转自:http://hpingliu.blog.ccidnet.com/blog-htm-do-showone-type-blog-itemid-144245-uid-48802.html]
作为测试,我们新建一个TUser对象,其image属性中,保存了一个图片文件的二进制内容。而其resume属性,我们以一个简单的字符串作为填充。
1 TUser user = new TUser();
2
3 user.setAge( new Integer( 20 ));
4
5 user.setName( " Shark " );
6
7 FilelnputStream imgis = new FileinputStream( " C:\\inimage.jpg "
8
9 Blob img = Hibernate.createBlob(imgis);
10
11 user.setlmage(img);
12
13 Clob resume = Hibernate.createClob( " This is Clob " );
14
15 user. setResume(resume);
16
17 Transaction tx = session.beginTransaction();
18
19 session.save(user);
20
21 tx.commit();
22
23
上面的代码中,我们通过Hibemate.createBlob和Hibemate.createClob创建了对应的Blob和Clob对象。其中Blob对象基于一个FileInputStream构建,而Clob对象基于一个字符串构建。
完成了写入操作,对应的读取操作代码如下:
1 // 假设库表记录的id字段等于3
2 TUser user = (TUser) session.load(TUger.elaa., load(TUser. class , new Integer( 3 ));
3 Clob resume = user.getResume();
4 // 通过Clob.getSubString()方法获取Clob字段内容
5 System.out.println( " User resume=> " + resume.getSubString( 1 ,( int )resume.length()));
6 Blob img = user.getImage();
7 // 通过Blob.getBinaryS=ream()方法获取二进制流
8 InputStream is = img.getBinaryStream();
9 FileOutputStream fos = new FileOutputStream( " C:\\outimage.jpg " );
10 byte [] buf = new byte ( 102400 );
11 int len;
12 while ((len = is.read(buf)) !=- 1 ) {
13 fos.write(buf, 0 ,len);
14 }
15 fos.close();
16 is.close();
17
18
通过上面的代码,我们完成了针对SQLServer的Blob/Clob型字段操作.看起来非常简单,不过,并非侮种数据库都如此友善。让我们接着来看看Oracle数据库下的Blob/Clob字段读写,
通过修改hibernate.cfg.xml中的Dialect配置完成数据库切换后,我们再次运行上面的TUser对象保存例程。
程序运行期间抛出异常:
Hibernate:select hibernate_sequence.nextval from dual
Hibernate:insert into T_USER (name, age, image,resume. id) values(?, ?, ?, ?, ?)
17:27:24,161 ERROR JDBCExceptionReporter:58 - - 不允许的操作: Streams type cannot be used in batching
17:27:24,171 ERROR Sessionlmpl:2399 - Could not synchronize database state with session
net.sf.hibernate.exception.GenericJDBCException:could not insert:[com.redsaga.hibernate.db.entity.TUser#6]
...
观察异常信息:streams type cannot be used in batching.这意味着Oracle JDBC不允许流操作以批量方式执行(Oracle CLOB采用流机制作为数据读写方式)。
这种错误一般发生在hibernate.cfg.xml中的hibernate jdbc.batch_size设定大于0的情况,将hibernate.jdbc.batch_size修改为0即可消除。
<hibernate-configuration>
<session-factory>
...
<property name='hibernate. jdbc.batch_size">0</property>
...
</session-factory>
</hibernate-configuration>
再次尝试启动代码,发现依然无法成功运行,抛出异常如下:
-
- Hibernate: select hibernate_sequence.nextval from dual
- Hibernate: insert into T--USER (name, age, image,resume,id) values(?,?,?,?,?)
- 19:02:21,054 ERROR JDBCExceptionReporter:58 一IO异常:Connection reset bypeer: socket write error
- 19:02:21,054 ERROR JDBCExceptionReporter:58 一I。异常:Connection reset by peer:socket write error
- 19:02:21 。064 ERROR JDBCExceptionReporter:58一Io异常:Connection reset by peer: socket wr主to error
- 19:02:21,064 ERROR Sessionlrnpl:2399 一Could not synchronize database state with session
- net.sf.hibernate.exception.GenericJDSCException: could not insert:[com.redsaga.hibernate.db.entity.TUser#27]
- ...
Hibernate: select hibernate_sequence.nextval from dual
Hibernate: insert into T--USER (name, age, image,resume,id) values(?,?,?,?,?)
19:02:21,054 ERROR JDBCExceptionReporter:58 一IO异常:Connection reset bypeer: socket write error
19:02:21,054 ERROR JDBCExceptionReporter:58 一I。异常:Connection reset by peer:socket write error
19:02:21 。064 ERROR JDBCExceptionReporter:58一Io异常:Connection reset by peer: socket wr主to error
19:02:21,064 ERROR Sessionlrnpl:2399 一Could not synchronize database state with session
net.sf.hibernate.exception.GenericJDSCException: could not insert:[com.redsaga.hibernate.db.entity.TUser#27]
...
为什么会出现这样的情况?
问题出在Oracce中Blob/Clob字段独特的访问方式,Oracle Blob/Clob字段本身拥有一个游标(cursor) , JDBC必须通过游标对Blob/ Clob字段进行操作,在Blob/Clob字段被创建之前,我们无法获取其游标句柄,这也就意味着,我们必须首先创建一个空Blob/Clob字段,再从这个空Blob/Clob字段获取游标,写入我们所期望保存的数据。
如果用JDBC代码来表示这个过程,则得到如下代码:
-
-
- dbconn.setAutoCommit(falee);
-
- PreparedStatenent preStmt=
- dbconn.prepareStatement(
- "insert into T_USER (name, age, id,image,resume) values
- (?,?,?,?,?)");
- preStmt.setString(1,"Shark");
- preStmt.setInt(2,20);
- preStmt.setInt(3,5);
-
- preStmt.setBlob(4 ,oracle.sgl.BLOB.empty_lob());
- preStmt.setClob(5,oracle.sgl.CLOB.empty_lob());
- preStmt.executeUpdate();
- preStmt.close():
-
- preStmt=
- dbconn.prepareStatement(
- "select image,resume from T_USER where id=?for update');
- preStmt.setint(l,5);
- ResultSet rset=preStmt.executeQuery();
-
-
- rset.next();
- oracle.sql.BLOB imqBlob = (oracle.sql.BLOB) rset.getBlob(1);
- oracle.sql.CLOB resClob = (oracle.sql.CLOB) rset.getClob(2);
-
- FileInputStream inStream = new FileinputStream("c\\inimage.jpg");
- OutputStream outStream = imgBlob.getBinaryOutputStream();
- byte[] buf=new byte[10240];
- int len;
- while((len=inStream.read(buf))>0){
- outStream.write(buf,0,len);
- }
- inStream.close();
- outStream.close():
-
- resClob.putString(1 ,"This is my Glob"
-
- preStmt= dbconn.prepareStatement("update T_USER set image=?, resume=? where id=?");
- preStmt.setBlob(1,imgBlob);
- preStmt.setClob(2,resClob):
- preStmt.setlnt(3 ,5);
- preStmt.executeUpdate();
- preStmt.close():
- dbconn.commit();
- dbconn.close():
//... 获取JDBC连接
dbconn.setAutoCommit(falee);
// =======插入数据,BLOB CLOB字段插入空值
PreparedStatenent preStmt=
dbconn.prepareStatement(
"insert into T_USER (name, age, id,image,resume) values
(?,?,?,?,?)");
preStmt.setString(1,"Shark");
preStmt.setInt(2,20);
preStmt.setInt(3,5);
// 通过oracle.sgl.BLOB/CLOB.empty_lob()方法构造空Blob/Clob对象
preStmt.setBlob(4 ,oracle.sgl.BLOB.empty_lob());
preStmt.setClob(5,oracle.sgl.CLOB.empty_lob());
preStmt.executeUpdate();
preStmt.close():
//========== 再次从库表读出,获得Blob/Clob句柄
preStmt=
dbconn.prepareStatement(
"select image,resume from T_USER where id=?for update');
preStmt.setint(l,5);
ResultSet rset=preStmt.executeQuery();
// 注意我们这里需要引用Oracle原生BLOB定义,如果使用了Weblogic JDBC Vendor
// 则应使用weblogic.jdbc.vendor.oracle. OracleThinBLob/OracleThinCLOb
rset.next();
oracle.sql.BLOB imqBlob = (oracle.sql.BLOB) rset.getBlob(1);
oracle.sql.CLOB resClob = (oracle.sql.CLOB) rset.getClob(2);
//======= 将二进创数据写入Blob
FileInputStream inStream = new FileinputStream("c\\inimage.jpg");
OutputStream outStream = imgBlob.getBinaryOutputStream();
byte[] buf=new byte[10240];//10K 读取缓存
int len;
while((len=inStream.read(buf))>0){
outStream.write(buf,0,len);
}
inStream.close();
outStream.close():
//======= 将字符串写入Clob
resClob.putString(1 ,"This is my Glob"
//======= 将Blob/Clob字段更新到数据序
preStmt= dbconn.prepareStatement("update T_USER set image=?, resume=? where id=?");
preStmt.setBlob(1,imgBlob);
preStmt.setClob(2,resClob):
preStmt.setlnt(3 ,5);
preStmt.executeUpdate();
preStmt.close():
dbconn.commit();
dbconn.close():
上面的代码说明了Oracle中Blob/Clob字段操作的一般机制,那么,基于Hibernate的持久层实现中,应该如何对Blob/Clob字段进行处理?
我们知道,Hibernate底层数据访问机制仍然是基于JDBC实现,那么也就意味着我们必须在Hibernate中模拟JDBC的访问流程:
-
- TUser user=new TUser();
- user.setAge(new Integer(20));
- user.setName("Shark');
- user.setImage(Hibernate.createSlob(new byte [1])):
- user.setResume(Hibernate.createClob(" "));
- Transaction tx=session.beginTransaction();
- session.save(user):
-
- session.flush();
-
- session.refresh(user, LockMode.UPGRADE);
-
- oracle.sql.BLOB blob=(oracle.sql.BLOB)user.getImage();
- OutputStream out=blob. getBinaryOutputStream();
- FileInputStream imgis=new FileInputStream("C:\\inimage.jpg");
- byte[] buf=new byte[10240];
- int len;
- while((len=imgis.read(buf))>0){
- out.write(buf,0,len);
- }
- imgis.close();
- out.close();
-
- oracle.sql.CLOB clob=(oracle.sgl.CLOB) user.getResume();
- java.io.Writer writer = clob.getCharacterOutputStream();
- writer.write("this is my resume');
- writer.close();
- session.save(user);
- tx.commit();
-
- <PRE class=java name="code">
-
- 实际应用中,对于Clob字段,我们可以简单地将其映射为String类型,不过在这种情况下需要注意,Oracle Thin Driver对Clob字段支持尚有欠缺,当Clob内容超出4000字节时将无法读取,而Oracle OCI Driver(需要在本地安装Oracle客户端组件)则可以成功地完成大容量Clob字段的操作。
- 上面的代码中,我们通过Session.save/flush/refresh方法的组合使用,实现了上面JDBC代码中的Blob/Clob访问逻辑。
- Blob/Clob 字段的Hibernate保存实现如上所述,相对来讲,读取则没有太多的障碍,之前的读取代码依然可以正常运行。
- 对于上面的实现,相信大家都感觉到了一些Bad Smell,如果Blob/Clob字段普遍存在,那么我们的持久层逻辑中可能遍布如此复杂的数据存储逻辑、并与数据库原生类紧密祸
- 如何解决这些问题?
- 回忆之前关于自定义数据类型的讨论。通过自定义数据类型我们可以对数据的通用特征进行抽象,那么,对于Oracle的Blob/Clob字段,我们是否可以也对其进行抽象,并以其作为所有Oracle Blob/Clob字段的映射类型?
- 下面的StringClobType实现了这一目标:
-
- <PRE class=java name="code">
- public class StringClobType implements UserType{
- private static final String ORACLE_DRIVER_NAME="Oracle JDBC driver";
- private static final int ORACLE_DRIVER_MAJOR_VERSION=9;
- private static final int ORACLE_DRIVER_MINOR_VERSION=0;
- public int[] sqlTypes(){
- return new int[] {Types.CLOB};
- }
- public Class returnedClass{
- return String.class;
- }
- public boolean equals(Object x, object y){
- return org.apache.commons.lang.ObjectUtils.equals(x, y);
- }
- public Object nullSafeGet(ResultSet rs, String[] names, Object owner)
- throws HibernateException,SQLException{
- Clob clob=rs.getClob(names(O]);
- return(clob==null ? null:clob.getSubString(l, (int) clob.length())):
- }
- public void nullSafeSet(PreparedStatement st ,Object value, int index)
- throws HibernateException, SQLException{
- DatabaseMetaData dbMetaData=st.getConnection().getMetaData();
- if (value==null)
- st.setNull(index, sqiTypes()(0));
- else
-
- if
- (ORACLE_DRIVER_NAME.equals(dbMetaData.getDriverName( ,))(
- if((dbMetaData.getDriverMajorVersion()
- >=ORACLE-DRIVER-MAJOR-VERSION)
- &&(dbMetaData.getDriverMinorVersion()
- >=ORACLE-DRIVER-MINOR-VERSION)) {
- try {
-
- Class oracleClobClass=Class.forName('oracle.sgl.CLOB");
-
- Class partypes[]=new Class[3];
- partypes[0]=Connection.class;
- partypes[1]=Boolean.TYPE;
- partypes(2]=Integer.TYPE;
- Method createTemporaryMethod=
- oracleClobClass.getDeclaredMethod(
- "createTemporaxy “,
- partypes);
- Field durationSessionField=
- oracleClobClass.getField("DURATION-SESSION");
- Object arglist[]=new 0bject[3]:
- Connection conn=
- st.getConnection().getMetaData().getConnection();
-
-
-
- Class oracleConnectionClass=
- Class.forName("oracle.jdbc.OracleConnection");
- if(!oracleConnectionClass
- .isAssignableFrom(conn.getClass())){
- throw new HibernateException(
- "Must be a oracle.jdbc.OracleConnection:.
- +conn.getClass().getName());
- }
- arglist[0] = conn;
- arglist(1] = Boolean.TRUE;
- arolist[2] = durationSessionField.get(null);
- Object tempClob =createTemporaryMethod.invoke(null,arglist);
- partypes=new Class[l];
- partypes[0]=Integer.TYPE;
- Method openMethod =oracleClobClass.getDeclaredMethod("open",partypes);
- Field modeReadWriteField =oracleClobClass.getField("MODE_READWRITE");
- arglist = new Object(l];
- arglis[0] = modeReadWriteField.get(null);
- openMethod.invoke(tempClob, arglist);
- Method getCharacterOutputStreamMethod=oracleClobClass.getDeclaredMethod("getCharacterOutputStream',null) ;
-
- Writer tempClobWriter =(Writer)getCharacterOutputStreamMethod.invoke(tempClob,null);
-
- tempClobwriter.write((String) value);
- tempClobWriter.flush();
- tempClobWriter.Close();
-
-
- Method closeMethod=oracleClobClass.getDeclaredMethod("close", null);
- closeMethod.invoke(tempClob, null);
- st.setClob(index, (Clob) tempClob);
- )catch (ClassNotFoundException e){
- throw new HibernateException("Unable to find a required class.\n"+e.getMessage()):
- }catch (NOSuchMethodException e){
- throw new HibernateException("Unable to find a required method.\n"+e.getMessage()):
- }catch (NoSuchFieldException e){
- throw new HibernateException("Unable to find a required field.\n"+e.getMessage());
- }catch (IllegalAccessException e){
- throw new HibernateException("Unable to access a required method or field.\n"+e.getMessage());
- catch (InvocationTargetException e){
- throw new HibernateException(e.getMessage());
- { catch (IOException e){
- throw new HibernateException(e.getMessage());
- }
- else {
- throw new HibernateException(
- "No CLOBS support.Use driver version"
- +ORACLE_DRIVER_MAJOR_VERSION
- +" ,minor"
- +ORACLE_DRIVER_MINOR_VERSION);
- }
- }else {
- String str = (String) value;
- StrinaReader r = new StringReader(str);
- St.setCharacterStream(index, r, str.length());
- }
- }
- public Object deepCopy(Object value){
- if(value==null)
- return null;
- return new String((String) value);
- }
- public boolean isMutable(){
- return false
- }
- }
-
- </PRE>
- <BR>
- <BR>上面这段代码,重点在于nullSafeSet方法的实现,nullSafeSet中通过Java Reflection机制,解除了编译期的Oralce JDBC原生类依赖。同时,借助Oracle JDBC提供的原生功能完成了Clob字段的写入,Clob字段的写入操作由于涉及特定数据库内部实现细节,这里就不多费唇舌,大家可参见Oracle JDBC Java Doc.
- <BR>这段代码是由笔者根据Ali Ibrahim, Scott Miller的代码修改而来的(原版请参见httpJ/www.hibemate, org /56.html ),支持Oracle 9以上版本,Oracle 8对应的实现请参见上述网址。
- <BR>同样的道理,读者可以根据以上例程,编写自己的ByteBlobType以实现byte[]到Blob的映射。
- <BR>另外,此代码必须运行在最新版的Oracle JDBC Driver上(笔者所用版本为Oracle9i9.2.0.5 for JDK1.4,如果使用9.2.0.3或之前版本则在新建l更却删除数据时可能会遇到“nomore data read from socket”错误)。
- <BR>
- <BR> </PRE