Tuples​

Constructing tuples​

A tuple constructor is an expression that consists of a sequence of comma-separated expressions enclosed in parentheses. It produces a tuple value:

"(" expr [, ... ] ")"

Declare a named tuple:

"(" identifier := expr [, ... ] ")"

All elements in a named tuple must have a name.

A tuple constructor automatically creates a corresponding tuple type.

Accessing elements​

An element of a tuple can be referenced in the form:

expr.element-index

Here, expr is any expression that has a tuple type, and element-index is either the zero-based index of the element or the name of an element in a named tuple.

Examples:

db> 

select (1, 'EdgeDB').0;

{1}

db> 

select (number := 1, name := 'EdgeDB').name;

{"EdgeDB"}

db> 

select (number := 1, name := 'EdgeDB').1;

{"EdgeDB"}

Nesting tuples​

Tuples can be nested:

db> 

select (nested_tuple := (1, 2)).nested_tuple.0;

{1}

Referencing a non-existent tuple element will result in an error:

db> 

select (1, 2).5;

EdgeQLError: 5 is not a member of a tuple

---- query context ----

    line 1
        > select (1, 2).3;

Type syntax​

A tuple type can be explicitly declared in an expression or schema declaration using the following syntax:

tuple "<" element-type, [element-type, ...] ">"

A named tuple:

tuple "<" element-name : element-type [, ... ] ">"

Any type can be used as a tuple element type.

Here’s an example of using this syntax in a schema definition:

type GameElement {
    required name: str;
    required position: tuple<x: int64, y: int64>;
}

Here’s a few examples of using tuple types in EdgeQL queries:

db> 

select <tuple<int64, str>>('1', 3);

{(1, '3')}

db> 

select <tuple<x: int64, y: int64>>(1, 2);

{(x := 1, y := 2)}

db> 

select (1, '3') is (tuple<int64, str>);

{true}

db> 

select ([1, 2], 'a') is (tuple<array<int64>, str>);

{true}

db> 

select ('foo', 42);

{('foo', 42)}

db> 

select (1, 2, 3) = (a := 1, b := 2, c := 3);

{true}