Connector/J 1.1.3 Changelog

Download | Release Notes | Changelog |

Release date: 1 Jul 2013

For the highlights of this release, see the release notes.

The revision number links will take you to the revision's page on Launchpad. On Launchpad you can view more details of the revision and view diffs of the code modified in that revision.

• Revision #454 Fri 2013-06-28 13:48:36 +0200

  • bump version

• Wed 2013-06-26 20:31:08 +0200

  • Remove useless long test

• Wed 2013-06-26 19:28:12 +0200

  • Remove @Override annotation for JDK7 methods (make Eclipse happy)

• Wed 2013-06-26 18:19:42 +0200

  • Fix tests (SSL detection)

• Wed 2013-06-26 18:00:55 +0200

  • Make compilable under by JDK7 - add missing methods for MySQLServerSidePreparedStatement

• Tue 2013-06-25 23:46:55 +0200

  • : Introduce new parameters serverSslCert, which allows validation for self-signed server certificates.

  • Server certificate (or server CA certificate) can be passed in 3 forms

  • sslServerCert=/path/to/cert.pem (path to certificate) sslServerCert=classpath:relative/path (path relative to the current classpath) or as string, sslServerCert=----BEGIN CERTIFICATE-----

• Sat 2013-06-22 21:53:04 +0200

  • : Fix executeBatch() methods to throw BatchUpdateException(), as per documentation.

  • Also, cleanup and simplify batch handling, remove complexity introduced with handler/ factory abstractions.

• Mon 2013-06-17 10:29:25 +0200

  • : JDBC types in metadata for FLOAT and INT UNSIGNED are different than in ConnectorJ

  • Fix ResultSetMetaData.getColumnType() to be more ConnectorJ compatible (FLOAT=>java.sql.Type.REAL, rather than java.sql.Type.FLOAT etc)

• Sun 2013-06-16 22:53:24 +0200

  • fix build

• Sat 2013-06-15 00:06:09 +0200

  • : Fix bug in prepared statement parsing code (skipping next character after backslash '')

• Fri 2013-06-14 15:22:27 +0200

  • remove dummy test

• Fri 2013-06-14 15:00:20 +0200

  • fix ResultSet.getDisplaySize() for character columns - due to UTF-8 usage it was 3x more than expected

• Fri 2013-06-14 14:58:19 +0200

  • : provide alternative implementation for DatabaseMetadata.getImportedKeys() that does not use information schema, but parses the output of "show create table" instead.

  • The reason for it is to speed up DatabaseMetadata.getImportedKeys() call , which can get very slow in certain scenarios involving hibernate

• Tue 2013-06-11 16:25:04 +0200

  • update version of surefire plugin in pom.xml, to be able to run single test

• Tue 2013-06-11 16:22:59 +0200

  • : DatabaseMetaData.getPrimaryKeys returns incorrect value for KEY_SEQ (ORDINAL_POSITION from INFORMATION_SCHEMA.COLUMNS is incorrectly returned)

  • The fix to join I_S.COLUMNS

• Fri 2013-06-07 18:57:11 +0200

  • : allow java.util.Date to be passed to PreparedStatement.setObject()

• Fri 2013-06-07 17:18:58 +0200

  • : An exception is thrown when using PreparedStatement.getResultSetMetaData(), prior to statement execution. Statement must contain at least one '?' placeholder to trigger the exception

  • The reason for the bug was incorrect reading of COM_STMT_PREPARE response in this case (wrong order of parameter definition block and column definition block)

• Tue 2013-05-28 10:35:24 +0200

  • : multi-row inserts should return multiple entries in getGeneratedKeys()

• Tue 2013-05-28 09:53:41 +0200

  • - Escape processing - allow for missing space it temporal escapes, for better compatibility with ConnectorJ

• Thu 2013-05-23 22:18:40 +0200

  • : for multi-row inserts, getGeneratedKeys() will return ResultSet with multiple entries. like ConnectorJ does.

• Wed 2013-05-22 00:04:43 +0200

  • - Fix threading issue in MySQLCallableStatement.

  • 3 operations - setting input parameters, caling stored procedure and reading off the output parameters must be "atomic". i.e it needs to be synchronized so that it does not interfere with other clients using the same Connection in a different thread.

• Tue 2013-05-21 13:55:57 +0200

  • CONJ37- don't connect socket if it is already connected - (i.e in case custom SocketFactory creates an already connected sockets)

• Thu 2013-05-09 13:45:22 +0200

  • fix named pipe remote access

• Tue 2013-05-07 23:02:02 +0200

  • : implement connections via named pipe on Windows (via pipe= syntax in the connect URL)

• Mon 2013-05-06 20:42:09 +0200

  • : Change assertion in PacketOutputStream.startPacket() to be IOException (so it can be caught). "Last packet not finished" condition can happen , if underlying stream is broken (e.g write operation failed)

• Fri 2013-05-03 00:05:34 +0200

  • Add build script to bazaar repository

• Fri 2013-05-03 00:04:52 +0200

  • MySQLProtocol has this like in executeQuery :

  • log.finest("Executing streamed query: " + dQuery);

  • Even if log is not activated (which is normally the case), even then relatively expensive string concatenation takes place.It should be avoided by using