Data types

How pgJDBC maps PostgreSQL types to Java types: temporal types (java.time), binary data, arrays, geometric types, large objects, and the infinity sentinel.

• Binary data (BYTEA) Choosing between the BYTEA column type and the large-object API for binary payloads: 1 GB memory ceiling on BYTEA, separate ACLs and manual deletion on large objects, and which JDBC getBytes / getBlob methods target each.
• Arrays Working with PostgreSQL array columns from JDBC: Connection.createArrayOf and the PGConnection overload for primitive arrays, the Java-to-PG type mapping table, and which element types travel as binary in PreparedStatement.setObject.
• Date and time Mapping PostgreSQL temporal types (DATE, TIME, TIMESTAMP, with and without time zone) to Java 8 java.time types via JDBC 4.2, the UTC-normalised return of OffsetDateTime, and why ZonedDateTime and Instant are unsupported.
• Date and timestamp infinity How the driver represents PostgreSQL -infinity and infinity timestamps: LocalDate.MIN / MAX (and analogous sentinels) for java.time, and PGStatement.DATE_NEGATIVE_INFINITY / DATE_POSITIVE_INFINITY for legacy Timestamp.
• Geometric types Reading and writing PostgreSQL geometric types (point, line, box, path, polygon, circle) through the org.postgresql.geometric classes, with an example PGcircle / PGpoint round-trip.
• Large objects Reading and writing PostgreSQL large objects (LO / OID) through the LargeObjectManager and LargeObject extension classes, which give random access to the object’s contents beyond what standard Blob / Clob allow.