = makeScopedA();
auto a4 = makeScopedA(1);

// Use as member variable
struct B
{
    typeof(scoped!A()) a; // note the trailing parentheses

    this(int i)
    {
        // construct member
        a = scoped!A(i);
    }
}

// Stack-allocate
auto b1 = B(5);
aRef = b1.a;
writeln(aRef.x); // 5
destroy(b1); // calls A's destructor for b1.a
// aRef is now an invalid reference

// Heap-allocate
auto b2 = new B(6);
writeln(b2.a.x); // 6
destroy(*b2); // calls A's destructor for b2.a

@system auto scoped(Args...)(auto ref Args args);

Returns the scoped object.

Parameters:

Args args Arguments to pass to T's constructor.

template Flag(string name)

Defines a simple, self-documenting yes/no flag. This makes it easy for APIs to define functions accepting flags without resorting to bool, which is opaque in calls, and without needing to define an enumerated type separately. Using Flag!"Name" instead of bool makes the flag's meaning visible in calls. Each yes/no flag has its own type, which makes confusions and mix-ups impossible.

Example Code calling getLine (usually far away from its definition) can't be understood without looking at the documentation, even by users familiar with the API:

string getLine(bool keepTerminator)
{
    ...
    if (keepTerminator) ...
    ...
}
...
auto line = getLine(false);

Assuming the reverse meaning (i.e. "ignoreTerminator") and inserting the wrong code compiles and runs with erroneous results.

After replacing the boolean parameter with an instantiation of Flag, code calling getLine can be easily read and understood even by people not fluent with the API:

string getLine(Flag!"keepTerminator" keepTerminator)
{
    ...
    if (keepTerminator) ...
    ...
}
...
auto line = getLine(Yes.keepTerminator);

The structs Yes and No are provided as shorthand for Flag!"Name".yes and Flag!"Name".no and are preferred for brevity and readability. These convenience structs mean it is usually unnecessary and counterproductive to create an alias of a Flag as a way of avoiding typing out the full type while specifying the affirmative or negative options.

Passing categorical data by means of unstructured bool parameters is classified under "simple-data coupling" by Steve McConnell in the Code Complete book, along with three other kinds of coupling. The author argues citing several studies that coupling has a negative effect on code quality. Flag offers a simple structuring method for passing yes/no flags to APIs.

Examples:

Flag!"abc" flag;

writeln(flag); // Flag!"abc".no
writeln(flag); // No.abc
assert(!flag);
if (flag) assert(0);

Examples:

auto flag = Yes.abc;

assert(flag);
writeln(flag); // Yes.abc
if (!flag) assert(0);
if (flag) {} else assert(0);

enum Flag: bool;

no: When creating a value of type Flag!"Name", use Flag!"Name".no for the negative option. When using a value of type Flag!"Name", compare it against Flag!"Name".no or just false or 0.
yes: When creating a value of type Flag!"Name", use Flag!"Name".yes for the affirmative option. When using a value of type Flag!"Name", compare it against Flag!"Name".yes.

struct Yes; struct No;

Convenience names that allow using e.g. Yes.encryption instead of Flag!"encryption".yes and No.encryption instead of Flag!"encryption".no.

Examples:

Flag!"abc" flag;

writeln(flag); // Flag!"abc".no
writeln(flag); // No.abc
assert(!flag);
if (flag) assert(0);

Examples:

auto flag = Yes.abc;

assert(flag);
writeln(flag); // Yes.abc
if (!flag) assert(0);
if (flag) {} else assert(0);

template isBitFlagEnum(E)

Detect whether an enum is of integral type and has only "flag" values (i.e. values with a bit count of exactly 1). Additionally, a zero value is allowed for compatibility with enums including a "None" value.

Examples:

enum A
{
    None,
    A = 1 << 0,
    B = 1 << 1,
    C = 1 << 2,
    D = 1 << 3,
}

static assert(isBitFlagEnum!A);

Examples:

Test an enum with default (consecutive) values

enum B
{
    A,
    B,
    C,
    D // D == 3
}

static assert(!isBitFlagEnum!B);

Examples:

Test an enum with non-integral values

enum C: double
{
    A = 1 << 0,
    B = 1 << 1
}

static assert(!isBitFlagEnum!C);

struct BitFlags(E, Flag!"unsafe" unsafe = No.unsafe) if (unsafe || isBitFlagEnum!E);

A typesafe structure for storing combinations of enum values.

This template defines a simple struct to represent bitwise OR combinations of enum values. It can be used if all the enum values are integral constants with a bit count of at most 1, or if the unsafe parameter is explicitly set to Yes. This is much safer than using the enum itself to store the OR combination, which can produce surprising effects like this:

enum E
{
    A = 1 << 0,
    B = 1 << 1
}
E e = E.A | E.B;
// will throw SwitchError
final switch (e)
{
    case E.A:
        return;
    case E.B:
        return;
}

Examples:

Set values with the | operator and test with &

enum Enum
{
    A = 1 << 0,
}

// A default constructed BitFlags has no value set
immutable BitFlags!Enum flags_empty;
assert(!flags_empty.A);

// Value can be set with the | operator
immutable flags_A = flags_empty | Enum.A;

// and tested using property access
assert(flags_A.A);

// or the & operator
assert(flags_A & Enum.A);
// which commutes.
assert(Enum.A & flags_A);

Examples:

A default constructed BitFlags has no value set

enum Enum
{
    None,
    A = 1 << 0,
    B = 1 << 1,
    C = 1 << 2
}

immutable BitFlags!Enum flags_empty;
assert(!(flags_empty & (Enum.A | Enum.B | Enum.C)));
assert(!(flags_empty & Enum.A) && !(flags_empty & Enum.B) && !(flags_empty & Enum.C));

Examples:

Binary operations: subtracting and intersecting flags

enum Enum
{
    A = 1 << 0,
    B = 1 << 1,
    C = 1 << 2,
}
immutable BitFlags!Enum flags_AB = BitFlags!Enum(Enum.A, Enum.B);
immutable BitFlags!Enum flags_BC = BitFlags!Enum(Enum.B, Enum.C);

// Use the ~ operator for subtracting flags
immutable BitFlags!Enum flags_B = flags_AB & ~BitFlags!Enum(Enum.A);
assert(!flags_B.A && flags_B.B && !flags_B.C);

// use & between BitFlags for intersection
writeln(flags_B); // (flags_BC & flags_AB)

Examples:

All the binary operators work in their assignment version

enum Enum
{
    A = 1 << 0,
    B = 1 << 1,
}

BitFlags!Enum flags_empty, temp, flags_AB;
flags_AB = Enum.A | Enum.B;

temp |= flags_AB;
writeln(temp); // (flags_empty | flags_AB)

temp = flags_empty;
temp |= Enum.B;
writeln(temp); // (flags_empty | Enum.B)

temp = flags_empty;
temp &= flags_AB;
writeln(temp); // (flags_empty & flags_AB)

temp = flags_empty;
temp &= Enum.A;
writeln(temp); // (flags_empty & Enum.A)

Examples:

Conversion to bool and int

enum Enum
{
    A = 1 << 0,
    B = 1 << 1,
}

BitFlags!Enum flags;

// BitFlags with no value set evaluate to false
assert(!flags);

// BitFlags with at least one value set evaluate to true
flags |= Enum.A;
assert(flags);

// This can be useful to check intersection between BitFlags
BitFlags!Enum flags_AB = Enum.A | Enum.B;
assert(flags & flags_AB);
assert(flags & Enum.A);

// You can of course get you raw value out of flags
auto value = cast(int) flags;
writeln(value); // Enum.A

Examples:

You need to specify the unsafe parameter for enums with custom values

enum UnsafeEnum
{
    A = 1,
    B = 2,
    C = 4,
    BC = B|C
}
static assert(!__traits(compiles, { BitFlags!UnsafeEnum flags; }));
BitFlags!(UnsafeEnum, Yes.unsafe) flags;

// property access tests for exact match of unsafe enums
flags.B = true;
assert(!flags.BC); // only B
flags.C = true;
assert(flags.BC); // both B and C
flags.B = false;
assert(!flags.BC); // only C

// property access sets all bits of unsafe enum group
flags = flags.init;
flags.BC = true;
assert(!flags.A && flags.B && flags.C);
flags.A = true;
flags.BC = false;
assert(flags.A && !flags.B && !flags.C);

template ReplaceType(From, To, T...)

Replaces all occurrences of From into To, in one or more types T. For example, ReplaceType!(int, uint, Tuple!(int, float)[string]) yields Tuple!(uint, float)[string]. The types in which replacement is performed may be arbitrarily complex, including qualifiers, built-in type constructors (pointers, arrays, associative arrays, functions, and delegates), and template instantiations; replacement proceeds transitively through the type definition. However, member types in structs or classes are not replaced because there are no ways to express the types resulting after replacement.

This is an advanced type manipulation necessary e.g. for replacing the placeholder type This in std.variant.Algebraic.

Returns:

ReplaceType aliases itself to the type(s) that result after replacement.

Examples:

static assert(
    is(ReplaceType!(int, string, int[]) == string[]) &&
    is(ReplaceType!(int, string, int[int]) == string[string]) &&
    is(ReplaceType!(int, string, const(int)[]) == const(string)[]) &&
    is(ReplaceType!(int, string, Tuple!(int[], float))
        == Tuple!(string[], float))
);

template ReplaceTypeUnless(alias pred, From, To, T...)

Like ReplaceType, but does not perform replacement in types for which pred evaluates to true.

Examples:

import std.traits : isArray;

static assert(
    is(ReplaceTypeUnless!(isArray, int, string, int*) == string*) &&
    is(ReplaceTypeUnless!(isArray, int, string, int[]) == int[]) &&
    is(ReplaceTypeUnless!(isArray, int, string, Tuple!(int, int[]))
        == Tuple!(string, int[]))
);

struct Ternary;

Ternary type with three truth values:

Ternary.yes for true
Ternary.no for false
Ternary.unknown as an unknown state

Also known as trinary, trivalent, or trilean.

See Also:

Three Valued Logic on Wikipedia

Examples:

Ternary a;
writeln(a); // Ternary.unknown

writeln(~Ternary.yes); // Ternary.no
writeln(~Ternary.no); // Ternary.yes
writeln(~Ternary.unknown); // Ternary.unknown

enum Ternary no; enum Ternary yes; enum Ternary unknown;

The possible states of the Ternary

pure nothrow @nogc @safe this(bool b); pure nothrow @nogc @safe void opAssign(bool b);

Construct and assign from a bool, receiving no for false and yes for true.

pure nothrow @nogc @safe this(const Ternary b);

Construct a ternary value from another ternary value

Ternary opUnary(string s)() if (s == "~"); Ternary opBinary(string s)(Ternary rhs) if (s == "|"); Ternary opBinary(string s)(Ternary rhs) if (s == "&"); Ternary opBinary(string s)(Ternary rhs) if (s == "^"); Ternary opBinary(string s)(bool rhs) if (s == "|" || s == "&" || s == "^");

Truth table for logical operations
a	b	˜a	a \| b	a & b	a ^ b
no	no	yes	no	no	no
no	yes		yes	no	yes
no	unknown		unknown	no	unknown
yes	no	no	yes	no	yes
yes	yes		yes	yes	no
yes	unknown		yes	unknown	unknown
unknown	no	unknown	unknown	no	unknown
unknown	yes		yes	unknown	unknown
unknown	unknown		unknown	unknown	unknown

struct RefCounted(T, RefCountedAutoInitialize autoInit = RefCountedAutoInitialize.yes);

The old version of SafeRefCounted, before borrow existed. Old code may be relying on @safety of some of the member functions which cannot be safe in the new scheme, and can avoid breakage by continuing to use this. SafeRefCounted should be preferred, as this type is outdated and unrecommended for new code.

Examples:

auto rc1 = RefCounted!int(5);
writeln(rc1); // 5
auto rc2 = rc1;
rc2 = 42;
writeln(rc1); // 42

RefCounted!(T, RefCountedAutoInitialize.no) refCounted(T)(T val);

Like safeRefCounted but used to initialize RefCounted instead. Intended for backwards compatibility, otherwise it is preferable to use safeRefCounted.

Examples:

static struct File
{
    static size_t nDestroyed;
    string name;
    @disable this(this); // not copyable
    ~this() { name = null; ++nDestroyed; }
}

auto file = File("name");
writeln(file.name); // "name"
static assert(!__traits(compiles, {auto file2 = file;}));
writeln(File.nDestroyed); // 0

{
    import std.algorithm.mutation : move;
    auto rcFile = refCounted(move(file));
    writeln(rcFile.name); // "name"
    writeln(File.nDestroyed); // 1
    writeln(file.name); // null

    auto rcFile2 = rcFile;
    writeln(rcFile.refCountedStore.refCount); // 2
    writeln(File.nDestroyed); // 1
}

writeln(File.nDestroyed); // 2

Library Reference

std.typecons