Builtin types

CHAR[ACTER] '(' length ')'	character string with length upperbound
VARCHAR '(' length ')' \| CHARACTER VARYING '(' length ')'	string with atmost length upperbound
CLOB \| CHARACTER LARGE OBJECT
BLOB \| BINARY LARGE OBJECT
DECIMAL '(' P ',' S ')' \| NUMERIC '(' P ',' S ')'	with precision P and scale S
TINYINT	8 bit integer
SMALLINT	16 bit integer
INT	32 bit integer
BIGINT	64 bit integer
REAL	32 bit floating point
DOUBLE [PRECISION]	64 bit floating point
BOOLEAN
DATE
TIME [ '(' posint ')' ] [WITH TIME ZONE]	time of day with precision and time zone
TIMESTAMP [ '(' posint ')' ] [WITH TIME ZONE]	date concatenated with unique time, precision and time zone
INTERVAL interval_qualifier	a temporal interval

The integer types align with the storage of 2,4, and 8 bytes. Their domain contains a NULL representation, which may cause unexpected side effects if you recast them to a larger size. The system does not provide overflow detection.

The numeric types are represented as fixed length integers, whose decimal point is produced during result rendering.

The types float and double map to the underlying implementation system. No special attention is given to the value NaN.

interval_qualifier :
{ YEAR | MONTH | DAY | HOUR | MINUTE | SECOND } [time_precision]
| start_field TO end_field

start_field :
{ YEAR | MONTH | DAY | HOUR | MINUTE } [time_precision]

end_field:
{ YEAR | MONTH | DAY | HOUR | MINUTE | SECOND [time_precision] }

time_precision :
'(' posint ')'