Using JDBC

This section will be a brief introduction to JDBC, addressing the basics of JDBC, issues, caveats, and so forth. For more detailed information, visit the JDBC website (http://java.sun.com/products/jdbc/), which has many good resources and will always provide the most up to date information. Also, the API documentation included with your JDK has detailed information on specific classes, methods, and fields. Look for the java.sql package.

JDBC has classes to represent most of the basic pieces of a program’s interaction with SQL. The classes are: Connection, Statement, ResultSet, Blob, and Clob, and they all map directly to some concept in SQL. JDBC also has helper classes, such as ResultSetMetaData and DatabaseMetaData, that represent meta-information. These are useful for when you’d like to get information about the capabilities of the database. They are also useful for getting the types of results returned by a query, either for debugging, or because you don’t know about the data you are dealing with.

PostgreSQL’s JDBC interface also provides classes to map to PostgreSQL’s non-standard extensions to JDBC’s SQL support. These non-standard extensions include: Fastpath, geometric types, native large objects, and a class that aids serialization of Java objects into the database.

Example 12-2 used a Connection object, representing a physical connection to the database. You can use this Connection object to create Statement objects. Statement objects are JDBC’s way of getting SQL statements to the database.

There are three main types of Statement objects: the base class Statement, the PreparedStatement, and the CallableStatement.

To create a Statement object, use the createStatement method as shown in Example 12-3:

Example 12-3 creates a Statement object named s, from the Connection object c. You can now use this Statement object to execute queries and updates on the database.

There are two main methods in the Statement class that are important. The first is executeQuery. This method takes one argument, the SQL statement to be executed, and returns an object of type ResultSet, which is discussed later. This method is used for executing queries which will return a set of data back, for instance, a SELECT statement. The ResultSet object returned represents the data resulting from the query.

Example 12-4 retrieves some data from the booktown database:

Example 12-4 creates a Statement object, and then uses that Statement object’s executeQuery method to execute the query SELECT * FROM books. You get back a ResultSet, and use that ResultSet to print out some of the information you got back.

The ResultSet object is our primary interface for fetching information from the database. It has two main features. It can step through the set of rows returned, and it can return the value for a specific column in that row. It works in a similar fashion to a standard Java Enumeration: it starts before the first element, and you use the next method to step through the rest of the elements.

next returns true if the ResultSet was able to step to the next results; that is to say, there are results to be read. The while loop in Example 12-4 will print out the first column of each of the rows returned. If no rows were returned, next will return false initially, representing this fact, and therefore nothing will be printed.

ResultSet can return values of all sorts of different types; Example 12-4 treats the first column as if it were a String. Fortunately, all standard SQL data types can be represented as String, so regardless of the type of the first column, you will be able to fetch the value of the first column and print it out. There are many other methods available on ResultSet, including methods for fetching all of the various SQL data types and converting them to native Java types. Consult the API documentation on ResultSet for more information.

The other important method is executeUpdate. This method, again, takes one argument, which is the SQL statement to be executed. The difference between executeQuery and executeUpdate is that executeUpdate is for executing statements that change data in the database. For example, use executeUpdate to execute a CREATE an INSERT or an UPDATE statement. executeUpdate returns an int, and the value of that int corresponds to the number of records that were modified.

Example 12-5 uses the executeUpdate method to insert a new row into the books table.

As mentioned earlier, besides the basic Statement object, there are two additional types of statements available in JDBC: PreparedStatements and CallableStatements. These two types are described later in this section.

In addition to these statements, this section also describes the use of the ResultSetMetaData and DatabaseMetaData objects. You can use these last two objects to interrogate JDBC for information about a given set of query results, or for information about your database. The ability to get such information at run-time enables you to dynamically execute any SQL statement, even one that is unknown when you write your program.

A PreparedStatement, in contrast to a CallableStatement, is used for SQL statements that are executed multiple times with different values. For instance, you might want to insert several values into a table, one after another. The advantage of the PreparedStatement is that it is pre-compiled, reducing the overhead of parsing SQL statements on every execution. Example 12-6 is an example of how a PreparedStatement might be used.

You can see that Example 12-6 prepares a statement in a similar fashion as before, except it uses a question mark (?) character in place of each value that you want to supply. Use the appropriate PreparedStatement set method (e.g., setInt, setString) to set each value. The specific set method that you use for a column depends on the data type of the column.

The PreparedStatement approach is useful because it avoids manual conversion of Java types to SQL types. For instance, the you do not have to worry about quoting or escaping when going to a text type.

Notice that the first parameter passed to a set method indicates the specific placeholder parameter (the question marks) that you are setting. A value of 1 corresponds to the first question mark, a value of 2 corresponds to the second, and so on.

The other strength of the PreparedStatement is that you can use it over and over again with new parameter values, rather than having to create a new Statement object for each new set of parameters. This approach is obviously more efficient, as only one object is created.

Use the set methods each time to specify new parameter values.

You can interrogate JDBC for detailed information about a query’s result set using a ResultSetMetaData object. ResultSetMetaData is a class that is used to find information about the ResultSet returned from a executeQuery call. It contains information about the number of columns, the types of data they contain, the names of the columns, and so on.

Two of the most common methods in the ResultSetMetaData are getColumnName and getColumnTypeName. These retrieve the name of a column, and the name of its associated data type, respectively, each in the form of a String.

Example 12-7 is an example of using the ResultSetMetaData to get the name and data type of the first column in a ResultSet called rs. This code could logically follow the acquisition of the ResultSet named rs in Example 12-4.

Overall, the PreparedStatement mechanism is considerably more robust than the Statement class.

There are many other useful methods in the ResultSetMetaData class, all of which are well documented in the JDK API documentation.