您的位置:首页 > 数据库 > Oracle

ORACLE中表,视图名变大写和BLOB操作的问题

2006-05-30 18:48 393 查看
1. 表,视图创建后名字全部变成大写
     通过JDBC中的DatabaseMetaData.getTables函数可以来查询当前数据库的MetaData,既可以查询一些数据表和视图等很多信息。但是,ORACLE在当你创建完成,名字全部变成了大写,如果在getTables中,参数名字输入的是小写,那么将查询不到该表或者该视图。
   其实,getTables函数也是通过SQL语句来查询一个叫做all_objects的表。
 
SELECT NULL AS table_cat,
       o.owner AS table_schem,
       o.object_name AS table_name,
       o.object_type AS table_type,
       NULL AS remarks
  FROM all_objects o
  WHERE o.owner LIKE ? ESCAPE '/'
    AND o.object_name LIKE ? ESCAPE '/'
    AND o.object_type IN ('xxx', 'VIEW')
  ORDER BY table_type, table_schem, table_name
 
2。关于BLOB数据类型的存取:
  ORACLE的新版本,其中BLOB的存取一直是个比较麻烦的事情,似乎它不兼容JDBC的标准。我在MYSQL下创建,查询的BLOB字段的SQL,在ORACLE上要出现一些问题:比如getBytes(),setBytes()要出现问题等等。
  网上讨论这个的还是比较多的,最简单的办法就是把BLOB改成LONG RAW,虽然LONG RAW最大存储容量是2GB,而BLOB是4GB,但是其实大多数情况都够了。
  如果非要使用BLOB,我这里从网上找到了一些贴子和代码:
http://community.csdn.net/Expert/FAQ/FAQ_Index.asp?id=197116
http://www.zhuoda.org/hofman/22956.html
package demo;
import java.sql.*;
import java.io.*;
import java.sql.PreparedStatement;
import java.util.*;

import oracle.jdbc.driver.*;
import oracle.sql.BLOB;

/**
* Insert record in the MEDIA table
* MEDIA (file_name varchar2(256), file_content BLOB);
*/
public class BlobOracle
{
private final static String hostname = "localhost";
private final static String port = "1521";
private final static String sid = "ORCL";
private final static String username = "scott";
private final static String password = "tiger";
private static String fileLocation;
private static Connection connection;

public BlobOracle()
{
}

/**
*
* @param args
*/
public static void main(String[] args)
{
try
{
if (args.length == 0)
{
System.out.println("/n/n Usage demo.BlobOracle ");
System.exit(0);
}

fileLocation = args[0];

setConnection();
insertBLOB();

} catch (Exception ex)
{
ex.printStackTrace();
} finally
{
}
}

private static void setConnection() throws SQLException
{
DriverManager.registerDriver(new oracle.jdbc.driver.OracleDriver());
connection = DriverManager.getConnection("jdbc:oracle:thin:@"+hostname+ ":"+ port +":"+ sid , username , password);
connection.setAutoCommit(false); // we must control the commit
}

private static void insertBLOB() throws SQLException, Exception
{
BLOB blob;
File file ;
FileInputStream is;
OutputStream os;

long ts1 = System.currentTimeMillis();

//Create a statement.
PreparedStatement pstmt = connection.prepareStatement("insert into media (file_name, file_content) values (?,empty_blob())");
file = new File(fileLocation);
String fileName = file.getName();
//Set the file name and execute the query
pstmt.setString(1, fileName);
pstmt.execute();

//Take back the record for update (we will insert the blob)
//supposely the file name is the PK
pstmt = connection.prepareStatement("select file_content from media where file_name = ? for update");
pstmt.setString(1, fileName);

//Execute the query, and we must have one record so take it
ResultSet rset = pstmt.executeQuery();
rset.next();

//Use the OracleDriver resultset, we take the blob locator
blob = ((OracleResultSet)rset).getBLOB("file_content");

is = new FileInputStream(file); //Create a stream from the file
// JDBC 2.0
//os = blob.getBinaryOutputStream(); //get the output stream from the Blob to insert it
// JDBC 3.0
os = blob.setBinaryStream(0); //get the output stream from the Blob to insert it

//Read the file by chuncks and insert them in the Blob. The chunk size come from the blob
byte[] chunk = new byte[blob.getChunkSize()];
int i=-1;
System.out.println("Inserting the Blob");
while((i = is.read(chunk))!=-1)
{
os.write(chunk,0,i); //Write the chunk
System.out.print('.'); // print progression
}

// When done close the streams
is.close();
os.close();

//Close the statement and commit
pstmt.close();
long ts2 = System.currentTimeMillis();

connection.commit();
connection.close();

System.out.println("/n"+ (ts2 - ts1) +" ms" );

}

}
内容来自用户分享和网络整理,不保证内容的准确性,如有侵权内容,可联系管理员处理 点击这里给我发消息