How to use filters in jfinal to monitor Druid for SQL execution?

Abstract ： What I want to do at first is to intercept through interceptors SQL perform , But it was tested and found that , The filter can listen to at least every SQL Execution and return results of . therefore , Record this exploration .

This article is shared from Huawei cloud community 《jfinal Filter monitoring in Druid Of SQL perform 【 May 07】》, author ：KevinQ .

What I want to do at first is to intercept through interceptors SQL perform , For example, similar to PageHelper This plug-in , Through interceptors or filters , Manually modify SQL sentence , To meet certain business requirements , For example, perform paging , Or restrict access to data permissions . But I found the information that the filter is not for this , This is what the database middleware does , such as MyCat etc. .

But it was tested and found that , The filter can listen to at least every SQL Execution and return results of . therefore , Record this exploration .

Configure filters

stay jfinal In the startup configuration class of , There is a function configPlugin(Plugins me) Function to configure the plug-in , This function will be in jfinal Call at startup , The argument to this function is Plugins me, This parameter is a plug-in manager , You can add plug-ins through this plug-in manager .

Database plug-ins Druid Is added in this function .

public void configPlugin(Plugins me) {    DruidPlugin druidPlugin = createDruidPlugin_holdoa();    druidPlugin.setPublicKey(p.get("publicKeydebug").trim());    wallFilter = new WallFilter();    wallFilter.setDbType("mysql");    druidPlugin_oa.addFilter(wallFilter);    druidPlugin_oa.addFilter(new StatFilter());    me.add(druidPlugin);}

We refer to WallFilter as well as StatFilter Also create a filter class ：

import com.alibaba.druid.filter.FilterEventAdapter;public class DataScopeFilter extends FilterEventAdapter {}

We found that FilterEventAdapter There are several methods in ：

public boolean statement_execute(FilterChain chain, StatementProxy statement, String sql) throws SQLException {...}protected void statementExecuteUpdateBefore(StatementProxy statement, String sql) {...}protected void statementExecuteUpdateAfter(StatementProxy statement, String sql, int updateCount) {...}protected void statementExecuteQueryBefore(StatementProxy statement, String sql) {...}protected void statementExecuteQueryAfter(StatementProxy statement, String sql, ResultSetProxy resultSet) {...}protected void statementExecuteBefore(StatementProxy statement, String sql) {...}protected void statementExecuteAfter(StatementProxy statement, String sql, boolean result) {...}

Let's repeat these methods to see （ exclude Update Method , Because we are more concerned with query statements ）

package xxxx.xxxx;import com.alibaba.druid.filter.FilterChain;import com.alibaba.druid.filter.FilterEventAdapter;import com.alibaba.druid.proxy.jdbc.ResultSetProxy;import com.alibaba.druid.proxy.jdbc.StatementProxy;import com.jfinal.kit.LogKit;import java.sql.SQLException;public class DataScopeFilter extends FilterEventAdapter {    @Override    public boolean statement_execute(FilterChain chain, StatementProxy statement, String sql) throws SQLException {        LogKit.info("statement_execute");        return super.statement_execute(chain, statement, sql);    }    @Override    protected void statementExecuteQueryBefore(StatementProxy statement, String sql) {        LogKit.info("statementExecuteQueryBefore");        super.statementExecuteQueryBefore(statement, sql);    }    @Override    protected void statementExecuteQueryAfter(StatementProxy statement, String sql, ResultSetProxy resultSet) {        LogKit.info("statementExecuteQueryAfter");        super.statementExecuteQueryAfter(statement, sql, resultSet);    }    @Override    protected void statementExecuteBefore(StatementProxy statement, String sql) {        LogKit.info("statementExecuteBefore");        super.statementExecuteBefore(statement, sql);    }    @Override    protected void statementExecuteAfter(StatementProxy statement, String sql, boolean result) {        LogKit.info("statementExecuteAfter");        super.statementExecuteAfter(statement, sql, result);    }    @Override    public ResultSetProxy statement_executeQuery(FilterChain chain, StatementProxy statement, String sql)            throws SQLException {        LogKit.info("statement_executeQuery");        return super.statement_executeQuery(chain, statement, sql);    }}

And then again config Add a filter to the configuration class ：

druidPlugin.addFilter(new DataScopeFilter());

The execution sequence of the initiation is ：

statement_executeQuerystatementExecuteQueryBeforestatementExecuteQueryAfter

View parent code , It is found that the execution logic is , First, execute statement_executeQuery, And then because the method of the parent is called , The parent method body is ：

@Override    public ResultSetProxy statement_executeQuery(FilterChain chain, StatementProxy statement, String sql)                                                                                                         throws SQLException {        statementExecuteQueryBefore(statement, sql);        try {            ResultSetProxy resultSet = super.statement_executeQuery(chain, statement, sql);            if (resultSet != null) {                statementExecuteQueryAfter(statement, sql, resultSet);                resultSetOpenAfter(resultSet);            }            return resultSet;        } catch (SQLException error) {            statement_executeErrorAfter(statement, sql, error);            throw error;        } catch (RuntimeException error) {            statement_executeErrorAfter(statement, sql, error);            throw error;        } catch (Error error) {            statement_executeErrorAfter(statement, sql, error);            throw error;        }    }

This further triggers statementExecuteQueryBefore Methods and statementExecuteQueryAfter Method .

So we , modify statement_executeQuery Method ：

 @Override    public ResultSetProxy statement_executeQuery(FilterChain chain, StatementProxy statement, String sql)            throws SQLException {        statementExecuteQueryBefore(statement, sql);        ResultSetProxy result = chain.statement_executeQuery(statement, sql);        statementExecuteQueryAfter(statement, sql, result);        return result;    }

such , Let the output result be ：

statementExecuteQueryBeforestatement_executeQuerystatementExecuteQueryAfter

We can do it in Before perhaps After Method , such as ： Record SQL The actual executor of , Operating time , Request execution SQL The interface of .

sql Be declared final type

Find out what's going on SQL stay Druid The corresponding class in is ：DruidPooledPreparedStatement, Its class structure is ：

public class DruidPooledPreparedStatement extends DruidPooledStatement implements PreparedStatement {    private final static Log              LOG = LogFactory.getLog(DruidPooledPreparedStatement.class);    private final PreparedStatementHolder holder;    private final PreparedStatement       stmt;    private final String                  sql;    ....}

This means , Once this class is created ,SQL After setting, it cannot be modified , therefore , We need to change SQL Words , It needs to be in prepared The object is modified to the corresponding execution before it is generated SQL.

In the course of debugging , It is found that the following method needs to be overwritten ：

@Override    public PreparedStatementProxy connection_prepareStatement(FilterChain chain, ConnectionProxy connection, String sql)            throws SQLException {        //  It can be modified SQL Purpose         sql += " LIMIT 1";        PreparedStatementProxy statement = super.connection_prepareStatement(chain, connection, sql);        statementPrepareAfter(statement);        return statement;    }

We can add custom here SQL Modify the logic , For example, add data permissions and so on .

 

Click to follow , The first time to learn about Huawei's new cloud technology ~