JDK 5 Documentation v1.3.1, Java 2 SDK 英文文档

JDBC^TM Guide: Getting Started

Appendix C: Change History

Removed references to the java.sql2 package. Substituted a proposal that splits the JDBC API into two packages: javax.sql and java.sql.
Removed descriptions of the classes and intefaces in javax.sql. We plan to add them in again later, or create a separate document for the extended JDBC API.
Removed the ScrollableResultSet interface and associated methods on other interfaces. Added methods for scrolling to the ResultSet interface
Added result set type, concurrency type, keyset size, fetch size, and fetch direction properties to Connection, Statement, and ResultSet interfaces.
Deprecated Statement.setCursorName and ResultSet.getCursorName. Use of "SELECT FOR UPDATE" is also deprecated. The new result set update methods can be used instead.
Added SQL99 APIs.
Added new metadata for persistent Java objects.

The Struct interface no longer extends the SQLData interface.
All occurrences of the SQLType interface have been removed and replaced with 'String'.
The Array.getArray() method now returns Object instead of Object[]. This allows an int[] array to be returned, for example.
The specification now states that a Java programming language array may be passed as an input parameter via PreparedStatement.setObject().
The semantics of ResultSet.isDeleted(), etc. is clarified.
The DatabaseMetaData.getClass() method has been dropped due to the difficulties involved in loading classes directly from a database.
The ResultSet.getBigDecimal() and CallableStatement.getBigDecimal() methods which take an `int scale' parameter have been deprecated.
Statement.getResultSetType() now throws SQLException
Added method Array.getBaseTypeName(). Array.getBaseType() now returns an int type code.
DatabaseMetaData.getUDTs() now allows a fully qualified SQL name which may contain wildcard characters in the typeNamePattern parameter.

Section 9.4: Compiled code examples and removed syntax errors.
Section 9.4.2: Added note on the independence of the Java programming language and SQL99 inheritance hierarchies.
Section 6.1: Removed the requirement that indirect updates, such as those performed by a trigger that is fired, be included in the update counts returned by a batch update.
Section 5.6: Clarified some technical details concerning inserting new rows in a result set.
Section 7.3.1: Noted that the TYPE_NAME returned by getTypeInfo() may contain a vendor specific type name when the type code is JAVA_ODBECT. Previously, the TYPE_NAME had to be null.
Removed method SQLInput.readStruct() and SQLOutput.writeStruct(). Read/ writeObject() should be used instead.
Chapter 5: Revised and simplified the scrollable result set model.
Section 11.3: Added note on character conversion.
Section 10.3: Added description of time zone support.

Section 5.5: Added rules for selecting a supported result set type and concurrency type.
Section 5.6: Added method ResultSet.cancelRowUpdates.
Section 5.6: Added more description on the semantics of the update methods.
Added Section 5.9.

Section 5.6: Added general description of the queries that yield updatable result sets.
Section 8.2: Clarified rules for passing blob and clob values as input parameters.
Sections 11.4 and 11.5: Added these sections to the document.

Chapter 6: Extended the semantics of executeBatch. Batch processing is now allowed to continue after an element in a batch has failed to execute successfully.
Added Section 10.6.