SQL JDBC JAVA：如果不存在（SELECT…）INSERT INTO（…）后的getGeneratedKeys（）

Question

I am now searching for a while not finding a solution for my problem. I would like to insert a new row into a table if it currently not exists. I need the auto-incremented key of that new entry or the existing key of the existing entry.

Also I'm not clear about the idea of the return-value of my INSERT, what if one column is inserted, value = 1, and what if the row exists? value = 0 or NULL or ?!

If possible I do NOT want to "BATCH-INSERT", I must use these two return-values for the next INSERT as foreign keys... or has anyone an idea how to do this all with batch inserts?

Thanks to everybody who can help!!

System: MS SQL 2012 standard database, JAVA SE 1.6, sqljdbc4

This is my code, that gives me a SQLException:

    public long executeUpdateActionStatement(String sqlStatement, int autoGeneratedKeys) {

    ResultSet rs = null;
    Statement stmt = null;
    int rowCount = 0;
    try {
        connect();
        stmt = connect.createStatement();

        rowCount = stmt.executeUpdate(sqlStatement, autoGeneratedKeys);


        // if no new row is inserted, later I do a select to get the key
        if (rowCount < 1) 
            return -1;

        stmt.getGeneratedKeys();    /* This step throws the exception ! */
        if (rs.next()) {
            long id = rs.getLong(1);
            System.out.println(" !! ==---> ID: " + id + " | : " + sqlStatement);
            return id;
        }


    } catch (SQLFeatureNotSupportedException feature) {
        feature.printStackTrace();
        System.out.println("|| ---->>> Fehler: SQLFeatureNotSupportedException: " + sqlStatement);

    } catch (SQLServerException sqlse) {
        sqlse.printStackTrace();
        System.out.println("|| ---->>> Fehler: SQLServerException: " + sqlStatement);
    } catch (SQLException e) {
        e.printStackTrace();
        System.out.println("|| ---->>> Fehler: SQLException: " + sqlStatement);
    }

    finally {

        if (rs != null) {
            try {
                rs.close();
            } catch (SQLException ex) {
                ex.printStackTrace();
            }
        }

        if (stmt != null) {
            try {
                stmt.close();
            } catch (SQLException ex) {
                ex.printStackTrace();
            }
        }
    }

    return -1;

}

The SQLException I get:

    com.microsoft.sqlserver.jdbc.SQLServerException: Die Anweisung muss ausgeführt werden, bevor Ergebnisse abgerufen werden können.
at com.microsoft.sqlserver.jdbc.SQLServerException.makeFromDriverError(SQLServerException.java:190)
at com.microsoft.sqlserver.jdbc.SQLServerStatement.getGeneratedKeys(SQLServerStatement.java:2040)

in english:

    The statement must be executed before any results can be obtained.

The SQL Statements:

    executeUpdateActionStatement("IF NOT EXISTS (SELECT GeoLocationId FROM GeoLocationCoordinates WHERE Longitude = " + longitude + 
    " AND Latitude = " + latitude + ") INSERT INTO GeoLocationCoordinates VALUES (" + longitude + ", " + latitude + ")", Statement.RETURN_GENERATED_KEYS); //Statement.RETURN_GENERATED_KEYS);

AND

    executeUpdateActionStatement("IF NOT EXISTS (SELECT PlaceId FROM Places WHERE TwitterPlaceId = \'" + p.getId() + "\') INSERT INTO Places VALUES (" + 
            placeData + ")", Statement.RETURN_GENERATED_KEYS);

The database-tables:

    /****** Object:  Table [dbo].[GeoLocationCoordinates]    Script Date: 16.08.2013 17:55:04 **************************************************************************/
    SET ANSI_NULLS ON
    GO

    SET QUOTED_IDENTIFIER ON
    GO

    CREATE TABLE [dbo].[GeoLocationCoordinates](
        [GeoLocationId] [bigint] IDENTITY(1,1) NOT NULL,
        [Longitude] [float] NOT NULL,
        [Latitude] [float] NOT NULL, CONSTRAINT [PK_GeoLocationCoordinates] PRIMARY KEY CLUSTERED (  [GeoLocationId] ASC
    )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
    ) ON [PRIMARY]

    GO

and

    /****** Object:  Table [dbo].[Places]    Script Date: 16.08.2013 17:57:16 **************************************************************************/
    SET ANSI_NULLS ON
    GO

    SET QUOTED_IDENTIFIER ON
    GO

    CREATE TABLE [dbo].[Places](
      [PlaceId] [bigint] IDENTITY(1,1) NOT NULL,
      [TwitterPlaceId] [nvarchar](255) NULL,
      [PlaceName] [nvarchar](255) NULL,
      [PlaceFullName] [nvarchar](255) NULL,
      [StreetAdress] [nvarchar](255) NULL,
      [Country] [nvarchar](255) NULL,
      [PlaceType] [nvarchar](255) NULL,
      [PlaceUrl] [nvarchar](255) NULL,
      [BoundingBoxType] [nvarchar](255) NULL,
      [CountryTwoLetterCode] [nchar](2) NULL,
     CONSTRAINT [PK_Places] PRIMARY KEY CLUSTERED 
    (
  [PlaceId] ASC
    )WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
    ) ON [PRIMARY]

    GO

Answer 1

This is not syntactically correct, but gives an idea, just use scope identity in a transaction to get the new key and use it on child tables..The scrope in the transaction gives you hback the number you want. You'd just need to store the number of the found value in other code, if it exists. Could probably do it in the same proc, ideally.

declare @returnvalue int


BEGIN TRANSACTION
INSERT INTO GeoLocationCoordinates(longitude,latitude)
values (@lat,@long)

SELECT @returnvalue = SCOPE_IDENTITY()

INSERT INTO childtable (field1,field2,parentkey)
VALUES (@field1,@field2,@returnvalue)
COMMIT TRANSACTION

Answer 2

executeUpdate returns either 0, if no rows were affected, or the number of rows that were affected.

What step in your code is throwing the exception?

EDIT 2: Okay, looks like the merge will work.

PreparedStatement stmt = connection.prepareStatement(
        "merge table_1 t1 " +
        "using (select 'testvalue' as col1) t2 " +
        " on t1.col1 = t2.col1 " +
        " when not matched by target then " + 
        " insert (col1) values('testvalue'); "  
        ,Statement.RETURN_GENERATED_KEYS);

int out = stmt.executeUpdate();
System.out.println("Return:  " + out);  
ResultSet rs = stmt.getGeneratedKeys();
if (rs.next()) {
    System.out.println("value:  " + rs.getLong(1));
}

If the value 'testvalue' doesn't exist in the destination table, it will be inserted. If it's inserted, the return value will be 1, and getGeneratedKeys will return the new row's identity column value. If it does exist, both methods will return 0.