我正在使用SQLServer 2008，但我有一个开发限制:我不能为它使用一个新的驱动程序，我必须使用“com.microsoft.jdbc.sqlserver”。SQLServerDriver”(我不能使用“com.microsoft.sqlserver.jdbc.SQLServerDriver”)。

这就是为什么解决方案conn.prepareStatement(sql, Statement.RETURN_GENERATED_KEYS)为我抛出了一个java.lang.AbstractMethodError。 在这种情况下，我找到了一个可能的解决方案，即微软提出的旧方案: 如何检索@@IDENTITY值使用JDBC

import java.sql.*; 
import java.io.*; 

public class IdentitySample
{
    public static void main(String args[])
    {
        try
        {
            String URL = "jdbc:microsoft:sqlserver://yourServer:1433;databasename=pubs";
            String userName = "yourUser";
            String password = "yourPassword";

            System.out.println( "Trying to connect to: " + URL); 

            //Register JDBC Driver
            Class.forName("com.microsoft.jdbc.sqlserver.SQLServerDriver").newInstance();

            //Connect to SQL Server
            Connection con = null;
            con = DriverManager.getConnection(URL,userName,password);
            System.out.println("Successfully connected to server"); 

            //Create statement and Execute using either a stored procecure or batch statement
            CallableStatement callstmt = null;

            callstmt = con.prepareCall("INSERT INTO myIdentTable (col2) VALUES (?);SELECT @@IDENTITY");
            callstmt.setString(1, "testInputBatch");
            System.out.println("Batch statement successfully executed"); 
            callstmt.execute();

            int iUpdCount = callstmt.getUpdateCount();
            boolean bMoreResults = true;
            ResultSet rs = null;
            int myIdentVal = -1; //to store the @@IDENTITY

            //While there are still more results or update counts
            //available, continue processing resultsets
            while (bMoreResults || iUpdCount!=-1)
            {           
                //NOTE: in order for output parameters to be available,
                //all resultsets must be processed

                rs = callstmt.getResultSet();                   

                //if rs is not null, we know we can get the results from the SELECT @@IDENTITY
                if (rs != null)
                {
                    rs.next();
                    myIdentVal = rs.getInt(1);
                }                   

                //Do something with the results here (not shown)

                //get the next resultset, if there is one
                //this call also implicitly closes the previously obtained ResultSet
                bMoreResults = callstmt.getMoreResults();
                iUpdCount = callstmt.getUpdateCount();
            }

            System.out.println( "@@IDENTITY is: " + myIdentVal);        

            //Close statement and connection 
            callstmt.close();
            con.close();
        }
        catch (Exception ex)
        {
            ex.printStackTrace();
        }

        try
        {
            System.out.println("Press any key to quit...");
            System.in.read();
        }
        catch (Exception e)
        {
        }
    }
}

这个方法对我很有效!

我希望这能有所帮助!

我只是想回答这个帖子，而不是评论。

接口java.sql.PreparedStatement

columnIndexes « You can use prepareStatement function that accepts columnIndexes and SQL statement. Where columnIndexes allowed constant flags are Statement.RETURN_GENERATED_KEYS1 or Statement.NO_GENERATED_KEYS[2], SQL statement that may contain one or more '?' IN parameter placeholders. SYNTAX « Connection.prepareStatement(String sql, int autoGeneratedKeys) Connection.prepareStatement(String sql, int[] columnIndexes) Example: PreparedStatement pstmt = conn.prepareStatement( insertSQL, Statement.RETURN_GENERATED_KEYS );

columnNames«列出columnNames，如'id'， 'uniqueID'， ....在包含应返回的自动生成键的目标表中。如果SQL语句不是INSERT语句，驱动程序将忽略它们。 语法« 连接。prepareStatement(String sql, String[] columnNames) 例子: String columnNames[] = new String[] {"id"}; PreparedStatement pstmt = conn.prepareStatement(insertSQL, columnNames);

完整的例子:

public static void insertAutoIncrement_SQL(String UserName, String Language, String Message) {
    String DB_URL = "jdbc:mysql://localhost:3306/test", DB_User = "root", DB_Password = "";

    String insertSQL = "INSERT INTO `unicodeinfo`( `UserName`, `Language`, `Message`) VALUES (?,?,?)";
            //"INSERT INTO `unicodeinfo`(`id`, `UserName`, `Language`, `Message`) VALUES (?,?,?,?)";
    int primkey = 0 ;
    try {
        Class.forName("com.mysql.jdbc.Driver").newInstance();
        Connection conn = DriverManager.getConnection(DB_URL, DB_User, DB_Password);

        String columnNames[] = new String[] { "id" };

        PreparedStatement pstmt = conn.prepareStatement( insertSQL, columnNames );
        pstmt.setString(1, UserName );
        pstmt.setString(2, Language );
        pstmt.setString(3, Message );

        if (pstmt.executeUpdate() > 0) {
            // Retrieves any auto-generated keys created as a result of executing this Statement object
            java.sql.ResultSet generatedKeys = pstmt.getGeneratedKeys();
            if ( generatedKeys.next() ) {
                primkey = generatedKeys.getInt(1);
            }
        }
        System.out.println("Record updated with id = "+primkey);
    } catch (InstantiationException | IllegalAccessException | ClassNotFoundException | SQLException e) {
        e.printStackTrace();
    }
}

使用Hibernate的NativeQuery，您需要返回一个ResultList而不是SingleResult，因为Hibernate修改了一个本机查询

INSERT INTO bla (a,b) VALUES (2,3) RETURNING id

like

INSERT INTO bla (a,b) VALUES (2,3) RETURNING id LIMIT 1

如果你试图得到一个单一的结果，这将导致大多数数据库(至少PostgreSQL)抛出一个语法错误。然后，您可以从列表中获取结果id(通常只包含一个项)。

您可以使用以下java代码来获取新的插入id。

ps = con.prepareStatement(query, Statement.RETURN_GENERATED_KEYS);
ps.setInt(1, quizid);
ps.setInt(2, userid);
ps.executeUpdate();

ResultSet rs = ps.getGeneratedKeys();
if (rs.next()) {
    lastInsertId = rs.getInt(1);
}

当使用Statement时遇到“不支持的功能”错误。RETURN_GENERATED_KEYS，试试这个:

String[] returnId = { "BATCHID" };
String sql = "INSERT INTO BATCH (BATCHNAME) VALUES ('aaaaaaa')";
PreparedStatement statement = connection.prepareStatement(sql, returnId);
int affectedRows = statement.executeUpdate();

if (affectedRows == 0) {
    throw new SQLException("Creating user failed, no rows affected.");
}

try (ResultSet rs = statement.getGeneratedKeys()) {
    if (rs.next()) {
        System.out.println(rs.getInt(1));
    }
    rs.close();
}

其中BATCHID是自动生成的id。

如果它是一个自动生成的键，那么你可以使用Statement#getGeneratedKeys()。您需要在与INSERT使用的语句相同的Statement上调用它。首先需要使用statement创建语句。RETURN_GENERATED_KEYS通知JDBC驱动程序返回密钥。

这里有一个基本的例子:

public void create(User user) throws SQLException {
    try (
        Connection connection = dataSource.getConnection();
        PreparedStatement statement = connection.prepareStatement(SQL_INSERT,
                                      Statement.RETURN_GENERATED_KEYS);
    ) {
        statement.setString(1, user.getName());
        statement.setString(2, user.getPassword());
        statement.setString(3, user.getEmail());
        // ...

        int affectedRows = statement.executeUpdate();

        if (affectedRows == 0) {
            throw new SQLException("Creating user failed, no rows affected.");
        }

        try (ResultSet generatedKeys = statement.getGeneratedKeys()) {
            if (generatedKeys.next()) {
                user.setId(generatedKeys.getLong(1));
            }
            else {
                throw new SQLException("Creating user failed, no ID obtained.");
            }
        }
    }
}

注意，您依赖于JDBC驱动程序来决定它是否工作。目前，最后的大多数版本都可以工作，但如果我是正确的，Oracle JDBC驱动程序在这方面仍然有些麻烦。MySQL和DB2已经支持它很多年了。PostgreSQL不久前开始支持它。我不能评论MSSQL，因为我从未使用过它。

对于Oracle，您可以在同一个事务中的INSERT语句后直接调用带有RETURNING子句或SELECT CURRVAL(sequencename)(或任何特定于db的语法)的CallableStatement来获得最后生成的密钥。看看这个答案。