Types

Scalars

Scalar types are the JSON scalar types with distinction between int and float.

Type	Example	Notes
null	`null`	Distinct from a missing object key
boolean	`true`, `false`
integer	`1`, `-5`	Exact 64-bit signed integers
float	`1.5`, `3.14`	Non-finite values error on insert
string	`'hello'`	UTF-8 text

Complex

Complex types are the JSON array type and object type.

Type	Example	Notes
array	`[1, 2, 3]`	Ordered sequence
object	`{ x: 1 }`	Insertion-ordered map of named fields

Cast

Scalar type names are callable as conversion functions: int(x), float(x), string(x), bool(x), and number(x). A cast is a normal function call — it nests and combines with operators. null propagates (int(null) is null); conversions that cannot succeed are runtime errors; object, array, and any are not callable and produce syntax errors.

Function	Description
`int(x)`	To integer: floats truncate toward zero; strings parse leniently (trimmed whitespace, float syntax accepted for int targets); bools → `1`/`0`
`float(x)`	To float: ints widen, strings parse, bools → `1.0`/`0.0`
`string(x)`	To string: the value's text form (`42`, `1.5`, `true`)
`bool(x)`	To bool: `0`/`0.0` → `false`, nonzero → `true`; the strings `true`/`false` (case-insensitive)
`number(x)`	To number: keeps int/float-ness; strings parse to the narrowest numeric variant

int(2.7)           -- 2
int('2.7')         -- 2
string(42)         -- "42"
float(3)           -- 3.0
int({a: 2.9}.a)    -- 2
typeof(int(2.7))   -- "int"

See the cast examples for full conversion matrices and error cases.