The instances of our entity class Book are special Java objects representing "entities" (or business objects), which can be serialized, or, in other words, turned into database records, or rows of a database table. Consequently, they can be shown as a table like in Table 3.2.

The data storage technology used in our example app is MySQL, and the (My)SQL code used to create the schema for the database table books is the following:

CREATE TABLE IF NOT EXISTS books (
  isbn VARCHAR(10) NOT NULL PRIMARY KEY,
  title VARCHAR(128),
  year SMALLINT
);

While it is also possible to create the database schema manually (with the help of CREATE TABLE statements such as the one above), we'll show later in this tutorial how the database schema can be automatically generated by JPA. In both cases, the database setup, including a user account and the associated rights (create, update, etc), must be done manually before the JPA application can connect to it.

The Book.add method takes care of creating a new Book instance and saving it to a database with the help of an 'entity manager':

public static void add( EntityManager em, UserTransaction ut, 
    String isbn, String title, int year) throws Exception {
  ut.begin();
  Book book = new Book( isbn, title, year);
  em.persist( book);
  ut.commit();
}

For storing the new object, the persist method of the given 'entity manager' is invoked. It is responsible for creating the corresponding SQL INSERT statement and executing it.

The instances of an entity class, such as Book, are retrieved from the database with the help of a corresponding query expressed in the Java Persistence Query Language (JPQL). These queries are similar to SQL queries. They use class names Instead of table names, property names instead of column names, and object variables instead of row variables.

In the Book.retrieveAll method, first a query asking for all Book instances is created, and then this query is executed with query.getResultList() assigning its answer set to the list variable books:

public static List<Book> retrieveAll( EntityManager em) {
  Query query = em.createQuery( "SELECT b FROM Book b", Book.class);
  List<Book> books = query.getResultList();
  return books;
}

For updating an existing Book instance we first retrieve it from the database with em.find, and then set those attributes the value of which has changed:

public static void update( EntityManager em, 
    UserTransaction ut, String isbn, String title, 
    int year) throws Exception {
  ut.begin();
  Book book = em.find( Book.class, isbn);
  if (!title.equals( book.getTitle())) book.setTitle( title);
  if (year != book.getYear()) book.setYear( year);
  ut.commit();
}

Notice that, when invoking the find method for retrieving an entity, the first argument must be a reference to the entity class concerned (here: Book.class), so the JPA runtime environment can identify the database table from which to retrieve the entity's data. The second argument must be the value of the entity's primary key.

Notice that in the update case, we do not have to use persist for saving the changes. Saving is automatically managed by the JPA runtime environment when we complete the transaction with ut.commit().

A book entity can be deleted from the database as shown in the following example code:

public static void destroy( EntityManager em, 
    UserTransaction ut, String isbn) throws Exception {
  ut.begin();
  Book book = em.find( Book.class, isbn);
  em.remove( book);
  ut.commit();
}

To delete an entity from the database, we first need to retrieve it with the help of the find method as in the update case. Then, the remove method has to be invoked by the 'entity manager', and finally the transaction is completed with ut.commit().

For being able to test our code, we may create some test data and save it in our database. We can use the following procedure for this:

public static void createTestData( EntityManager em, 
    UserTransaction ut) throws Exception {
  Book book = null;
  Book.clearData( em, ut);  // first clear the books table
  ut.begin();
  book = new Book("006251587X","Weaving the Web", 2000);
  em.persist( book);
  book = new Book("0465026567","Gödel, Escher, Bach", 1999);
  em.persist( book);
  book = new Book("0465030793","I Am A Strange Loop", 2008);
  em.persist( book);
  ut.commit();
}

After clearing the database, we successively create 3 instances of the Book entity class and save them with the help of persist.

The following procedure clears our database by deleting all rows:

public static void clearData( EntityManager em, 
    UserTransaction ut) throws Exception {
  ut.begin();
  Query deleteStatement = em.createQuery( "DELETE FROM Book");
  deleteStatement.executeUpdate();
  ut.commit();
}

JPA does not provide a direct method to drop the entire population of a specific class from the database. However, this can be easily obtained by using a JPQL statement as shown in the above code. The JPQL code can be read as: delete all rows from the database table associated with the entity class Book.