Description:
We recently upgraded from
old: /usr/sbin/mysqld --version
/usr/sbin/mysqld Ver 4.0.24_Debian-10-log for pc-linux-gnu on i386 (Source distribution)
to
new:~$ mysql --version
mysql Ver 14.7 Distrib 4.1.13, for pc-linux-gnu (i686) using readline 4.3
new:~$ /usr/sbin/mysqld --version -debug
(no clue why mysqld doesn't want to tell me what version it is when I do not have root permission)
Any, it appears that
int com.mysql.jdbc.ResultSet.getNativeInt(int columnIndex)
fails and an int of "2" is displayed as "65538"
When doing the same query as I do with jdbc with the windows MySQL Query Browser 1.1.14, the correct result, namely "2" is displayed.
I stepped through with the source with the below test case with createTable = false, the correct result of 2 is returned. When doing createTable = true, I see the following difference in the ResultSet.getNativeInt, f.getMysqlType() no longer is
case MysqlDefs.FIELD_TYPE_LONG:
but
case MysqlDefs.FIELD_TYPE_SHORT:
if (!f.isUnsigned()) {
return getNativeShort(columnIndex + 1);
}
return getNativeShort(columnIndex + 1) + 65536;
and obviously 65536 is added.
Originally, the field was created with
CREATE TABLE upgradedExistingTable (
lang_id SMALLINT UNSIGNED NOT NULL DEFAULT 1,
Using MyISAM Tables and on the client side java version "1.5.0_04"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_04-b05)
Java HotSpot(TM) Client VM (build 1.5.0_04-b05, mixed mode) on windows XP
server side: Linux 2.6.9-11.EL #1 Fri May 20 18:17:57 EDT 2005 i686 i686 i386 GNU/Linux
How to repeat:
ResultSet rs = null;
Statement stmt = null;
try {
PreparedStatement selUtfStringStmt = null;
boolean createTable = false;
if (createTable) {
stmt = this.conn.createStatement(
java.sql.ResultSet.TYPE_FORWARD_ONLY,
java.sql.ResultSet.CONCUR_UPDATABLE);
stmt.executeUpdate("DROP TABLE IF EXISTS bugInt");
stmt.executeUpdate("CREATE TABLE bugInt ("
+ "id INT NOT NULL AUTO_INCREMENT UNIQUE, "
+ "lang_id INT NOT NULL, PRIMARY KEY (id))");
stmt
.executeUpdate("INSERT INTO bugInt (lang_id) VALUES (2);");
selUtfStringStmt = this.conn
.prepareStatement("SELECT lang_id FROM bugInt WHERE id = '1' ");
} else {
selUtfStringStmt = this.conn
.prepareStatement("SELECT lang_id FROM upgradedExistingTable WHERE id= '1' ");
}
log.debug("selUtfStringStmt: "
+ ((DelegatingPreparedStatement) selUtfStringStmt)
.getDelegate().toString());
rs = selUtfStringStmt.executeQuery();
int i = 0;
if (rs == null) {
log.debug("row set is null!");
}
if (!rs.next()) {
log.debug("no more results" + i);
} else {
do {
i++;
log.debug("lang: " + rs.getInt("lang_id"));
} while (rs.next());
log.debug(i + " user found.");
}
} finally {
if (rs != null) {
try {
rs.close();
} catch (SQLException ex) {
// ignore
}
}
}
Suggested fix:
If there was a CVS, I'd happily compile and test with a changed driver
Description: We recently upgraded from old: /usr/sbin/mysqld --version /usr/sbin/mysqld Ver 4.0.24_Debian-10-log for pc-linux-gnu on i386 (Source distribution) to new:~$ mysql --version mysql Ver 14.7 Distrib 4.1.13, for pc-linux-gnu (i686) using readline 4.3 new:~$ /usr/sbin/mysqld --version -debug (no clue why mysqld doesn't want to tell me what version it is when I do not have root permission) Any, it appears that int com.mysql.jdbc.ResultSet.getNativeInt(int columnIndex) fails and an int of "2" is displayed as "65538" When doing the same query as I do with jdbc with the windows MySQL Query Browser 1.1.14, the correct result, namely "2" is displayed. I stepped through with the source with the below test case with createTable = false, the correct result of 2 is returned. When doing createTable = true, I see the following difference in the ResultSet.getNativeInt, f.getMysqlType() no longer is case MysqlDefs.FIELD_TYPE_LONG: but case MysqlDefs.FIELD_TYPE_SHORT: if (!f.isUnsigned()) { return getNativeShort(columnIndex + 1); } return getNativeShort(columnIndex + 1) + 65536; and obviously 65536 is added. Originally, the field was created with CREATE TABLE upgradedExistingTable ( lang_id SMALLINT UNSIGNED NOT NULL DEFAULT 1, Using MyISAM Tables and on the client side java version "1.5.0_04" Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_04-b05) Java HotSpot(TM) Client VM (build 1.5.0_04-b05, mixed mode) on windows XP server side: Linux 2.6.9-11.EL #1 Fri May 20 18:17:57 EDT 2005 i686 i686 i386 GNU/Linux How to repeat: ResultSet rs = null; Statement stmt = null; try { PreparedStatement selUtfStringStmt = null; boolean createTable = false; if (createTable) { stmt = this.conn.createStatement( java.sql.ResultSet.TYPE_FORWARD_ONLY, java.sql.ResultSet.CONCUR_UPDATABLE); stmt.executeUpdate("DROP TABLE IF EXISTS bugInt"); stmt.executeUpdate("CREATE TABLE bugInt (" + "id INT NOT NULL AUTO_INCREMENT UNIQUE, " + "lang_id INT NOT NULL, PRIMARY KEY (id))"); stmt .executeUpdate("INSERT INTO bugInt (lang_id) VALUES (2);"); selUtfStringStmt = this.conn .prepareStatement("SELECT lang_id FROM bugInt WHERE id = '1' "); } else { selUtfStringStmt = this.conn .prepareStatement("SELECT lang_id FROM upgradedExistingTable WHERE id= '1' "); } log.debug("selUtfStringStmt: " + ((DelegatingPreparedStatement) selUtfStringStmt) .getDelegate().toString()); rs = selUtfStringStmt.executeQuery(); int i = 0; if (rs == null) { log.debug("row set is null!"); } if (!rs.next()) { log.debug("no more results" + i); } else { do { i++; log.debug("lang: " + rs.getInt("lang_id")); } while (rs.next()); log.debug(i + " user found."); } } finally { if (rs != null) { try { rs.close(); } catch (SQLException ex) { // ignore } } } Suggested fix: If there was a CVS, I'd happily compile and test with a changed driver