Type

These functions can be used for generating and coercing data to specific data types. These functions are useful when accepting input values in client libraries, and ensuring that they are the desired type within SQL statements.

type::array

The type::array function converts a value into an array.

type::array(array|range) -> array

The following example shows this function, and its output:

type::array(1..=3);

[1, 2, 3]

This is the equivalent of using <array> to cast a value to an array.

type::bool

The type::bool function converts a value into a boolean.

type::bool(bool|string) -> bool

The following example shows this function, and its output:

type::bool("true");

true

This is the equivalent of using <bool> to cast a value to a boolean.

type::bytes

The type::bytes function converts a value into bytes.

type::bytes(bytes|string) -> bool

The following example shows this function, and its output:

type::bytes("A few bytes");

b"4120666577206279746573"

This is the equivalent of using <bytes> to cast a value to bytes.

type::date

Available since: v4.0.0

The type::date function converts a value into a date.

type::date(date|datetime|timestamp|string) -> date

The following example shows this function, and its output:

[type::date('2026-10-27'), type::date(d'2026-10-27T12:30:45Z')];

[pd'2026-10-27', pd'2026-10-27']

A datetime is converted to its date in UTC, so the time of day is dropped.

This is the equivalent of using <date> to cast a value. The method form is .to_date().

type::datetime

The type::datetime function converts a value into a datetime.

type::datetime(datetime|string) -> datetime

The following example shows this function, and its output:

type::datetime("2022-04-27T18:12:27+00:00");

d'2022-04-27T18:12:27Z'

This is the equivalent of using <datetime> to cast a value to a datetime.

type::decimal

The type::decimal function converts a value into a decimal.

type::decimal(decimal|float|int|number|string) -> decimal

The following example shows this function, and its output:

type::decimal("12345");

12345dec

This is the equivalent of using <decimal> to cast a value to a decimal.

type::duration

The type::duration function converts a value into a duration.

type::duration(duration|string) -> duration

The following example shows this function, and its output:

type::duration("4h");

4h

This is the equivalent of using <duration> to cast a value to a duration.

type::f16

Available since: v4.0.0

The type::f16 function converts a value into a 16-bit floating point number. A value that is out of range for f16 is refused.

type::f16(decimal|float|int|number|string) -> f16

The following example shows this function, and its output:

type::f16(1.5);

1.5f16

This is the equivalent of using <f16> to cast a value.

type::f32

Available since: v4.0.0

The type::f32 function converts a value into a 32-bit floating point number. A value that is out of range for f32 is refused.

type::f32(decimal|float|int|number|string) -> f32

The following example shows this function, and its output:

type::f32(1.5);

1.5f32

This is the equivalent of using <f32> to cast a value.

type::field

The type::field function projects a single field within a SELECT statement.

type::field(string)

The following example shows this function, and its output:

CREATE person:test SET title = 'Mr',
  name.first = 'Tobie',
  name.last = 'Morgan Hitchcock';

LET $param = 'name.first';

SELECT type::field($param), type::field('name.last') FROM person;

SELECT VALUE { 'firstname': type::field($param),
  lastname: type::field('name.last') } FROM person;

SELECT VALUE [type::field($param),
  type::field('name.last')] FROM person;

[
	{
		id: person:test,
		title: 'Mr',
		name: {
			first: 'Tobie',
			last: 'Morgan Hitchcock',
	    }
	}
]

From SurrealDB 3.0, this function can be used after the OMIT clause of a SELECT statement.

LET $omit = "id";
CREATE person SET name = "Galen", surname = "Pathwarden", age = 19;
SELECT * OMIT type::field($omit) FROM person;

[
	{
		age: 19,
		name: 'Galen',
		surname: 'Pathwarden'
	}
]

type::fields

The type::fields function projects one or more fields within a SELECT statement.

type::fields(array<string>)

The following example shows this function, and its output:

CREATE person:test SET title = 'Mr',
  name.first = 'Tobie',
  name.last = 'Morgan Hitchcock';

LET $param = ['name.first', 'name.last'];

SELECT type::fields($param), type::fields(['title']) FROM person;

SELECT VALUE { 'names': type::fields($param) } FROM person;

SELECT VALUE type::fields($param) FROM person;

[
	{
		id: person:test,
		title: 'Mr',
		name: {
			first: 'Tobie',
			last: 'Morgan Hitchcock',
		}
	}
]

From SurrealDB 3.0, this function can be used after the OMIT clause of a SELECT statement.

LET $omit = ["id", "age"];
CREATE person SET name = "Galen", surname = "Pathwarden", age = 19;
SELECT * OMIT type::fields($omit) FROM person;

[
	{
		name: 'Galen',
		surname: 'Pathwarden'
	}
]

type::file

Available since: v3.0.0

The type::file function converts two strings representing a bucket name and a key into a file pointer.

type::file($bucket: string, $key: string) -> file

An example of a file pointer created using this function:

type::file("my_bucket", "file_name")

f"my_bucket:/file_name"

The following query shows the equivalent file pointer when created using the f prefix:

type::file("my_bucket", "file_name") == f"my_bucket:/file_name";

true

Once a bucket has been defined, operations using one of the file functions can be performed on the file pointer.

DEFINE BUCKET my_bucket BACKEND "memory";

type::file("my_bucket", "file_name").put("Some data inside");
type::file("my_bucket", "file_name").get();

b"536F6D65206461746120696E73696465"

type::float

The type::float function converts a value into a float.

type::float(decimal|float|int|number|string) -> float

The following example shows this function, and its output:

type::float("12345");

12345f

This is the equivalent of using <float> to cast a value to a float.

type::i8

Available since: v4.0.0

The type::i8 function converts a value into an 8-bit signed integer. A value that is out of range for i8, or that has a fractional part, is refused.

type::i8(decimal|float|int|number|string) -> i8

The following example shows this function, and its output:

type::i8('-128');

-128i8

This is the equivalent of using <i8> to cast a value.

type::i16

Available since: v4.0.0

The type::i16 function converts a value into a 16-bit signed integer. A value that is out of range for i16, or that has a fractional part, is refused.

type::i16(decimal|float|int|number|string) -> i16

The following example shows this function, and its output:

type::i16(1000);

1000i16

This is the equivalent of using <i16> to cast a value.

type::i32

Available since: v4.0.0

The type::i32 function converts a value into a 32-bit signed integer. A value that is out of range for i32, or that has a fractional part, is refused.

type::i32(decimal|float|int|number|string) -> i32

The following example shows this function, and its output:

type::i32(100000);

100000i32

This is the equivalent of using <i32> to cast a value.

type::int

The type::int function converts a value into an integer.

type::int(decimal|float|int|number|string) -> int

The following example shows this function, and its output:

type::int("12345");

12345

This is the equivalent of using <int> to cast a value to an int.

type::interval

Available since: v4.0.0

The type::interval function converts a value into an interval.

type::interval(interval|duration|time|string) -> interval

The following example shows this function, and its output:

[type::interval('2 days'), type::interval(1h30m)];

[ivl'P2D', ivl'PT1H30M']

This is the equivalent of using <interval> to cast a value. The method form is .to_interval().

type::number

The type::number function converts a value into a number.

type::number(decimal|float|int|number|string) -> number

The following example shows this function, and its output:

type::number("12345");

12345

This is the equivalent of using <number> to cast a value to a number.

type:of

Available since: v3.0.0

The type::of function returns a string denoting the type of a value.

type::of(value) -> string

type::of(2022dec);        -- 'decimal';
type::of(["some", 9]);    -- 'array';
type::of((50.0, 9.9));    -- 'geometry<point>'

type::point

The type::point function converts a value into a geometry point.

type::point(array|point) -> point

The following example shows this function, and its output:

type::point([ 51.509865, -0.118092 ]);

(51.509865, -0.118092)

type::range

The type::range function converts a value into a range. It accepts a single argument, either a range or an array with two values. If the argument is an array, it will be converted into a range, similar to casting.

type::range(range|array) -> range<record>

The following example shows this function, and its output:

type::range([1, 2]);
//- 1..2

type::range(1..10);
//- 1..10

type::range([1,9,4]);
//- 'Expected a range but cannot convert [1, 9, 4] into a range'

type::record

type::record($table: any, $key: any) -> record

LET $tb = "person";
LET $id = "tobie";
RETURN type::record($tb, $id);

FOR $data IN [
	{
		id: 9,
		name: 'Billy'
	},
	{
		id: 10,
		name: 'Bobby'
	}
] {
	UPSERT type::record('person', $data.id) CONTENT $data;
};

FOR $data IN [
	{
		name: 'Billy',
		num: 9
	},
    {
		name: 'Bobby',
		num: 10
	},
].map(|$o| {
    id: $o.num,
    name: $o.name
}) {
    UPSERT type::record('person', $data.id) CONTENT $data;
};

type::record("person", person:mat);

person:mat

type::record(person:tobie, 'person'); -- person:tobie
type::record(person:tobie, 'cat');

error: record is not in table `cat`

type::string

The type::string function converts any value except NONE, NULL, and bytes into a string.

type::string(any) -> string

The following example shows this function, and its output:

type::string(12345);

'12345'

This is the equivalent of using <string> to cast a value to a string.

type::string_lossy

Available since: v3.0.0

The type::string_lossy function converts any value except NONE, NULL, and bytes into a string. In the case of bytes, it will not return an error if the bytes are not valid UTF-8. Instead, invalid bytes will be replaced with the character � (U+FFFD REPLACEMENT CHARACTER, used in Unicode to represent a decoding error).

type::string(any) -> string

The following example shows this function, and its output:

-- Contains some invalid bytes
type::string_lossy(<bytes>[83,
  117,
  114,
  255,
  114,
  101,
  97,
  254,
  108,
  68,
  66]);
-- valid bytes
type::string_lossy(<bytes>[ 83,
  117,
  114,
  114,
  101,
  97,
  108,
  68,
  66 ]);

-------- Query --------

'Sur�rea�lDB'

-------- Query --------

'SurrealDB'

This is similar to using <string> to cast a value to a string, except that an input of bytes will not fail.

type::table

The type::table function converts a value into a table name.

type::table(record|string) -> string

The following example shows this function, and its output:

[
  type::table("person"),
  type::table(cat:one)
];

[person, cat]

As of version 2.0, SurrealDB no longer eagerly parses strings into record IDs. As such, the output of the last item ("dog:two") in the following example will differ. In version 1.x, it will be eagerly parsed into a record ID after which the dog table name will be returned, while in later editions it will be treated as a string and converted into the table name dog:two. As of version 3.0, a number is no longer accepted as input, because a number on its own is not a valid table name.

[
  type::table(cat:one),
  type::table("dog"),
  type::table("dog:two"),
];

[
	cat,
	dog,
	`dog:two`
]

type::time

Available since: v4.0.0

The type::time function converts a value into a time of day.

type::time(time|datetime|timestamp|string) -> time

The following example shows this function, and its output:

[type::time('12:30:00'), type::time(d'2026-10-27T12:30:45Z')];

[pt'12:30:00', pt'12:30:45']

This is the equivalent of using <time> to cast a value. The method form is .to_time().

type::timestamp

Available since: v4.0.0

The type::timestamp function converts a value into a timestamp.

type::timestamp(timestamp|date|datetime|string) -> timestamp

The following example shows this function, and its output:

[type::timestamp('2026-10-27T12:30:00'), type::timestamp(pd'2026-10-27')];

[ts'2026-10-27T12:30:00', ts'2026-10-27T00:00:00']

This is the equivalent of using <timestamp> to cast a value. The method form is .to_timestamp().

type::u8

Available since: v4.0.0

The type::u8 function converts a value into an 8-bit unsigned integer. A value that is out of range for u8, or that has a fractional part, is refused.

type::u8(decimal|float|int|number|string) -> u8

The following example shows this function, and its output:

type::u8(255);

255u8

This is the equivalent of using <u8> to cast a value.

type::u16

Available since: v4.0.0

The type::u16 function converts a value into a 16-bit unsigned integer. A value that is out of range for u16, or that has a fractional part, is refused.

type::u16(decimal|float|int|number|string) -> u16

The following example shows this function, and its output:

type::u16(1000);

1000u16

This is the equivalent of using <u16> to cast a value.

type::u32

Available since: v4.0.0

The type::u32 function converts a value into a 32-bit unsigned integer. A value that is out of range for u32, or that has a fractional part, is refused.

type::u32(decimal|float|int|number|string) -> u32

The following example shows this function, and its output:

type::u32(100000);

100000u32

This is the equivalent of using <u32> to cast a value.

type::u64

Available since: v4.0.0

The type::u64 function converts a value into a 64-bit unsigned integer. A value that is out of range for u64, or that has a fractional part, is refused.

type::u64(decimal|float|int|number|string) -> u64

The following example shows this function, and its output:

type::u64('18446744073709551615');

18446744073709551615u64

This is the equivalent of using <u64> to cast a value.

type::uuid

The type::uuid function converts a value into a UUID.

type::uuid(string|uuid) -> uuid

The following example shows this function, and its output:

type::uuid("0191f946-936f-7223-bef5-aebbc527ad80");

u'0191f946-936f-7223-bef5-aebbc527ad80'

type::is_array

The type::is_array function checks if the passed value is of type array.

type::is_array(any) -> bool

The following example shows this function, and its output:

type::is_array([ 'a', 'b', 'c' ]);

true

type::is_bool

The type::is_bool function checks if the passed value is of type bool.

type::is_bool(any) -> bool

The following example shows this function, and its output:

type::is_bool(true);

true

type::is_bytes

The type::is_bytes function checks if the passed value is of type bytes.

type::is_bytes(any) -> bool

The following example shows this function, and its output:

type::is_bytes("I am not bytes");

false

type::is_collection

The type::is_collection function checks if the passed value is of type collection.

type::is_collection(any) -> bool

The following example shows this function, and its output:

type::is_collection("I am not a collection");

false

type::is_date

Available since: v4.0.0

The type::is_date function checks if the passed value is of type date.

type::is_date(any) -> bool

The following example shows this function, and its output:

[type::is_date(pd'2026-10-27'), type::is_date(d'2026-10-27')];

[true, false]

type::is_datetime

The type::is_datetime function checks if the passed value is of type datetime.

type::is_datetime(any) -> bool

The following example shows this function, and its output:

type::is_datetime(time::now());

true

type::is_decimal

The type::is_decimal function checks if the passed value is of type decimal.

type::is_decimal(any) -> bool

The following example shows this function, and its output:

type::is_decimal(<decimal>
  13.5719384719384719385639856394139476937756394756);

true

type::is_duration

The type::is_duration function checks if the passed value is of type duration.

type::is_duration(any) -> bool

The following example shows this function, and its output:

type::is_duration('1970-01-01T00:00:00');

false

type::is_f16

Available since: v4.0.0

The type::is_f16 function checks if the passed value is of type f16.

type::is_f16(any) -> bool

The following example shows this function, and its output:

type::is_f16(1.5f16);

true

type::is_f32

Available since: v4.0.0

The type::is_f32 function checks if the passed value is of type f32.

type::is_f32(any) -> bool

The following example shows this function, and its output:

type::is_f32(1.5f32);

true

type::is_float

The type::is_float function checks if the passed value is of type float. From SurrealDB 4.0, a sized float such as 1.5f32 is not of type float, so this function returns false for it. Use type::is_number() to accept a number of any type.

type::is_float(any) -> bool

The following example shows this function, and its output:

type::is_float(<float> 41.5);

true

type::is_geometry

The type::is_geometry function checks if the passed value is of type geometry.

type::is_geometry(any) -> bool

The following example shows this function, and its output:

type::is_geometry((-0.118092, 51.509865));

true

type::is_i8

Available since: v4.0.0

The type::is_i8 function checks if the passed value is of type i8.

type::is_i8(any) -> bool

The following example shows this function, and its output:

type::is_i8(5i8);

true

type::is_i16

Available since: v4.0.0

The type::is_i16 function checks if the passed value is of type i16.

type::is_i16(any) -> bool

The following example shows this function, and its output:

type::is_i16(5i16);

true

type::is_i32

Available since: v4.0.0

The type::is_i32 function checks if the passed value is of type i32.

type::is_i32(any) -> bool

The following example shows this function, and its output:

type::is_i32(5i32);

true

type::is_int

The type::is_int function checks if the passed value is of type int. From SurrealDB 4.0, a sized integer such as 5u8 or 5u64 is not of type int, so this function returns false for it. Use type::is_number() to accept a number of any type.

type::is_int(any) -> bool

The following example shows this function, and its output:

type::is_int(<int> 123);

true

type::is_interval

Available since: v4.0.0

The type::is_interval function checks if the passed value is of type interval.

type::is_interval(any) -> bool

The following example shows this function, and its output:

[type::is_interval(ivl'1 day'), type::is_interval(1d)];

[true, false]

type::is_line

The type::is_line function checks if the passed value is of type line.

type::is_line(any) -> bool

The following example shows this function, and its output:

type::is_line("I am not a line");

false

type::is_none

The type::is_none function checks if the passed value is of type none.

type::is_none(any) -> bool

The following example shows this function, and its output:

type::is_none(NONE);

true

type::is_null

The type::is_null function checks if the passed value is of type null.

type::is_null(any) -> bool

The following example shows this function, and its output:

type::is_null(NULL);

true

type::is_multiline

The type::is_multiline function checks if the passed value is of type multiline.

type::is_multiline(any) -> bool

The following example shows this function, and its output:

type::is_multiline("I am not a multiline");

false

type::is_multipoint

The type::is_multipoint function checks if the passed value is of type multipoint.

type::is_multipoint(any) -> bool

The following example shows this function, and its output:

type::is_multipoint("I am not a multipoint");

false

type::is_multipolygon

The type::is_multipolygon function checks if the passed value is of type multipolygon.

type::is_multipolygon(any) -> bool

The following example shows this function, and its output:

type::is_multipolygon("I am not a multipolygon");

false

type::is_number

The type::is_number function checks if the passed value is of type number.

type::is_number(any) -> bool

The following example shows this function, and its output:

type::is_number(123);

true

type::is_object

The type::is_object function checks if the passed value is of type object.

type::is_object(any) -> bool

The following example shows this function, and its output:

type::is_object({ hello: 'world' });

true

type::is_point

The type::is_point function checks if the passed value is of type point.

type::is_point(any) -> bool

The following example shows this function, and its output:

type::is_point((-0.118092, 51.509865));

true

type::is_polygon

The type::is_polygon function checks if the passed value is of type polygon.

type::is_polygon(any) -> bool

The following example shows this function, and its output:

type::is_polygon("I am not a polygon");

false

type::is_range

The type::is_range function checks if the passed value is of type range.

type::is_range(any) -> bool

type::is_range(0..1);
//- true

-- method syntax
(0..1).is_range();
//- true

type::is_record

The type::is_record function checks if the passed value is of type record.

type::is_record(any) -> bool

The following example shows this function, and its output:

type::is_record(user:tobie);

true

Validate a table

type::is_record(user:tobie, 'test');

-- false

type::is_string

The type::is_string function checks if the passed value is of type string.

type::is_string(any) -> bool

The following example shows this function, and its output:

type::is_string("abc");

true

type::is_time

Available since: v4.0.0

The type::is_time function checks if the passed value is of type time.

type::is_time(any) -> bool

The following example shows this function, and its output:

[type::is_time(pt'12:30:00'), type::is_time(12h)];

[true, false]

type::is_timestamp

Available since: v4.0.0

The type::is_timestamp function checks if the passed value is of type timestamp.

type::is_timestamp(any) -> bool

The following example shows this function, and its output:

[type::is_timestamp(ts'2026-10-27T12:30:00'), type::is_timestamp(pd'2026-10-27')];

[true, false]

type::is_u8

Available since: v4.0.0

The type::is_u8 function checks if the passed value is of type u8. A number of another type with the same value, such as 5 or 5u16, is not a u8.

type::is_u8(any) -> bool

The following example shows this function, and its output:

type::is_u8(5);

false

type::is_u16

Available since: v4.0.0

The type::is_u16 function checks if the passed value is of type u16.

type::is_u16(any) -> bool

The following example shows this function, and its output:

type::is_u16(5u16);

true

type::is_u32

Available since: v4.0.0

The type::is_u32 function checks if the passed value is of type u32.

type::is_u32(any) -> bool

The following example shows this function, and its output:

type::is_u32(5u32);

true

type::is_u64

Available since: v4.0.0

The type::is_u64 function checks if the passed value is of type u64.

type::is_u64(any) -> bool

The following example shows this function, and its output:

type::is_u64(5u64);

true

type::is_uuid

The type::is_uuid function checks if the passed value is of type uuid.

type::is_uuid(any) -> bool

The following example shows this function, and its output:

type::is_uuid(u"018a6680-bef9-701b-9025-e1754f296a0f");

true

Method chaining

Method chaining allows functions to be called using the . dot operator on a value of a certain type instead of the full path of the function followed by the value.

-- Traditional syntax
type::is_record(r"person:aeon", "cat");

-- Method chaining syntax
r"person:aeon".is_record("cat");

false