hibernate-orm/reference/en/modules/session_api.xml

<chapter id="objectstate">
    <title>Working with objects</title>

    <para>
        Hibernate is a full object/relational mapping solution that not only shields
        the developer from the details of the underlying database management
        system, but also offers <emphasis>state management</emphasis> of objects. This is,
        contrary to the management of SQL <literal>statements</literal> in common JDBC/SQL
        persistence layers, a very natural object-oriented view of persistence in Java
        applications.
    </para>

    <para>
        In other words, Hibernate application developers should always think about the
        <emphasis>state</emphasis> of their objects, and not necessarily about the
        execution of SQL statements. This part is taken care of by Hibernate and is only
        relevant for the application developer when tuning the performance of the system.
    </para>

    <sect1 id="objectstate-overview">
        <title>Hibernate object states</title>

        <para>
            Hibernate defines and supports the following object states:
        </para>

        <itemizedlist>
            <listitem>
                <para>
                    <emphasis>Transient</emphasis> - an object is transient if it has just
                    been instantiated using the <literal>new</literal> operator, and it
                    is not associated with a Hibernate <literal>Session</literal>. It has no
                    persistent representation in the database and no identifier value has been
                    assigned. Transient instances will be destroyed by the garbage collector if
                    the application doesn't hold a reference anymore. Use the Hibernate
                    <literal>Session</literal> to make an object persistent (and let Hibernate
                    take care of the SQL statements that need to be executed for this transition).
                </para>
            </listitem>
            <listitem>
                <para>
                    <emphasis>Persistent</emphasis> - a persistent instance has a representation
                    in the database and an identifier value. It might just have been saved or loaded,
                    however, it is by definition in the scope of a <literal>Session</literal>.
                    Hibernate will detect any changes made to an object in persistent state and
                    synchronize the state with the database when the unit of work completes.
                    Developers don't execute manual <literal>UPDATE</literal> statements, or
                    <literal>DELETE</literal> statements when an object should be made transient.
                </para>
            </listitem>
            <listitem>
                <para>
                    <emphasis>Detached</emphasis> - a detached instance is an object that has been
                    persistent, but its <literal>Session</literal> has been closed. The reference
                    to the object is still valid, of course, and the detached instance might even
                    be modified in this state. A detached instance can be reattached to a new
                    <literal>Session</literal> at a later point in time, making it (and all the
                    modifications) persistent again. This feature enables a programming model for
                    long running units of work that require user think-time. We call them
                    <emphasis>application transactions</emphasis>, i.e. a unit of work from the
                    point of view of the user.
                </para>
            </listitem>
        </itemizedlist>

        <para>
            We'll now discuss the states and state transitions (and the Hibernate methods that
            trigger a transition) in more detail.
        </para>

    </sect1>

    <sect1 id="objectstate-makingpersistent" revision="1">
        <title>Making objects persistent</title>

        <para>
            Newly instantiated instances of a a persistent class are considered 
            <emphasis>transient</emphasis> by Hibernate. We can make a transient 
            instance <emphasis>persistent</emphasis> by associating it with a 
            session:
        </para>

        <programlisting><![CDATA[DomesticCat fritz = new DomesticCat();
fritz.setColor(Color.GINGER);
fritz.setSex('M');
fritz.setName("Fritz");
Long generatedId = (Long) sess.save(fritz);]]></programlisting>

        <para>
            If <literal>Cat</literal> has a generated identifier, the identifier is
            generated and assigned to the <literal>cat</literal> when <literal>save()</literal> 
            is called. If <literal>Cat</literal> has an <literal>assigned</literal>
            identifier, or a composite key, the identifier should be assigned to 
            the <literal>cat</literal> instance before calling <literal>save()</literal>.
            You may also use <literal>persist()</literal> instead of <literal>save()</literal>,
            with the semantics defined in the EJB3 early draft.
        </para>
        
        <para>
            Alternatively, you may assign the identifier using an overloaded version
            of <literal>save()</literal>.
        </para>

<programlisting><![CDATA[DomesticCat pk = new DomesticCat();
pk.setColor(Color.TABBY);
pk.setSex('F');
pk.setName("PK");
pk.setKittens( new HashSet() );
pk.addKitten(fritz);
sess.save( pk, new Long(1234) );]]></programlisting>
        
        <para>
            If the object you make persistent has associated objects (e.g. the
            <literal>kittens</literal> collection in the previous example),
            these objects may be made persistent in any order you like unless you
            have a <literal>NOT NULL</literal> constraint upon a foreign key column.
            There is never a risk of violating foreign key constraints. However, you 
            might violate a <literal>NOT NULL</literal> constraint if you
            <literal>save()</literal> the objects in the wrong order.
        </para>
        
        <para>
            Usually you don't bother with this detail, as you'll very likely use Hibernate's
            <emphasis>transitive persistence</emphasis> feature to save the associated
            objects automatically. Then, even <literal>NOT NULL</literal>
            constraint violations don't occur - Hibernate will take care of everything.
            Transitive persistence is discussed later in this chapter.
        </para>
        
    </sect1>

    <sect1 id="objectstate-loading">
        <title>Loading an object</title>

        <para>
            The <literal>load()</literal> methods of <literal>Session</literal> gives you
            a way to retrieve a persistent instance if you already know its identifier. 
            <literal>load()</literal> takes a class object and will load the state into 
            a newly instantiated instance of that class, in persistent state.
        </para>

        <programlisting><![CDATA[Cat fritz = (Cat) sess.load(Cat.class, generatedId);]]></programlisting>

<programlisting><![CDATA[// you need to wrap primitive identifiers
long id = 1234;
DomesticCat pk = (DomesticCat) sess.load( DomesticCat.class, new Long(id) );]]></programlisting>

        <para>
            Alternatively, you can load state into a given instance:
        </para>

<programlisting><![CDATA[Cat cat = new DomesticCat();
// load pk's state into cat
sess.load( cat, new Long(pkId) );
Set kittens = cat.getKittens();]]></programlisting>

        <para>
            Note that <literal>load()</literal> will throw an unrecoverable exception if 
            there is no matching database row. If the class is mapped with a proxy, 
            <literal>load()</literal> just returns an uninitialized proxy and does not 
            actually hit the database until you invoke a method of the proxy. This 
            behaviour is very useful if you wish to create an association to an object
            without actually loading it from the database. It also allows multiple
            instances to be loaded as a batch if <literal>batch-size</literal> is
            defined for the class mapping.
        </para>
        
        <para>
            If you are not certain that a matching row exists, you should use the 
            <literal>get()</literal> method, which hits the database immediately and 
            returns null if there is no matching row.
        </para>
        
        <programlisting><![CDATA[Cat cat = (Cat) sess.get(Cat.class, id);
if (cat==null) {
    cat = new Cat();
    sess.save(cat, id);
}
return cat;]]></programlisting>

        <para>
            You may even load an object using an SQL <literal>SELECT ... FOR UPDATE</literal>,
            using a <literal>LockMode</literal>. See the API documentation for more information.
        </para>

        <programlisting><![CDATA[Cat cat = (Cat) sess.get(Cat.class, id, LockMode.UPGRADE);]]></programlisting>
        
        <para>
            Note that any associated instances or contained collections are 
            <emphasis>not</emphasis> selected <literal>FOR UPDATE</literal>, unless you decide
            to specify <literal>lock</literal> or <literal>all</literal> as a
            cascade style for the association.
        </para>
        
        <para>
            It is possible to re-load an object and all its collections at any time, using the 
            <literal>refresh()</literal> method. This is useful when database triggers are used to
            initialize some of the properties of the object.
        </para>
        
        <programlisting><![CDATA[sess.save(cat);
sess.flush(); //force the SQL INSERT
sess.refresh(cat); //re-read the state (after the trigger executes)]]></programlisting>

        <para>
            An important question usually appears at this point: How much does Hibernate load
            from the database and how many SQL <literal>SELECT</literal>s will it use? This
            depends on the <emphasis>fetching strategy</emphasis> and is explained in
            <xref linkend="performance-fetching"/>.
        </para>

    </sect1>

    <sect1 id="objectstate-querying" revision="1">
        <title>Querying</title>

        <para>
            If you don't know the identifiers of the objects you are looking for, 
            you need a query. Hibernate supports an easy-to-use but powerful object 
            oriented query language (HQL). For programmatic query creation, Hibernate
            supports a sophisticated Criteria and Example query feature (QBC and QBE).
            You may also express your query in the native SQL of your database, with
            optional support from Hibernate for result set conversion into objects.
        </para>

        <sect2 id="objectstate-querying-executing">
            <title>Executing queries</title>

            <para>
                HQL and native SQL queries are represented with an instance of <literal>org.hibernate.Query</literal>.
                This interface offers methods for parameter binding, result set handling, and for the execution
                of the actual query. You always obtain a <literal>Query</literal> using the current
                <literal>Session</literal>:
            </para>

        <programlisting><![CDATA[List cats = session.createQuery(
    "from Cat as cat where cat.birthdate < ?")
    .setDate(0, date)
    .list();

List mothers = session.createQuery(
    "select mother from Cat as cat join cat.mother as mother where cat.name = ?")
    .setString(0, name)
    .list();

List kittens = session.createQuery(
    "from Cat as cat where cat.mother = ?")
    .setEntity(0, pk)
    .list();

Cat mother = (Cat) session.createQuery(
    "select cat.mother from Cat as cat where cat = ?")
    .setEntity(0, izi)
    .uniqueResult();]]></programlisting>

            <para>
                A query is usually executed by invoking <literal>list()</literal>, the
                result of the query will be loaded completely into a collection in memory.
                Entity instances retrieved by a query are in persistent state. The
                <literal>uniqueResult()</literal> method offers a shortcut if you
                know your query will only return a single object.
            </para>

            <sect3 id="objectstate-querying-executing-iterate">
                <title>Iterating results</title>

                <para>
                    Occasionally, you might be able to achieve better performance by
                    executing the query using the <literal>iterate()</literal> method.
                    This will only usually be the case if you expect that the actual
                    entity instances returned by the query will already be in the session
                    or second-level cache. If they are not already cached,
                    <literal>iterate()</literal> will be slower than <literal>list()</literal>
                    and might require many database hits for a simple query, usually
                    <emphasis>1</emphasis> for the initial select which only returns identifiers,
                    and <emphasis>n</emphasis> additional selects to initialize the actual instances.
                </para>

                <programlisting><![CDATA[// fetch ids
Iterator iter = sess.createQuery("from eg.Qux q order by q.likeliness").iterate();
while ( iter.hasNext() ) {
    Qux qux = (Qux) iter.next();  // fetch the object
    // something we couldnt express in the query
    if ( qux.calculateComplicatedAlgorithm() ) {
        // delete the current instance
        iter.remove();
        // dont need to process the rest
        break;
    }
}]]></programlisting>
            </sect3>
            
            <sect3 id="objectstate-querying-executing-tuples">
                <title>Queries that return tuples</title>

                <para>
                    Hibernate queries sometimes return tuples of objects, in which case each tuple
                    is returned as an array:
                </para>

                <programlisting><![CDATA[Iterator kittensAndMothers = sess.createQuery(
            "select kitten, mother from Cat kitten join kitten.mother mother")
            .list()
            .iterator();

while ( kittensAndMothers.hasNext() ) {
    Object[] tuple = (Object[]) kittensAndMothers.next();
    Cat kitten  = tuple[0];
    Cat mother  = tuple[1];
    ....
}]]></programlisting>

            </sect3>

            <sect3 id="objectstate-querying-executing-scalar" revision="1">
                <title>Scalar results</title>

                <para>
                    Queries may specify a property of a class in the <literal>select</literal> clause.
                    They may even call SQL aggregate functions. Properties or aggregates are considered
                    "scalar" results (and not entities in persistent state).
                </para>

                <programlisting><![CDATA[Iterator results = sess.createQuery(
        "select cat.color, min(cat.birthdate), count(cat) from Cat cat " +
        "group by cat.color")
        .list()
        .iterator();

while ( results.hasNext() ) {
    Object[] row = (Object[]) results.next();
    Color type = (Color) row[0];
    Date oldest = (Date) row[1];
    Integer count = (Integer) row[2];
    .....
}]]></programlisting>

            </sect3>

            <sect3 id="objectstate-querying-executing-parameters">
                <title>Bind parameters</title>

                <para>
                    Methods on <literal>Query</literal> are provided for binding values to
                    named parameters or JDBC-style <literal>?</literal> parameters. 
                    <emphasis>Contrary to JDBC, Hibernate numbers parameters from zero.</emphasis>
                    Named parameters are identifiers of the form <literal>:name</literal> in 
                    the query string. The advantages of named parameters are:
                </para>

                <itemizedlist spacing="compact">
                    <listitem>
                        <para>
                            named parameters are insensitive to the order they occur in the
                            query string
                        </para>
                    </listitem>
                    <listitem>
                        <para>
                            they may occur multiple times in the same query
                        </para>
                    </listitem>
                    <listitem>
                        <para>
                            they are self-documenting
                        </para>
                    </listitem>
                </itemizedlist>

                <programlisting><![CDATA[//named parameter (preferred)
Query q = sess.createQuery("from DomesticCat cat where cat.name = :name");
q.setString("name", "Fritz");
Iterator cats = q.iterate();]]></programlisting>

                <programlisting><![CDATA[//positional parameter
Query q = sess.createQuery("from DomesticCat cat where cat.name = ?");
q.setString(0, "Izi");
Iterator cats = q.iterate();]]></programlisting>

                <programlisting><![CDATA[//named parameter list
List names = new ArrayList();
names.add("Izi");
names.add("Fritz");
Query q = sess.createQuery("from DomesticCat cat where cat.name in (:namesList)");
q.setParameterList("namesList", names);
List cats = q.list();]]></programlisting>

            </sect3>

            <sect3 id="objectstate-querying-executing-pagination">
                <title>Pagination</title>

                <para>
                    If you need to specify bounds upon your result set (the maximum number of rows
                    you want to retrieve and / or the first row you want to retrieve) you should
                    use methods of the <literal>Query</literal> interface:
                </para>

                <programlisting><![CDATA[Query q = sess.createQuery("from DomesticCat cat");
q.setFirstResult(20);
q.setMaxResults(10);
List cats = q.list();]]></programlisting>

                <para>
                    Hibernate knows how to translate this limit query into the native
                    SQL of your DBMS.
                </para>

            </sect3>

            <sect3 id="objectstate-querying-executing-scrolling">
                <title>Scrollable iteration</title>

                <para>
                    If your JDBC driver supports scrollable <literal>ResultSet</literal>s, the
                    <literal>Query</literal> interface may be used to obtain a
                    <literal>ScrollableResults</literal> object, which allows flexible
                    navigation of the query results.
                </para>

                <programlisting><![CDATA[Query q = sess.createQuery("select cat.name, cat from DomesticCat cat " +
                            "order by cat.name");
ScrollableResults cats = q.scroll();
if ( cats.first() ) {

    // find the first name on each page of an alphabetical list of cats by name
    firstNamesOfPages = new ArrayList();
    do {
        String name = cats.getString(0);
        firstNamesOfPages.add(name);
    }
    while ( cats.scroll(PAGE_SIZE) );

    // Now get the first page of cats
    pageOfCats = new ArrayList();
    cats.beforeFirst();
    int i=0;
    while( ( PAGE_SIZE > i++ ) && cats.next() ) pageOfCats.add( cats.get(1) );

}
cats.close()]]></programlisting>

                <para>
                    Note that an open database connection (and cursor) is required for this
                    functionality, use <literal>setMaxResult()</literal>/<literal>setFirstResult()</literal>
                    if you need offline pagination functionality.
                </para>

            </sect3>

            <sect3 id="objectstate-querying-executing-named">
                <title>Externalizing named queries</title>

                <para>
                    You may also define named queries in the mapping document. (Remember to use a
                    <literal>CDATA</literal> section if your query contains characters that could
                    be interpreted as markup.)
                </para>

                <programlisting><![CDATA[<query name="eg.DomesticCat.by.name.and.minimum.weight"><![CDATA[
    from eg.DomesticCat as cat
        where cat.name = ?
        and cat.weight > ?
] ]></query>]]></programlisting>

                <para>
                    Parameter binding and executing is done programatically:
                </para>

                <programlisting><![CDATA[Query q = sess.getNamedQuery("eg.DomesticCat.by.name.and.minimum.weight");
q.setString(0, name);
q.setInt(1, minWeight);
List cats = q.list();]]></programlisting>

                <para>
                    Note that the actual program code is independent of the query language that
                    is used, you may also define native SQL queries in metadata, or migrate
                    existing queries to Hibernate by placing them in mapping files.
                </para>

            </sect3>

        </sect2>

        <sect2 id="objectstate-filtering" revision="1">
            <title>Filtering collections</title>
            <para>
                A collection <emphasis>filter</emphasis> is a special type of query that may be applied to
                a persistent collection or array. The query string may refer to <literal>this</literal>,
                meaning the current collection element.
            </para>

            <programlisting><![CDATA[Collection blackKittens = session.createFilter(
    pk.getKittens(), 
    "where this.color = ?")
    .setParameter( Color.BLACK, Hibernate.custom(ColorUserType.class) )
    .list()
);]]></programlisting>
        
            <para>
                The returned collection is considered a bag, and it's a copy of the given
                collection. The original collection is not modified (this is contrary to
                the implication of the name "filter", but consistent with expected behavior).
            </para>

            <para>
                Observe that filters do not require a <literal>from</literal> clause (though they may have
                one if required). Filters are not limited to returning the collection elements themselves.
            </para>

            <programlisting><![CDATA[Collection blackKittenMates = session.createFilter(
    pk.getKittens(), 
    "select this.mate where this.color = eg.Color.BLACK.intValue")
    .list();]]></programlisting>

            <para>
                Even an empty filter query is useful, e.g. to load a subset of elements in a
                huge collection:
            </para>

            <programlisting><![CDATA[Collection tenKittens = session.createFilter(
    mother.getKittens(), "")
    .setFirstResult(0).setMaxResults(10)
    .list();]]></programlisting>

        </sect2>

        <sect2 id="objecstate-querying-criteria" revision="1">
           <title>Criteria queries</title>

            <para>
                HQL is extremely powerful but some developers prefer to build queries dynamically,
                using an object-oriented API, rather than building query strings. Hibernate provides
                an intuitive <literal>Criteria</literal> query API for these cases:
            </para>

            <programlisting><![CDATA[Criteria crit = session.createCriteria(Cat.class);
crit.add( Expression.eq( "color", eg.Color.BLACK ) );
crit.setMaxResults(10);
List cats = crit.list();]]></programlisting>
    
            <para>
                The <literal>Criteria</literal> and the associated <literal>Example</literal>
                API are discussed in more detail in <xref linkend="querycriteria"/>.
            </para>

        </sect2>

        <sect2 id="objectstate-querying-nativesql" revision="2">
            <title>Queries in native SQL</title>

            <para>
                You may express a query in SQL, using <literal>createSQLQuery()</literal> and
                let Hibernate take care of the mapping from result sets to objects. Note
                that you may at any time call <literal>session.connection()</literal> and
                use the JDBC <literal>Connection</literal> directly. If you chose to use the
                Hibernate API, you must enclose SQL aliases in braces:
            </para>

            <programlisting><![CDATA[List cats = session.createSQLQuery(
    "SELECT {cat.*} FROM CAT {cat} WHERE ROWNUM<10",
    "cat",
    Cat.class
).list();]]></programlisting>
                
            <programlisting><![CDATA[List cats = session.createSQLQuery(
    "SELECT {cat}.ID AS {cat.id}, {cat}.SEX AS {cat.sex}, " +
           "{cat}.MATE AS {cat.mate}, {cat}.SUBCLASS AS {cat.class}, ... " +
    "FROM CAT {cat} WHERE ROWNUM<10",
    "cat",
    Cat.class
).list()]]></programlisting>

            <para>
                SQL queries may contain named and positional parameters, just like Hibernate queries.
                More information about native SQL queries in Hibernate can be found in
                <xref linkend="querysql"/>.
            </para>

        </sect2>

    </sect1>

    <sect1 id="objectstate-modifying" revision="1">
        <title>Modifying persistent objects</title>

        <para>
            <emphasis>Transactional persistent instances</emphasis> (ie. objects loaded, saved, created or
            queried by the <literal>Session</literal>) may be manipulated by the application
            and any changes to persistent state will be persisted when the <literal>Session</literal>
            is <emphasis>flushed</emphasis> (discussed later in this chapter). There is no need
            to call a particular method (like <literal>update()</literal>, which has a different
            purpose) to make your modifications persistent. So the most straightforward way to update
            the state of an object is to <literal>load()</literal> it,
            and then manipulate it directly, while the <literal>Session</literal> is open:
        </para>

        <programlisting><![CDATA[DomesticCat cat = (DomesticCat) sess.load( Cat.class, new Long(69) );
cat.setName("PK");
sess.flush();  // changes to cat are automatically detected and persisted]]></programlisting>

        <para>
            Sometimes this programming model is inefficient since it would require both an SQL
            <literal>SELECT</literal> (to load an object) and an SQL <literal>UPDATE</literal>
            (to persist its updated state) in the same session. Therefore Hibernate offers an
            alternate approach, using detached instances.
        </para>

        <para>
            <emphasis>Note that Hibernate does not offer its own API for direct execution of
            <literal>UPDATE</literal> or <literal>DELETE</literal> statements. Hibernate is a
            <emphasis>state management</emphasis> service, you don't have to think in
            <emphasis>statements</emphasis> to use it. JDBC is a perfect API for executing
            SQL statements, you can get a JDBC <literal>Connection</literal> at any time
            by calling <literal>session.connection()</literal>. Furthermore, the notion
            of mass operations conflicts with object/relational mapping for online
            transaction processing-oriented applications. Future versions of Hibernate
            may however provide special mass operation functions. See <xref linkend="batch"/>
            for some possible batch operation tricks.</emphasis>
        </para>

    </sect1>

    <sect1 id="objectstate-detached" revision="2">
        <title>Modifying detached objects</title>

        <para>
            Many applications need to retrieve an object in one transaction, send it to the
            UI layer for manipulation, then save the changes in a new transaction.
            Applications that use this kind of approach in a high-concurrency environment
            usually use versioned  data to ensure isolation for the "long" unit of work.
        </para>

        <para>
            Hibernate supports this model by providing for reattachment of detached instances
            using the <literal>Session.update()</literal> or <literal>Session.merge()</literal>
            methods:
        </para>

        <programlisting><![CDATA[// in the first session
Cat cat = (Cat) firstSession.load(Cat.class, catId);
Cat potentialMate = new Cat();
firstSession.save(potentialMate);

// in a higher layer of the application
cat.setMate(potentialMate);

// later, in a new session
secondSession.update(cat);  // update cat
secondSession.update(mate); // update mate]]></programlisting>

        <para>
            If the <literal>Cat</literal> with identifier <literal>catId</literal> had already
            been loaded  by <literal>secondSession</literal> when the application tried to
            reattach it, an exception would have been thrown.
        </para>

        <para>
            Use <literal>update()</literal> if you are sure that the session does
            not contain an already persistent instance with the same identifier, and
            <literal>merge()</literal> if you want to merge your modifications at any time
            without consideration of the state of the session. In other words, <literal>update()</literal>
            is usually the first method you would call in a fresh session, ensuring that
            reattachment of your detached instances is the first operation that is executed.
        </para>

        <para>
            The application should individually <literal>update()</literal> detached instances
            reachable from the given detached instance if and <emphasis>only</emphasis> if it wants
            their state also updated. This can be automated of course, using <emphasis>transitive
            persistence</emphasis>, see <xref linkend="objectstate-transitive"/>.
        </para>

        <para>
            The <literal>lock()</literal> method also allows an application to reassociate
            an object with a new session. However, the detached instance has to be unmodified!
        </para>

        <programlisting><![CDATA[//just reassociate:
sess.lock(fritz, LockMode.NONE);
//do a version check, then reassociate:
sess.lock(izi, LockMode.READ);
//do a version check, using SELECT ... FOR UPDATE, then reassociate:
sess.lock(pk, LockMode.UPGRADE);]]></programlisting>

        <para>
            Note that <literal>lock()</literal> can be used with various
            <literal>LockMode</literal>s, see the API documentation and the
            chapter on transaction handling for more information. Reattachment is not
            the only usecase for <literal>lock()</literal>.
        </para>

        <para>
            Other models for long units of work are discussed in <xref linkend="transactions-optimistic"/>.
        </para>

    </sect1>

    <sect1 id="objectstate-saveorupdate">
        <title>Automatic state detection</title>

        <para>
            Hibernate users have requested a general purpose method that either saves a
            transient instance by generating a new identifier or updates/reattaches
            the detached instances associated with its current identifier.
            The <literal>saveOrUpdate()</literal> method implements this functionality.
        </para>

        <programlisting><![CDATA[// in the first session
Cat cat = (Cat) firstSession.load(Cat.class, catID);

// in a higher tier of the application
Cat mate = new Cat();
cat.setMate(mate);

// later, in a new session
secondSession.saveOrUpdate(cat);   // update existing state (cat has a non-null id)
secondSession.saveOrUpdate(mate);  // save the new instance (mate has a null id)]]></programlisting>

        <para>
            The usage and semantics of <literal>saveOrUpdate()</literal> seems to be confusing
            for new users. Firstly, so long as you are not trying to use instances from one session
            in another new session, you should not need to use <literal>update()</literal>,
            <literal>saveOrUpdate()</literal>, or <literal>merge()</literal>. Some whole
            applications will never use either of these methods.
        </para>

        <para>
            Usually <literal>update()</literal> or <literal>saveOrUpdate()</literal> are used in
            the following scenario:
        </para>

        <itemizedlist spacing="compact">
            <listitem>
                <para>
                    the application loads an object in the first session
                </para>
            </listitem>
            <listitem>
                <para>
                    the object is passed up to the UI tier
                </para>
            </listitem>
            <listitem>
                <para>
                    some modifications are made to the object
                </para>
            </listitem>
            <listitem>
                <para>
                    the object is passed back down to the business logic tier
                </para>
            </listitem>
            <listitem>
                <para>
                    the application persists these modifications by calling
                    <literal>update()</literal> in a second session
                </para>
            </listitem>
        </itemizedlist>

        <para>
            <literal>saveOrUpdate()</literal> does the following:
        </para>

        <itemizedlist spacing="compact">
            <listitem>
                <para>
                    if the object is already persistent in this session, do nothing
                </para>
            </listitem>
            <listitem>
                <para>
                    if another object associated with the session has the same identifier, 
                    throw an exception
                </para>
            </listitem>
            <listitem>
                <para>
                    if the object has no identifier property, <literal>save()</literal> it
                </para>
            </listitem>
            <listitem>
                <para>
                    if the object's identifier has the value assigned to a newly instantiated
                    object, <literal>save()</literal> it
                </para>
            </listitem>
            <listitem>
                <para>
                    if the object is versioned (by a <literal>&lt;version&gt;</literal> or
                    <literal>&lt;timestamp&gt;</literal>), and the version property value
                    is the same value assigned to a newly instantiated object, 
                    <literal>save()</literal> it
                </para>
            </listitem>
            <listitem>
                <para>
                    otherwise <literal>update()</literal> the object
                </para>
            </listitem>
        </itemizedlist>

        <para>
            and <literal>merge()</literal> is very different:
        </para>

        <itemizedlist spacing="compact">
            <listitem>
                <para>
                    if there is a persistent instance with the same identifier currently 
                    associated with the session, copy the state of the given object onto 
                    the persistent instance
                </para>
            </listitem>
            <listitem>
                <para>
                    if there is no persistent instance currently associated with the session, 
                    try to load it from the database, or create a new persistent instance
                </para>
            </listitem>
            <listitem>
                <para>
                    the persistent instance is returned
                </para>
            </listitem>
            <listitem>
                <para>
                    the given instance does not become associated with the session, it
                    remains detached
                </para>
            </listitem>
        </itemizedlist>

    </sect1>

    <sect1 id="objectstate-deleting" revision="1">
        <title>Deleting persistent objects</title>

        <para>
            <literal>Session.delete()</literal> will remove an object's state from the database.
            Of course, your application might still hold a reference to a deleted object.
            It's best to think of <literal>delete()</literal> as making a persistent instance
            transient.
        </para>

        <programlisting><![CDATA[sess.delete(cat);]]></programlisting>

        <para>
            You may delete objects in any order you like, without risk of foreign key
            constraint violations. It is still possible to violate a <literal>NOT
            NULL</literal> constraint on a foreign key column by deleting objects in
            the wrong order, e.g. if you delete the parent, but forget to delete the
            children.
        </para>

    </sect1>
    
    <sect1 id="objectstate-replicating" revision="1">
    	<title>Replicating object between two different datastores</title>
    	
    	<para>
    	    It is occasionally useful to be able to take a graph of persistent instances
    	    and make them persistent in a different datastore, without regenerating identifier
    	    values.
    	</para>
    	
        <programlisting><![CDATA[//retrieve a cat from one database
Session session1 = factory1.openSession();
Transaction tx1 = session1.beginTransaction();
Cat cat = session1.get(Cat.class, catId);
tx1.commit();
session1.close();

//reconcile with a second database
Session session2 = factory2.openSession();
Transaction tx2 = session2.beginTransaction();
session2.replicate(cat, ReplicationMode.LATEST_VERSION);
tx2.commit();
session2.close();]]></programlisting>

        <para>
            The <literal>ReplicationMode</literal> determines how <literal>replicate()</literal>
            will deal with conflicts with existing rows in the database.
        </para>
        
        <itemizedlist spacing="compact">
            <listitem>
                <para>
                    <literal>ReplicationMode.IGNORE</literal> - ignore the object when there is
                    an existing database row with the same identifier
                </para>
            </listitem>
            <listitem>
                <para>
                    <literal>ReplicationMode.OVERWRITE</literal> - overwrite any existing database 
                    row with the same identifier
                </para>
            </listitem>
            <listitem>
                <para>
                    <literal>ReplicationMode.EXCEPTION</literal> - throw an exception if there is
                    an existing database row with the same identifier
                </para>
            </listitem>
            <listitem>
                <para>
                    <literal>ReplicationMode.LATEST_VERSION</literal> - overwrite the row if its
                    version number is earlier than the version number of the object, or ignore
                    the object otherwise
                </para>
            </listitem>
        </itemizedlist>

        <para>
            Usecases for this feature include reconciling data entered into different database
            instances, upgrading system configuration information during product upgrades,
            rolling back changes made during non-ACID transactions and more.
        </para>
    	
    </sect1>

    <sect1 id="objectstate-flushing">
        <title>Flushing the Session</title>

        <para>
            From time to time the <literal>Session</literal> will execute the SQL statements 
            needed to synchronize the JDBC connection's state with the state of objects held in 
            memory. This process, <emphasis>flush</emphasis>, occurs by default at the following 
            points
        </para>

        <itemizedlist spacing="compact">
            <listitem>
                <para>
                    before some query executions
                </para>
            </listitem>
            <listitem>
                <para>
                    from <literal>org.hibernate.Transaction.commit()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    from <literal>Session.flush()</literal>
                </para>
            </listitem>
        </itemizedlist>

        <para>
            The SQL statements are issued in the following order
        </para>

        <orderedlist spacing="compact">
            <listitem>
                <para>
                    all entity insertions, in the same order the corresponding objects
                    were saved using <literal>Session.save()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    all entity updates
                </para>
            </listitem>
            <listitem>
                <para>
                    all collection deletions
                </para>
            </listitem>
            <listitem>
                <para>
                    all collection element deletions, updates and insertions
                </para>
            </listitem>
            <listitem>
                <para>
                    all collection insertions
                </para>
            </listitem>
            <listitem>
                <para>
                    all entity deletions, in the same order the corresponding objects
                    were deleted using <literal>Session.delete()</literal>
                </para>
            </listitem>
        </orderedlist>

        <para>
            (An exception is that objects using <literal>native</literal> ID generation are 
            inserted when they are saved.)
        </para>

        <para>
            Except when you explicity <literal>flush()</literal>, there are absolutely no 
            guarantees about <emphasis>when</emphasis> the <literal>Session</literal> executes 
            the JDBC calls, only the <emphasis>order</emphasis> in which they are executed.
            However, Hibernate does guarantee that the <literal>Query.list(..)</literal> 
            will never return stale data; nor will they return the wrong data.
        </para>

        <para>
            It is possible to change the default behavior so that flush occurs less frequently.
            The <literal>FlushMode</literal> class defines three different modes: only flush
            at commit time (and only when the Hibernate <literal>Transaction</literal> API
            is used), flush automatically using the explained routine, or never flush unless
            <literal>flush()</literal> is called explicitly. The last mode is useful for long running
            units of work, where a <literal>Session</literal> is kept open and disconnected for
            a long time (see <xref linkend="transactions-optimistic-longsession"/>).
        </para>

        <programlisting><![CDATA[sess = sf.openSession();
Transaction tx = sess.beginTransaction();
sess.setFlushMode(FlushMode.COMMIT); // allow queries to return stale state

Cat izi = (Cat) sess.load(Cat.class, id);
izi.setName(iznizi);

// might return stale data
sess.find("from Cat as cat left outer join cat.kittens kitten");

// change to izi is not flushed!
...
tx.commit(); // flush occurs
sess.close();]]></programlisting>

        <para>
            During flush, an exception might occur (e.g. if a DML operation violates a constraint).
            Since handling exceptions involves some understanding of Hibernate's transactional 
            behavior, we discuss it in <xref linkend="transactions"/>.
        </para>

    </sect1>

    <sect1 id="objectstate-transitive">
        <title>Transitive persistence</title>

        <para>
            It is quite cumbersome to save, delete, or reattach individual objects,
            especially if you deal with a graph of associated objects. A common case is
            a parent/child relationship. Consider the following example:
        </para>

        <para>
            If the children in a parent/child relationship would be value typed (e.g. a collection
            of addresses or strings), their lifecycle would depend on the parent and no
            further action would be required for convenient "cascading" of state changes.
            When the parent is saved, the value-typed child objects are saved as
            well, when the parent is deleted, the children will be deleted, etc. This
            even works for operations such as the removal of a child from the collection;
            Hibernate will detect this and, since value-typed objects can't have shared
            references, delete the child from the database.
        </para>

        <para>
            Now consider the same scenario with parent and child objects being entities,
            not value-types (e.g. categories and items, or parent and child cats). Entities
            have their own lifecycle, support shared references (so removing an entity from
            the collection does not mean it can be deleted), and there is by default no
            cascading of state from one entity to any other associated entities. Hibernate
            does not implement <emphasis>persistence by reachability</emphasis> by default.
        </para>

        <para>
            For each basic operation of the Hibernate session - including <literal>persist(), merge(),
            saveOrUpdate(), delete(), lock(), refresh(), evict(), replicate()</literal> - there is a 
            corresponding cascade style. Respectively, the cascade styles are named <literal>create, 
            merge, save-update, delete, lock, refresh, evict, replicate</literal>. If you want an 
            operation to be cascaded along an association, you must indicate that in the mapping
            document. For example:
        </para>
        
        <programlisting><![CDATA[<one-to-one name="person" cascade="persist"/>]]></programlisting>
        
        <para>
            Cascade styles my be combined:
        </para>
        
        <programlisting><![CDATA[<one-to-one name="person" cascade="persist,delete,lock"/>]]></programlisting>
        
        <para>
            You may even use <literal>cascade="all"</literal> to specify that <emphasis>all</emphasis>
            operations should be cascaded along the association. The default <literal>cascade="none"</literal>
            specifies that no operations are to be cascaded.
        </para>
        
        <para>
            A special cascade style, <literal>delete-orphan</literal>, applies only to one-to-many
            associations, and indicates that the <literal>delete()</literal> operation should
            be applied to any child object that is removed from the association.
        </para>


        <para>
            Recommendations:
        </para>

       <itemizedlist spacing="compact">
            <listitem>
                <para>
                    It doesn't usually make sense to enable cascade on a <literal>&lt;many-to-one&gt;</literal>
                    or <literal>&lt;many-to-many&gt;</literal> association. Cascade is often useful for 
                    <literal>&lt;one-to-one&gt;</literal> and <literal>&lt;one-to-many&gt;</literal>
                    associations.
                </para>
            </listitem>
            <listitem>
                <para>
                    If the child object's lifespan is bounded by the lifespan of the of the parent
                    object make it a <emphasis>lifecycle object</emphasis> by specifying
                    <literal>cascade="all,delete-orphan"</literal>.
                </para>
            </listitem>
            <listitem>
                <para>
                    Otherwise, you might not need cascade at all. But if you think that you will often be
                    working with the parent and children together in the same transaction, and you want to save 
                    yourself some typing, consider using <literal>cascade="persist,merge,save-update"</literal>.
                </para>
            </listitem>
        </itemizedlist>

        <para>
            Mapping an association (either a single valued association, or a collection) with 
            <literal>cascade="all"</literal> marks the association as a 
            <emphasis>parent/child</emphasis> style relationship where save/update/delete of the 
            parent results in save/update/delete of the child or children.
        </para>
        <para>
            Futhermore, a mere reference to a child from a persistent parent will result in 
            save/update of the child. This metaphor is incomplete, however. A child which becomes 
            unreferenced by its parent is <emphasis>not</emphasis> automatically deleted, except 
            in the case of a <literal>&lt;one-to-many&gt;</literal> association mapped with
            <literal>cascade="delete-orphan"</literal>. The precise semantics of cascading 
            operations for a parent/child relationship are as follows:
        </para>

       <itemizedlist spacing="compact">
            <listitem>
                <para>
                    If a parent is passed to <literal>persist()</literal>, all children are passed to 
                    <literal>persist()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    If a parent is passed to <literal>merge()</literal>, all children are passed to 
                    <literal>merge()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    If a parent is passed to <literal>save()</literal>, <literal>update()</literal> or 
                    <literal>saveOrUpdate()</literal>, all children are passed to <literal>saveOrUpdate()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    If a transient or detached child becomes referenced by a persistent parent, 
                    it is passed to <literal>saveOrUpdate()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    If a parent is deleted, all children are passed to <literal>delete()</literal>
                </para>
            </listitem>
            <listitem>
                <para>
                    If a child is dereferenced by a persistent parent, <emphasis>nothing
                    special happens</emphasis> - the application should explicitly delete 
                    the child if necessary - unless <literal>cascade="delete-orphan"</literal>, 
                    in which case the "orphaned" child is deleted.
                </para>
            </listitem>
        </itemizedlist>

    </sect1>

    <sect1 id="objectstate-metadata">
        <title>Using metadata</title>

        <para>
            Hibernate requires a very rich meta-level model of all entity and value types. From time
            to time, this model is very useful to the application itself. For example, the application
            might use Hibernate's metadata to implement a "smart" deep-copy algorithm that understands
            which objects should be copied (eg. mutable value types) and which should not (eg. 
            immutable value types and, possibly, associated entities).
        </para>
        <para>
            Hibernate exposes metadata via the <literal>ClassMetadata</literal> and
            <literal>CollectionMetadata</literal> interfaces and the <literal>Type</literal>
            hierarchy. Instances of the metadata interfaces may be obtained from the 
            <literal>SessionFactory</literal>.
        </para>

        <programlisting><![CDATA[Cat fritz = ......;
ClassMetadata catMeta = sessionfactory.getClassMetadata(Cat.class);

Object[] propertyValues = catMeta.getPropertyValues(fritz);
String[] propertyNames = catMeta.getPropertyNames();
Type[] propertyTypes = catMeta.getPropertyTypes();

// get a Map of all properties which are not collections or associations
Map namedValues = new HashMap();
for ( int i=0; i<propertyNames.length; i++ ) {
    if ( !propertyTypes[i].isEntityType() && !propertyTypes[i].isCollectionType() ) {
        namedValues.put( propertyNames[i], propertyValues[i] );
    }
}]]></programlisting>
        
    </sect1>

</chapter>