6. Calcite添加自定义函数

1. 简介

在上篇博文中介绍了如何使用calcite进行sql验证, 但是真正在实际生产环境中我们可能需要使用到

1. 用户自定义函数(UDF): 通过代码实现对应的函数逻辑并注册给calcite
   • sql验证: 将UDF信息注册给calcite, SqlValidator.validator验证阶段即可通过验证
   • sql执行: calcite通过调用UDF逻辑实现函数逻辑
2. 自定义db函数: 数据库中创建的自定义函数
   • sql验证: 将自定义的db函数信息注册给calcite, SqlValidator.validator验证阶段即可通过验证
   • sql执行: 下推到db执行对应的db函数

此时我们需要将自定义的函数注册到calcite中, 用于sql验证和执行. 例如注册一个简单的函数 如: 将数据库中的性别字段值做字典转换.

2. Maven

<dependency>
    <groupId>org.apache.calcite</groupId>
    <artifactId>calcite-core</artifactId>
    <version>1.37.0</version>
</dependency>

<dependency>
    <groupId>mysql</groupId>
    <artifactId>mysql-connector-java</artifactId>
    <version>8.0.33</version>
</dependency>

2. UDF

如上述所说, UDF是将用户自定义的方法注册为函数使用的, 首先看一下calcite是如何注册UDF的

SchemaPlus#add(String name, Function function);

其Function的实现类如下:

1. 定义UDF实现

   public class Udf {
       public static String dictSex(String code) {
           if (StringUtils.isBlank(code)) {
               return code;
           }
           if (StringUtils.equals(code, "1")) {
               return "男";
           } else if (StringUtils.equals(code, "2")) {
               return "女";
           }
           else {
               return "未知";
           }
       }
   }
   

2. 把dictSex方法注册到calcite中, 因为上述的方法输入返回的都是单一值, 所以直接注册为标量函数即可(如果是聚合函数可以使用AggregateFunction)

   // 指定函数名称 和 对应函数的class & method name
   rootSchema.add("dict_sex", ScalarFunctionImpl.create(Udf.class, "dictSex"));
   

3. 测试执行

   final ResultSet resultSet = statement.executeQuery("SELECT username, dict_sex(sex) sex_name FROM `user`");
   printResultSet(resultSet);
   

   表数据如下

   

   输出结果

   c.l.c.CalciteFuncTest - [printResultSet,86] - Number of columns: 2
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=男, username=张三}
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=女, username=李四}
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=女, username=张铁牛}
   

3. 自定义db函数

首先 我们定义一个db 函数实现字典值的转换

DELIMITER //

CREATE FUNCTION dict_sex(code VARCHAR(10))
RETURNS VARCHAR(10)
DETERMINISTIC
BEGIN
    -- 如果code为空或只包含空白字符，则直接返回code
    IF code IS NULL OR TRIM(code) = '' THEN
        RETURN code;
    END IF;
    -- 如果code为'1'则返回'男'
    IF code = '1' THEN
        RETURN '男';
    -- 如果code为'2'则返回'女'
    ELSEIF code = '2' THEN
        RETURN '女';
    ELSE
        RETURN '未知';
    END IF;
END //

DELIMITER ;

验证函数功能

ok, 函数创建完成, 我们将函数注册到calcite中

calcite中sqlfunction有很多其已经实现的类, 我们这里使用SqlBasicFunction来创建我们的函数

1. 定义SqlFunction

   /*
    * SqlBasicFunction create(String name, SqlReturnTypeInference returnTypeInference, SqlOperandTypeChecker operandTypeChecker)
    * name: 函数名称
    * returnTypeInference: 返回值类型
    * operandTypeChecker: 函数入参的校验器
    */
   SqlFunction DICT_SEX = SqlBasicFunction.create("dict_sex", ReturnTypes.VARCHAR, OperandTypes.family(SqlTypeFamily.CHARACTER));
   

2. 注册SqlFunction

   从上篇博文中我们知道, calcite的sql函数都注册到了SqlStdOperatorTable类中, 所以我们只需要将自定义的函数注册进即可

   final SqlStdOperatorTable sqlStdOperatorTable = SqlStdOperatorTable.instance();
   sqlStdOperatorTable.register(DICT_SEX);
   

   对, 就这么简单. 因为SqlStdOperatorTable类是单例模式, 所以我们可以随时随地的进行注册, 其验证逻辑就可以直接调用了

   当然, 看了其他博客大多数都是继承SqlStdOperatorTable类实现自定义SqlStdOperatorTable的 如下, 最后使用自己的SqlStdOperatorTable即可

   public static class SqlCustomOperatorTable extends SqlStdOperatorTable {
       private static SqlCustomOperatorTable instance;
       // 只需要申明为成员变量即可, instance.init() 的时候会反射取变量进行注册
       public static final SqlFunction DICT_SEX = SqlBasicFunction.create("dict_sex", ReturnTypes.VARCHAR, OperandTypes.family(SqlTypeFamily.CHARACTER));
   
       public static synchronized SqlCustomOperatorTable instance() {
           if (instance == null) {
               instance = new SqlCustomOperatorTable();
               instance.init();
           }
   
           return instance;
       }
   
       /**
        * 如果想修改获取函数的过程, 可以重写此方法
        */
       @Override
       protected void lookUpOperators(String name, boolean caseSensitive, Consumer<SqlOperator> consumer) {
           super.lookUpOperators(name, caseSensitive, consumer);
       }
   }
   

3. 测试执行

   final ResultSet resultSet = statement.executeQuery("SELECT username, dict_sex(sex) sex_name FROM `user`");
   printResultSet(resultSet);
   

   输出结果

   c.l.c.CalciteFuncTest - [printResultSet,86] - Number of columns: 2
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=男, username=张三}
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=女, username=李四}
   c.l.c.CalciteFuncTest - [printResultSet,98] - {sex_name=女, username=张铁牛}
   

   经测试: 如果udf 和 sqlfunction 同时存在的时候 优先使用udf

4. 完整代码

4.1 udf

package com.ldx.calcite;

import com.google.common.collect.Maps;
import com.mysql.cj.jdbc.MysqlDataSource;
import lombok.SneakyThrows;
import lombok.extern.slf4j.Slf4j;
import org.apache.calcite.adapter.jdbc.JdbcSchema;
import org.apache.calcite.config.Lex;
import org.apache.calcite.jdbc.CalciteConnection;
import org.apache.calcite.schema.SchemaPlus;
import org.apache.calcite.schema.impl.ScalarFunctionImpl;
import org.junit.jupiter.api.BeforeAll;
import org.junit.jupiter.api.Test;

import javax.sql.DataSource;
import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.ResultSet;
import java.sql.ResultSetMetaData;
import java.sql.SQLException;
import java.sql.Statement;
import java.util.Map;
import java.util.Properties;

import static org.apache.calcite.config.CalciteConnectionProperty.LEX;

@Slf4j
public class CalciteFuncWithUdfTest {
    private static Statement statement;

    @BeforeAll
    @SneakyThrows
    public static void beforeAll() {
        Properties info = new Properties();
        // 不区分sql大小写
        info.setProperty("caseSensitive", "false");
        info.setProperty(LEX.camelName(), Lex.MYSQL.name());
        // 创建Calcite连接
        Connection connection = DriverManager.getConnection("jdbc:calcite:", info);
        CalciteConnection calciteConnection = connection.unwrap(CalciteConnection.class);
        // 构建RootSchema，在Calcite中，RootSchema是所有数据源schema的parent，多个不同数据源schema可以挂在同一个RootSchema下
        SchemaPlus rootSchema = calciteConnection.getRootSchema();
        // 设置默认的schema, 如果不设置sql中需要加上对应数据源的名称
        calciteConnection.setSchema("my_mysql");
        final DataSource mysqlDataSource = getMysqlDataSource();
        final JdbcSchema schemaWithMysql = JdbcSchema.create(rootSchema, "my_mysql", mysqlDataSource, "test", null);
        final SchemaPlus myMysqlSchema = rootSchema.add("my_mysql", schemaWithMysql);
        // 全局注册
        rootSchema.add("dict_sex", ScalarFunctionImpl.create(Udf.class, "dictSex"));
        statement = calciteConnection.createStatement();
        // 只注册到mysql schema中
        // myMysqlSchema.add("dict_sex", ScalarFunctionImpl.create(Udf.class, "dictSex"));
        // 创建SQL语句执行查询
        statement = calciteConnection.createStatement();
    }

    @Test
    @SneakyThrows
    public void test_udf_func() {
        final ResultSet resultSet = statement.executeQuery("SELECT username, dict_sex(sex) sex_name FROM `user`");
        printResultSet(resultSet);
    }

    private static DataSource getMysqlDataSource() {
        MysqlDataSource dataSource = new MysqlDataSource();
        dataSource.setUrl("jdbc:mysql://localhost:3306/test");
        dataSource.setUser("root");
        dataSource.setPassword("123456");

        return dataSource;
    }

    public static void printResultSet(ResultSet resultSet) throws SQLException {
        // 获取 ResultSet 元数据
        ResultSetMetaData metaData = resultSet.getMetaData();
        // 获取列数
        int columnCount = metaData.getColumnCount();
        log.info("Number of columns: {}",columnCount);

        // 遍历 ResultSet 并打印结果
        while (resultSet.next()) {
            final Map<String, String> item = Maps.newHashMap();
            // 遍历每一列并打印
            for (int i = 1; i <= columnCount; i++) {
                String columnName = metaData.getColumnName(i);
                String columnValue = resultSet.getString(i);
                item.put(columnName, columnValue);
            }

            log.info(item.toString());
        }
    }
}

4.2 db func

package com.ldx.calcite;

import com.google.common.collect.Maps;
import com.mysql.cj.jdbc.MysqlDataSource;
import lombok.SneakyThrows;
import lombok.extern.slf4j.Slf4j;
import org.apache.calcite.adapter.jdbc.JdbcSchema;
import org.apache.calcite.config.Lex;
import org.apache.calcite.jdbc.CalciteConnection;
import org.apache.calcite.schema.SchemaPlus;
import org.apache.calcite.sql.SqlBasicFunction;
import org.apache.calcite.sql.SqlFunction;
import org.apache.calcite.sql.SqlOperator;
import org.apache.calcite.sql.fun.SqlStdOperatorTable;
import org.apache.calcite.sql.type.OperandTypes;
import org.apache.calcite.sql.type.ReturnTypes;
import org.apache.calcite.sql.type.SqlTypeFamily;
import org.junit.jupiter.api.BeforeAll;
import org.junit.jupiter.api.Test;

import javax.sql.DataSource;
import java.sql.Connection;
import java.sql.DriverManager;
import java.sql.ResultSet;
import java.sql.ResultSetMetaData;
import java.sql.SQLException;
import java.sql.Statement;
import java.util.Map;
import java.util.Properties;
import java.util.function.Consumer;

import static org.apache.calcite.config.CalciteConnectionProperty.LEX;

@Slf4j
public class CalciteFuncWithDbTest {
    private static Statement statement;

    public static final SqlFunction DICT_SEX = SqlBasicFunction.create("dict_sex", ReturnTypes.VARCHAR, OperandTypes.family(SqlTypeFamily.CHARACTER));

    @BeforeAll
    @SneakyThrows
    public static void beforeAll() {
        Properties info = new Properties();
        // 不区分sql大小写
        info.setProperty("caseSensitive", "false");
        info.setProperty(LEX.camelName(), Lex.MYSQL.name());
        // 创建Calcite连接
        Connection connection = DriverManager.getConnection("jdbc:calcite:", info);
        CalciteConnection calciteConnection = connection.unwrap(CalciteConnection.class);
        // 构建RootSchema，在Calcite中，RootSchema是所有数据源schema的parent，多个不同数据源schema可以挂在同一个RootSchema下
        SchemaPlus rootSchema = calciteConnection.getRootSchema();
        // 设置默认的schema, 如果不设置sql中需要加上对应数据源的名称
        calciteConnection.setSchema("my_mysql");
        final DataSource mysqlDataSource = getMysqlDataSource();
        final JdbcSchema schemaWithMysql = JdbcSchema.create(rootSchema, "my_mysql", mysqlDataSource, "test", null);
        rootSchema.add("my_mysql", schemaWithMysql);
        final SqlStdOperatorTable sqlStdOperatorTable = SqlStdOperatorTable.instance();
        sqlStdOperatorTable.register(DICT_SEX);
        statement = calciteConnection.createStatement();
    }

    @Test
    @SneakyThrows
    public void test_db_func() {
        final ResultSet resultSet = statement.executeQuery("SELECT dict_sex(sex) sex_name FROM `user`");
        printResultSet(resultSet);
    }

    private static DataSource getMysqlDataSource() {
        MysqlDataSource dataSource = new MysqlDataSource();
        dataSource.setUrl("jdbc:mysql://localhost:3306/test");
        dataSource.setUser("root");
        dataSource.setPassword("123456");

        return dataSource;
    }

    public static void printResultSet(ResultSet resultSet) throws SQLException {
        // 获取 ResultSet 元数据
        ResultSetMetaData metaData = resultSet.getMetaData();
        // 获取列数
        int columnCount = metaData.getColumnCount();
        log.info("Number of columns: {}",columnCount);

        while (resultSet.next()) {
            final Map<String, String> item = Maps.newHashMap();
            // 遍历每一列并打印
            for (int i = 1; i <= columnCount; i++) {
                String columnName = metaData.getColumnName(i);
                String columnValue = resultSet.getString(i);
                item.put(columnName, columnValue);
            }

            log.info(item.toString());
        }
    }

    public static class SqlCustomOperatorTable extends SqlStdOperatorTable {
        private static SqlCustomOperatorTable instance;
        // 只需要申明为成员变量即可, instance.init() 的时候会反射取变量进行注册
        public static final SqlFunction DICT_SEX = SqlBasicFunction.create("dict_sex", ReturnTypes.VARCHAR, OperandTypes.family(SqlTypeFamily.CHARACTER));

        public static synchronized SqlCustomOperatorTable instance() {
            if (instance == null) {
                instance = new SqlCustomOperatorTable();
                instance.init();
            }

            return instance;
        }

        /**
         * 如果想修改获取函数的过程, 可以重写此方法
         */
        @Override
        protected void lookUpOperators(String name, boolean caseSensitive, Consumer<SqlOperator> consumer) {
            super.lookUpOperators(name, caseSensitive, consumer);
        }
    }

}