4.5.1 Parquet Data Type Conversions

See: Parquet Types.

Table 4.1 Parquet Data Type Conversions

MySQL Data Type	Parquet Logical Type	Comments
BOOL	STRING	Valid values: TRUE, FALSE, T, F, true, false, 0, 1
BOOL	INT	All signed and unsigned integers with any bit width, but the value must be 0 or 1
BOOL	None	Physical Type FLOAT or DOUBLE
CHAR, VARCHAR, TEXT	DECIMAL	The maximum precision is 38 and Lakehouse supports the Physical Type FIXED_LEN_BYTE_ARRAY
CHAR, VARCHAR, TEXT	STRING
CHAR, VARCHAR, TEXT	INT	All signed and unsigned integers
CHAR, VARCHAR, TEXT	None	Physical Type FLOAT or DOUBLE
DATE	STRING
DATE	DATE
DATE	TIMESTAMP	isAdjustedToUTC = false, timeUnit = x: Lakehouse only loads the date portion
DATE	None	Physical Type INT96: Lakehouse only loads the date portion
DATETIME	STRING	DATETIME accepts the range from 0000-00-00 to 9999-12-12
DATETIME	DATE	DATETIME accepts the range from 0000-00-00 to 9999-12-12
DATETIME	TIME	isAdjustedToUTC = false, timeUnit = x: DATETIME accepts the range from 0000-00-00 to 9999-12-12
DATETIME	TIMESTAMP	isAdjustedToUTC = false, timeUnit = x: DATETIME accepts the range from 0000-00-00 to 9999-12-12
DATETIME	None	Physical Type INT96: DATETIME accepts the range from 0000-00-00 to 9999-12-12
DECIMAL	DECIMAL	The maximum precision is 38 and Lakehouse supports the Physical Type FIXED_LEN_BYTE_ARRAY
DECIMAL	STRING
DECIMAL	INT	All signed and unsigned integers
DECIMAL	None	Physical Type FLOAT or DOUBLE
DOUBLE, FLOAT	DECIMAL	The maximum precision is 38 and Lakehouse supports the Physical Type FIXED_LEN_BYTE_ARRAY
DOUBLE, FLOAT	STRING
DOUBLE, FLOAT	INT	All signed and unsigned integers
DOUBLE, FLOAT	None	Physical Type FLOAT or DOUBLE
ENUM	STRING
ENUM	ENUM	Physical Type BYTE_ARRAY or FIXED_LEN_BYTE_ARRAY
ENUM	INT	All signed and unsigned integers
ENUM	None	Physical Type FLOAT or DOUBLE
JSON	STRING
JSON	JSON
JSON	None	Physical Type BYTE_ARRAY, Binary encoded. The non-UTF-8 encoding should be JSON binary, see: The JSON Data Type.
TIME	STRING
TIME	TIME	isAdjustedToUTC = false, timeUnit = x
TIMESTAMP	STRING	TIMESTAMP accepts the range from 1970-01-01 to 2038-01-19. Lakehouse might not load out of range values correctly. Use DATETIME instead.
TIMESTAMP	DATE	TIMESTAMP accepts the range from 1970-01-01 to 2038-01-19. Lakehouse might not load out of range values correctly. Use DATETIME instead.
TIMESTAMP	TIME	isAdjustedToUTC = false, timeUnit = x: TIMESTAMP accepts the range from 1970-01-01 to 2038-01-19. Lakehouse might not load out of range values correctly. Use DATETIME instead.
TIMESTAMP	TIMESTAMP	isAdjustedToUTC = false, timeUnit = x: TIMESTAMP accepts the range from 1970-01-01 to 2038-01-19. Lakehouse might not load out of range values correctly. Use DATETIME instead.
TIMESTAMP	None	Physical Type INT96: TIMESTAMP accepts the range from 1970-01-01 to 2038-01-19. Lakehouse might not load out of range values correctly. Use DATETIME instead.
All signed and unsigned TINYINT, SMALLINT, MEDIUMINT, INTEGER, INT, BIGINT	DECIMAL	The maximum precision is 38 and Lakehouse supports the Physical Type FIXED_LEN_BYTE_ARRAY
All signed and unsigned TINYINT, SMALLINT, MEDIUMINT, INTEGER, INT, BIGINT	STRING
All signed and unsigned TINYINT, SMALLINT, MEDIUMINT, INTEGER, INT, BIGINT	INT	All signed and unsigned integers
All signed and unsigned TINYINT, SMALLINT, MEDIUMINT, INTEGER, INT, BIGINT	None	Physical Type FLOAT or DOUBLE
YEAR	STRING	The value should be in the range 0 to 99 or the range 1901 to 2155
YEAR	INT	The value should be in the range 0 to 99 or the range 1901 to 2155
YEAR	None	Physical Type FLOAT or DOUBLE: The value should be in the range 0 to 99 or the range 1901 to 2155
YEAR	None	Physical Type INT96: The value should be in the range 1901 to 2155, Lakehouse only loads the year portion
YEAR	DATE	The value should be in the range 1901 to 2155, Lakehouse only loads the year portion
YEAR	TIMESTAMP	isAdjustedToUTC = false: The value should be in the range 1901 to 2155, Lakehouse only loads the year portion