MySQL 8.0 中统计信息直方图的尝试

create table test
(
    id int auto_increment primary key,
    name varchar(100),
    create_date datetime ,
    index (create_date desc)
);


USE `db01`$$

DROP PROCEDURE IF EXISTS `insert_test_data`$$

CREATE DEFINER=`root`@`%` PROCEDURE `insert_test_data`()
BEGIN
    DECLARE v_loop INT;
    SET v_loop = 100000;
    WHILE v_loop>0 DO
        INSERT INTO test(NAME,create_date)VALUES (UUID(),DATE_ADD(NOW(),INTERVAL -RAND()*100000 MINUTE) );
        SET v_loop = v_loop - 1;
    END WHILE;
END$$

DELIMITER ;

DELIMITER $$

USE `db01`$$

DROP PROCEDURE IF EXISTS `parse_column_statistics`$$

CREATE DEFINER=`root`@`%` PROCEDURE `parse_column_statistics`(
    IN `p_schema_name` VARCHAR(200),
    IN `p_table_name` VARCHAR(200),
    IN `p_column_name` VARCHAR(200)
)
BEGIN
    
    DECLARE v_histogram TEXT;
    -- get the special HISTOGRAM
    SELECT HISTOGRAM->>‘$."buckets"‘ INTO v_HISTOGRAM 
    FROM   information_schema.column_statistics
    WHERE schema_name =  p_schema_name 
    AND table_name = p_table_name 
    AND column_name = p_column_name; 
    
    -- remove the first and last [ and ] char
    SET v_histogram = SUBSTRING(v_HISTOGRAM,2,LENGTH(v_HISTOGRAM)-2);

    DROP TABLE IF EXISTS t_buckets ;
    CREATE TEMPORARY TABLE t_buckets
    (
        id INT AUTO_INCREMENT PRIMARY KEY,
        buckets_content VARCHAR(500)
    );
    
    -- split by "]," and get single bucket content    
    WHILE (INSTR(v_histogram,‘],‘)>0) DO
        INSERT INTO t_buckets(buckets_content)
        SELECT SUBSTRING(v_histogram,1,INSTR(v_histogram,‘],‘));
        SET v_HISTOGRAM = SUBSTRING(v_histogram,INSTR(v_histogram,‘],‘)+2,LENGTH(v_histogram));    
    END WHILE;
　　 
    INSERT INTO t_buckets(buckets_content) 
    SELECT v_histogram;
    
    -- get the basic statistics data
    WITH cte AS
    (
        SELECT 
        HISTOGRAM->>‘$."last-updated"‘ AS last_updated,
        HISTOGRAM->>‘$."number-of-buckets-specified"‘ AS number_of_buckets_specified
        FROM INFORMATION_SCHEMA.COLUMN_STATISTICS
        WHERE schema_name =  p_schema_name 
        AND table_name = p_table_name 
        AND column_name = p_column_name
    )
    SELECT 
        CASE WHEN id = 1 THEN p_schema_name ELSE ‘‘ END AS schema_name,
        CASE WHEN id = 1 THEN p_table_name ELSE ‘‘ END AS table_name,
        CASE WHEN id = 1 THEN p_column_name ELSE ‘‘ END AS column_name, 
        CASE WHEN id = 1 THEN last_updated ELSE ‘‘ END AS last_updated,
        CASE WHEN id = 1 THEN number_of_buckets_specified ELSE ‘‘ END AS ‘number_of_buckets_specified‘ ,
        id AS buckets_specified_index,
        buckets_content
    FROM
    (
        SELECT * FROM cte,t_buckets
    )t;

END$$

DELIMITER ;