แก้ไข

แชร์ผ่าน


Tuple types (C# reference)

The tuples feature provides concise syntax to group multiple data elements in a lightweight data structure. The following example shows how you can declare a tuple variable, initialize it, and access its data members:

(double, int) t1 = (4.5, 3);
Console.WriteLine($"Tuple with elements {t1.Item1} and {t1.Item2}.");
// Output:
// Tuple with elements 4.5 and 3.

(double Sum, int Count) t2 = (4.5, 3);
Console.WriteLine($"Sum of {t2.Count} elements is {t2.Sum}.");
// Output:
// Sum of 3 elements is 4.5.

As the preceding example shows, to define a tuple type, you specify types of all its data members and, optionally, the field names. You can't define methods in a tuple type, but you can use the methods provided by .NET, as the following example shows:

(double, int) t = (4.5, 3);
Console.WriteLine(t.ToString());
Console.WriteLine($"Hash code of {t} is {t.GetHashCode()}.");
// Output:
// (4.5, 3)
// Hash code of (4.5, 3) is 718460086.

Tuple types support equality operators == and !=. For more information, see the Tuple equality section.

Tuple types are value types; tuple elements are public fields. That makes tuples mutable value types.

You can define tuples with an arbitrary large number of elements:

var t =
(1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
11, 12, 13, 14, 15, 16, 17, 18,
19, 20, 21, 22, 23, 24, 25, 26);
Console.WriteLine(t.Item26);  // output: 26

Use cases of tuples

One of the most common use cases of tuples is as a method return type. That is, instead of defining out method parameters, you can group method results in a tuple return type, as the following example shows:

int[] xs = new int[] { 4, 7, 9 };
var limits = FindMinMax(xs);
Console.WriteLine($"Limits of [{string.Join(" ", xs)}] are {limits.min} and {limits.max}");
// Output:
// Limits of [4 7 9] are 4 and 9

int[] ys = new int[] { -9, 0, 67, 100 };
var (minimum, maximum) = FindMinMax(ys);
Console.WriteLine($"Limits of [{string.Join(" ", ys)}] are {minimum} and {maximum}");
// Output:
// Limits of [-9 0 67 100] are -9 and 100

(int min, int max) FindMinMax(int[] input)
{
    if (input is null || input.Length == 0)
    {
        throw new ArgumentException("Cannot find minimum and maximum of a null or empty array.");
    }

    // Initialize min to MaxValue so every value in the input
    // is less than this initial value.
    var min = int.MaxValue;
    // Initialize max to MinValue so every value in the input
    // is greater than this initial value.
    var max = int.MinValue;
    foreach (var i in input)
    {
        if (i < min)
        {
            min = i;
        }
        if (i > max)
        {
            max = i;
        }
    }
    return (min, max);
}

As the preceding example shows, you can work with the returned tuple instance directly or deconstruct it in separate variables.

You can also use tuple types instead of anonymous types; for example, in LINQ queries. For more information, see Choosing between anonymous and tuple types.

Typically, you use tuples to group loosely related data elements. In public APIs, consider defining a class or a structure type.

Tuple field names

You explicitly specify tuple fields names in a tuple initialization expression or in the definition of a tuple type, as the following example shows:

var t = (Sum: 4.5, Count: 3);
Console.WriteLine($"Sum of {t.Count} elements is {t.Sum}.");

(double Sum, int Count) d = (4.5, 3);
Console.WriteLine($"Sum of {d.Count} elements is {d.Sum}.");

If you don't specify a field name, it may be inferred from the name of the corresponding variable in a tuple initialization expression, as the following example shows:

var sum = 4.5;
var count = 3;
var t = (sum, count);
Console.WriteLine($"Sum of {t.count} elements is {t.sum}.");

That's called tuple projection initializers. The name of a variable isn't projected onto a tuple field name in the following cases:

  • The candidate name is a member name of a tuple type, for example, Item3, ToString, or Rest.
  • The candidate name is a duplicate of another tuple field name, either explicit or implicit.

In the preceding cases, you either explicitly specify the name of a field or access a field by its default name.

The default names of tuple fields are Item1, Item2, Item3 and so on. You can always use the default name of a field, even when a field name is specified explicitly or inferred, as the following example shows:

var a = 1;
var t = (a, b: 2, 3);
Console.WriteLine($"The 1st element is {t.Item1} (same as {t.a}).");
Console.WriteLine($"The 2nd element is {t.Item2} (same as {t.b}).");
Console.WriteLine($"The 3rd element is {t.Item3}.");
// Output:
// The 1st element is 1 (same as 1).
// The 2nd element is 2 (same as 2).
// The 3rd element is 3.

Tuple assignment and tuple equality comparisons don't take field names into account.

At compile time, the compiler replaces non-default field names with the corresponding default names. As a result, explicitly specified or inferred field names aren't available at run time.

Tip

Enable .NET code style rule IDE0037 to set a preference on inferred or explicit tuple field names.

Beginning with C# 12, you can specify an alias for a tuple type with a using directive. The following example adds a global using alias for a tuple type with two integer values for an allowed Min and Max value:

global using BandPass = (int Min, int Max);

After declaring the alias, you can use the BandPass name as an alias for that tuple type:

BandPass bracket = (40, 100);
Console.WriteLine($"The bandpass filter is {bracket.Min} to {bracket.Max}");

An alias doesn't introduce a new type, but only creates a synonym for an existing type. You can deconstruct a tuple declared with the BandPass alias the same as you can with its underlying tuple type:

(int a , int b) = bracket;
Console.WriteLine($"The bracket is {a} to {b}");

As with tuple assignment or deconstruction, the tuple member names don't need to match; the types do.

Similarly, a second alias with the same arity and member types can be used interchangeably with the original alias. You could declare a second alias:

using Range = (int Minimum, int Maximum);

You can assign a Range tuple to a BandPass tuple. As with all tuple assignment, the field names need not match, only the types and the arity.

Range r = bracket;
Console.WriteLine($"The range is {r.Minimum} to {r.Maximum}");

An alias for a tuple type provides more semantic information when you use tuples. It doesn't introduce a new type. To provide type safety, you should declare a positional record instead.

Tuple assignment and deconstruction

C# supports assignment between tuple types that satisfy both of the following conditions:

  • both tuple types have the same number of elements
  • for each tuple position, the type of the right-hand tuple element is the same as or implicitly convertible to the type of the corresponding left-hand tuple element

Tuple element values are assigned following the order of tuple elements. The names of tuple fields are ignored and not assigned, as the following example shows:

(int, double) t1 = (17, 3.14);
(double First, double Second) t2 = (0.0, 1.0);
t2 = t1;
Console.WriteLine($"{nameof(t2)}: {t2.First} and {t2.Second}");
// Output:
// t2: 17 and 3.14

(double A, double B) t3 = (2.0, 3.0);
t3 = t2;
Console.WriteLine($"{nameof(t3)}: {t3.A} and {t3.B}");
// Output:
// t3: 17 and 3.14

You can also use the assignment operator = to deconstruct a tuple instance in separate variables. You can do that in many ways:

  • Use the var keyword outside the parentheses to declare implicitly typed variables and let the compiler infer their types:

    var t = ("post office", 3.6);
    var (destination, distance) = t;
    Console.WriteLine($"Distance to {destination} is {distance} kilometers.");
    // Output:
    // Distance to post office is 3.6 kilometers.
    
  • Explicitly declare the type of each variable inside parentheses:

    var t = ("post office", 3.6);
    (string destination, double distance) = t;
    Console.WriteLine($"Distance to {destination} is {distance} kilometers.");
    // Output:
    // Distance to post office is 3.6 kilometers.
    
  • Declare some types explicitly and other types implicitly (with var) inside the parentheses:

    var t = ("post office", 3.6);
    (var destination, double distance) = t;
    Console.WriteLine($"Distance to {destination} is {distance} kilometers.");
    // Output:
    // Distance to post office is 3.6 kilometers.
    
  • Use existing variables:

    var destination = string.Empty;
    var distance = 0.0;
    
    var t = ("post office", 3.6);
    (destination, distance) = t;
    Console.WriteLine($"Distance to {destination} is {distance} kilometers.");
    // Output:
    // Distance to post office is 3.6 kilometers.
    

The destination of a deconstruct expression can include both existing variables and variables declared in the deconstruction declaration.

You can also combine deconstruction with pattern matching to inspect the characteristics of fields in a tuple. The following example loops through several integers and prints those that are divisible by 3. It deconstructs the tuple result of Int32.DivRem and matches against a Remainder of 0:

for (int i = 4; i < 20;  i++)
{
    if (Math.DivRem(i, 3) is ( Quotient: var q, Remainder: 0 ))
    {
        Console.WriteLine($"{i} is divisible by 3, with quotient {q}");
    }
}

For more information about deconstruction of tuples and other types, see Deconstructing tuples and other types.

Tuple equality

Tuple types support the == and != operators. These operators compare members of the left-hand operand with the corresponding members of the right-hand operand following the order of tuple elements.

(int a, byte b) left = (5, 10);
(long a, int b) right = (5, 10);
Console.WriteLine(left == right);  // output: True
Console.WriteLine(left != right);  // output: False

var t1 = (A: 5, B: 10);
var t2 = (B: 5, A: 10);
Console.WriteLine(t1 == t2);  // output: True
Console.WriteLine(t1 != t2);  // output: False

As the preceding example shows, the == and != operations don't take into account tuple field names.

Two tuples are comparable when both of the following conditions are satisfied:

  • Both tuples have the same number of elements. For example, t1 != t2 doesn't compile if t1 and t2 have different numbers of elements.
  • For each tuple position, the corresponding elements from the left-hand and right-hand tuple operands are comparable with the == and != operators. For example, (1, (2, 3)) == ((1, 2), 3) doesn't compile because 1 isn't comparable with (1, 2).

The == and != operators compare tuples in short-circuiting way. That is, an operation stops as soon as it meets a pair of non equal elements or reaches the ends of tuples. However, before any comparison, all tuple elements are evaluated, as the following example shows:

Console.WriteLine((Display(1), Display(2)) == (Display(3), Display(4)));

int Display(int s)
{
    Console.WriteLine(s);
    return s;
}
// Output:
// 1
// 2
// 3
// 4
// False

Tuples as out parameters

Typically, you refactor a method that has out parameters into a method that returns a tuple. However, there are cases in which an out parameter can be of a tuple type. The following example shows how to work with tuples as out parameters:

var limitsLookup = new Dictionary<int, (int Min, int Max)>()
{
    [2] = (4, 10),
    [4] = (10, 20),
    [6] = (0, 23)
};

if (limitsLookup.TryGetValue(4, out (int Min, int Max) limits))
{
    Console.WriteLine($"Found limits: min is {limits.Min}, max is {limits.Max}");
}
// Output:
// Found limits: min is 10, max is 20

Tuples vs System.Tuple

C# tuples, which are backed by System.ValueTuple types, are different from tuples that are represented by System.Tuple types. The main differences are as follows:

  • System.ValueTuple types are value types. System.Tuple types are reference types.
  • System.ValueTuple types are mutable. System.Tuple types are immutable.
  • Data members of System.ValueTuple types are fields. Data members of System.Tuple types are properties.

C# language specification

For more information, see:

See also