Count

This function can be used when counting field values and expressions.

count

The count function counts the number of times that the function is called. In a SELECT that aggregates with GROUP BY or GROUP ALL, that call count is the size of each group.

When every field in the projection is a bare zero-argument count() (optionally aliased), SurrealDB implies GROUP ALL. Before this change, a bare count() returned the constant 1 once per record.

count() -> 1

If a value is given as the first argument, then this function checks whether a given value is truthy. This is useful for returning the total number of rows which match a certain condition in a SELECT with a GROUP BY or GROUP ALL clause.

count(any) -> number

If an array is given, this function counts the number of items in the array which are truthy. If, instead, you want to count the total number of items in the given array, then use the array::len() function.

count(array) -> number

The following example shows this function, and its output, when used in a RETURN statement:

RETURN count();

-- 1

RETURN count(true);

-- 1

RETURN count(10 > 15);

-- 0

RETURN count([ 1, 2, 3, null, 0, false, (15 > 10), rand::uuid() ]);

5

The following examples show this function being used in a SELECT statement with a GROUP ALL clause. From 3.2.5, a projection made only of bare count() implies GROUP ALL; the examples keep the explicit form.

SELECT 
	count() 
FROM [
	{ age: 33 }, 
	{ age: 45 }, 
	{ age: 39 }
] 
GROUP ALL;

[
	{ count: 3 }
]

SELECT 
	count(age > 35) 
FROM [
	{ age: 33 }, 
	{ age: 45 }, 
	{ age: 39 }
] 
GROUP ALL;

[
	{ count: 2 }
]

An advanced example of the count function can be seen below:

SELECT
	country,
	count(age > 30) AS total
FROM [
	{ age: 33, country: 'GBR' },
	{ age: 45, country: 'GBR' },
	{ age: 39, country: 'USA' },
	{ age: 29, country: 'GBR' },
	{ age: 43, country: 'USA' }
]
GROUP BY country;

[
	{
		country: 'GBR',
		total: 2
	},
	{
		country: 'USA',
		total: 2
	}
]

Using a COUNT index with count()

A COUNT index can be defined to speed up count() when used with a GROUP ALL clause. This allows count() to access a single stored value when it is called instead of iterating over the entire table. From 3.2.5, a bare count() projection implies GROUP ALL; prefer the explicit form in examples and production queries.

CREATE user;
-- One record in table, very fast
SELECT count() FROM user GROUP ALL;

-- 10,000 new records,
-- count() takes a bit longer than before
CREATE |user:10000| RETURN NONE;
SELECT count() FROM user GROUP ALL;

-- Add index, wait a moment for it to build
DEFINE INDEX user_count ON user COUNT;
-- count() very performant again
SELECT count() FROM user GROUP ALL;