Bug #61134 bad row affected values with INSERT ON DUPLICATE KEY UPDATE

Description:
Hi,

We get different return values when using INSERT ON DUPLICATE KEY UPDATE with jdbc connector than using mysql commandline. In commandline we get a value of 2 when the statement INSERT ON DUPLICATE KEY update a row, while we get a value of 3 with the jdbc connector

How to repeat:
mysql> use test
Database changed
mysql> CREATE TABLE `LgstApplication` (
    ->   `a` varchar(25) NOT NULL COMMENT 'Format is: CountryCode+Number+Kind',
    ->   `b` varchar(50) DEFAULT NULL COMMENT 'Questel XAP',
    ->   `c` date DEFAULT NULL COMMENT 'Questel APD\nPRS L014EP\nXLEV L007EP',
    ->   `d` enum('INVENTION','DESIGN','UTILITY_MODEL','PLANT_PATENT','TRANSLATION') DEFAULT NULL COMMENT '(see Enumd class)',
    ->   `e` date DEFAULT NULL COMMENT 'First Date of Issued Grant',
    ->   `f` date DEFAULT NULL COMMENT 'if PAP is present takes PAP\nelse if FD contains the XAP [xxxxCC-xxxxxxx] take the Oldest FD dates \n',
    ->   `g` varchar(255) DEFAULT NULL COMMENT 'Linked to the f',
    ->   `h` varchar(50) DEFAULT NULL,
    ->   `EED` date DEFAULT NULL COMMENT 'Expected Expiration Date',
    ->   `j` date DEFAULT NULL COMMENT 'If a user override the EED value\n',
    ->   `k` tinyint(1) DEFAULT NULL COMMENT 'If the OverridenEEDValue is permanent or if the next event on this application will re-compute the EED and the OverridenEEDValue will be reset to null\n',
    ->   `l` tinyint(1) NOT NULL DEFAULT '1',
    ->   `m` varchar(25) DEFAULT NULL,
    ->   `n` varchar(25) DEFAULT NULL,
    ->   `o` smallint(5) unsigned DEFAULT NULL,
    ->   `p` tinyint(1) NOT NULL DEFAULT '1',
    ->   `q` tinyint(1) NOT NULL DEFAULT '0',
    ->   `r` tinyint(1) DEFAULT NULL COMMENT 'From Questel NO field.\nUS only, if a disclaimer is noted',
    ->   `s` tinyint(3) unsigned NOT NULL COMMENT '2 for WO, 1 for EP, 0 for others, this column is used only for sort reasons',
    ->   PRIMARY KEY (`a`),
    ->   KEY `idx1` (`l`)
    -> ) ENGINE=InnoDB DEFAULT CHARSET=latin1;
Query OK, 0 rows affected (0.04 sec)

mysql> INSERT INTO LgstApplication (a,b,c,d,e,f,g,h,EED,j,k,l,m,n,o,p,q,r,s) VALUES ('CA458556A','1984CA-0458556','1984-07-10','INVENTION',NULL,NULL,NULL,NULL,NULL,NULL,0,1,NULL,NULL,0,1,0,0,0) ON DUPLICATE KEY UPDATE b = VALUES(b), c = VALUES(c), d = VALUES(d), e = VALUES(e), f = VALUES(f), g = VALUES(g), h = VALUES(h), m = VALUES(m), n = VALUES(n), o = VALUES(o),p = VALUES(p), r = VALUES(r), s = VALUES(s);
Query OK, 1 row affected (0.00 sec)

mysql> INSERT INTO LgstApplication (a,b,c,d,e,f,g,h,EED,j,k,l,m,n,o,p,q,r,s) VALUES ('CA458556A','1984CA-0458556','1984-07-11','INVENTION','1989-03-07',NULL,NULL,NULL,NULL,NULL,0,1,NULL,NULL,0,1,0,0,0) ON DUPLICATE KEY UPDATE b = VALUES(b), c = VALUES(c), d = VALUES(d), e = VALUES(e), f = VALUES(f), g = VALUES(g), h = VALUES(h), m = VALUES(m), n = VALUES(n), o = VALUES(o),p = VALUES(p), r = VALUES(r), s = VALUES(s);
Query OK, 2 rows affected (0.00 sec)

Using jdbc connector 5.1.16 :

/*****sample code*****/
/*** you will need to adapt it with your configuration ***/

   public static void main(String arg[]) throws Exception{

      System.setProperty("JREP_PROPERTIES_FILE", "file.properties");

      Environment.loadProperties();

      String driver = System.getProperty(Environment.JREP_DB_DRIVER);

      try{

         Class.forName(driver);

         Connection connexion = DriverManager.getConnection(

               System.getProperty(Environment.JREP_DB_URL),

               System.getProperty(Environment.JREP_DB_USER),

               System.getProperty(Environment.JREP_DB_PWD));

         Statement instruction = connexion.createStatement();

         int i1 = instruction.executeUpdate("delete from LgstApplication");

         System.out.println("i1="+i1);

         int i2 = instruction.executeUpdate("INSERT INTO LgstApplication (a,b,c,d,e,f,g,h,EED,j,k,l,m,n,o,p,q,r,s) VALUES ('CA458556A','1984CA-0458556','1984-07-10','INVENTION',NULL,NULL,NULL,NULL,NULL,NULL,0,1,NULL,NULL,0,1,0,0,0) ON DUPLICATE KEY UPDATE b = VALUES(b), c = VALUES(c), d = VALUES(d), e = VALUES(e), f = VALUES(f), g = VALUES(g), h = VALUES(h), m = VALUES(m), n = VALUES(n), o = VALUES(o),p = VALUES(p), r = VALUES(r), s = VALUES(s)");

         System.out.println("i2="+i2);

         int i3 = instruction.executeUpdate("INSERT INTO LgstApplication (a,b,c,d,e,f,g,h,EED,j,k,l,m,n,o,p,q,r,s) VALUES ('CA458556A','1984CA-0458556','1984-07-11','INVENTION','1989-03-07',NULL,NULL,NULL,NULL,NULL,0,1,NULL,NULL,0,1,0,0,0) ON DUPLICATE KEY UPDATE b = VALUES(b), c = VALUES(c), d = VALUES(d), e = VALUES(e), f = VALUES(f), g = VALUES(g), h = VALUES(h), m = VALUES(m), n = VALUES(n), o = VALUES(o),p = VALUES(p), r = VALUES(r), s = VALUES(s)");

         System.out.println("i3="+i3);
      }

      catch (Exception e){

         System.out.println("Driver error : "+e);

      }
   }
/***end of sample code***/

Output : 

i1=1
i2=1
i3=3  <--- you see the differences ?

This is probably a duplicate of bug #39352. Had you tired to use useAffectedRows connection property?

it works better with it :)

thank you Valeriy