The JoinRowSet interface provides a mechanism for combining related data from different RowSet objects into one JoinRowSet object, which represents an SQL JOIN. In other words, a JoinRowSet object acts as a container for the data from RowSet objects that form an SQL JOIN relationship.

The Joinable interface provides the methods for setting, retrieving, and unsetting a match column, the basis for establishing an SQL JOIN relationship. The match column may alternatively be set by supplying it to the appropriate version of the JointRowSet method addRowSet.

1.0 Overview

Disconnected RowSet objects (CachedRowSet objects and implementations extending the CachedRowSet interface) do not have a standard way to establish an SQL JOIN between RowSet objects without the expensive operation of reconnecting to the data source. The JoinRowSet interface is specifically designed to address this need.

Any RowSet object can be added to a JoinRowSet object to become part of an SQL JOIN relationship. This means that both connected and disconnected RowSet objects can be part of a JOIN. RowSet objects operating in a connected environment (JdbcRowSet objects) are encouraged to use the database to which they are already connected to establish SQL JOIN relationships between tables directly. However, it is possible for a JdbcRowSet object to be added to a JoinRowSet object if necessary.

Any number of RowSet objects can be added to an instance of JoinRowSet provided that they can be related in an SQL JOIN. By definition, the SQL JOIN statement is used to combine the data contained in two or more relational database tables based upon a common attribute. The Joinable interface provides the methods for establishing a common attribute, which is done by setting a match column. The match column commonly coincides with the primary key, but there is no requirement that the match column be the same as the primary key. By establishing and then enforcing column matches, a JoinRowSet object establishes JOIN relationships between RowSet objects without the assistance of an available relational database.

The type of JOIN to be established is determined by setting one of the JoinRowSet constants using the method setJoinType. The following SQL JOIN types can be set:

• CROSS_JOIN
• FULL_JOIN
• INNER_JOIN - the default if no JOIN type has been set
• LEFT_OUTER_JOIN
• RIGHT_OUTER_JOIN

Note that if no type is set, the JOIN will automatically be an inner join. The comments for the fields in the JoinRowSet interface explain these JOIN types, which are standard SQL JOIN types.

2.0 Using a JoinRowSet Object for Creating a JOIN

When a JoinRowSet object is created, it is empty. The first RowSet object to be added becomes the basis for the JOIN relationship. Applications must determine which column in each of the RowSet objects to be added to the JoinRowSet object should be the match column. All of the RowSet objects must contain a match column, and the values in each match column must be ones that can be compared to values in the other match columns. The columns do not have to have the same name, though they often do, and they do not have to store the exact same data type as long as the data types can be compared.

A match column can be be set in two ways:

• By calling the Joinable method setMatchColumn
  This is the only method that can set the match column before a RowSet object is added to a JoinRowSet object. The RowSet object must have implemented the Joinable interface in order to use the method setMatchColumn. Once the match column value has been set, this method can be used to reset the match column at any time.
• By calling one of the versions of the JoinRowSet method addRowSet that takes a column name or number (or an array of column names or numbers)
  Four of the five addRowSet methods take a match column as a parameter. These four methods set or reset the match column at the time a RowSet object is being added to a JoinRowSet object.

3.0 Sample Usage

The following code fragment adds two CachedRowSet objects to a JoinRowSet object. Note that in this example, no SQL JOIN type is set, so the default JOIN type, which is INNER_JOIN, is established.

In the following code fragment, the table EMPLOYEES, whose match column is set to the first column (EMP_ID), is added to the JoinRowSet object jrs. Then the table ESSP_BONUS_PLAN, whose match column is likewise the EMP_ID column, is added. When this second table is added to jrs, only the rows in ESSP_BONUS_PLAN whose EMP_ID value matches an EMP_ID value in the EMPLOYEES table are added. In this case, everyone in the bonus plan is an employee, so all of the rows in the table ESSP_BONUS_PLAN are added to the JoinRowSet object. In this example, both CachedRowSet objects being added have implemented the Joinable interface and can therefore call the Joinable method setMatchColumn.

     JoinRowSet jrs = new JoinRowSetImpl();
 
     ResultSet rs1 = stmt.executeQuery("SELECT * FROM EMPLOYEES");
     CachedRowSet empl = new CachedRowSetImpl();
     empl.populate(rs1);
     empl.setMatchColumn(1); 
     jrs.addRowSet(empl);
 
     ResultSet rs2 = stmt.executeQuery("SELECT * FROM ESSP_BONUS_PLAN");
     CachedRowSet bonus = new CachedRowSetImpl();
     bonus.populate(rs2);
     bonus.setMatchColumn(1); // EMP_ID is the first column
     jrs.addRowSet(bonus);
 

At this point, jrs is an inside JOIN of the two RowSet objects based on their EMP_ID columns. The application can now browse the combined data as if it were browsing one single RowSet object. Because jrs is itself a RowSet object, an application can navigate or modify it using RowSet methods.

     jrs.first();
     int employeeID = jrs.getInt(1);
     String employeeName = jrs.getString(2);
 

Note that because the SQL JOIN must be enforced when an application adds a second or subsequent RowSet object, there may be an initial degradation in performance while the JOIN is being performed.

The following code fragment adds an additional CachedRowSet object. In this case, the match column (EMP_ID) is set when the CachedRowSet object is added to the JoinRowSet object.

     ResultSet rs3 = stmt.executeQuery("SELECT * FROM 401K_CONTRIB");
     CachedRowSet fourO1k = new CachedRowSetImpl();
     four01k.populate(rs3);
     jrs.addRowSet(four01k, 1);
 

The JoinRowSet object jrs now contains values from all three tables. The data in each row in four01k in which the value for the EMP_ID column matches a value for the EMP_ID column in jrs has been added to jrs.

4.0 JoinRowSet Methods

The JoinRowSet interface supplies several methods for adding RowSet objects and for getting information about the JoinRowSet object.

• Methods for adding one or more RowSet objects
  These methods allow an application to add one RowSet object at a time or to add multiple RowSet objects at one time. In either case, the methods may specify the match column for each RowSet object being added.
• Methods for getting information
  One method retrieves the RowSet objects in the JoinRowSet object, and another method retrieves the RowSet names. A third method retrieves either the SQL WHERE clause used behind the scenes to form the JOIN or a text description of what the WHERE clause does.
• Methods related to the type of JOIN
  One method sets the JOIN type, and five methods find out whether the JoinRowSet object supports a given type.
• A method to make a separate copy of the JoinRowSet object
  This method creates a copy that can be persisted to the data source.