小编典典

将具有相同ID的行合并为一行,以保留所有数据

sql

我有一张表,上面有一些植物的记录。植物可以具有多个名称,该表将其显示为不同的记录。该表称为new_plantsname

plantid name
1       tree
1       rose
2       bush
3       tree
3       bush
3       rose

这继续超过3000条记录

我想要的是将具有相同Plantid的记录组合在一起,并在不同的列中显示不同的名称:

plantid name1 name2 name3 ...
1       tree  rose  NULL
2       shrub NULL  NULL
3       tree  rose  bush

等等

我也想将结果保存到新表中


阅读 121

收藏
2021-04-28

共1个答案

小编典典

基本上,这是一个PIVOT(您未指定RDBMS)我假设MySQL并且它没有PIVOT函数,因此您将需要使用带有CASE语句的聚合函数来复制它。此解决方案rownumber为每行添加一个,以便您可以确定将多少个name值转换为列。

如果您知道name要拥有多少个值,则可以对值进行硬编码:

select plantid,
  max(case when nameRn = 'name1' then name end) Name1,
  max(case when nameRn = 'name2' then name end) Name2,
  max(case when nameRn = 'name3' then name end) Name3
from
(
  select plantid, name,
      concat('name', @num := if(@plantid = `plantid`, @num + 1, 1)) as nameRn,
      @plantid := `plantid` as dummy
  from
  (
    select plantid, name, @rn:=@rn+1 overall_row_num
    from yourtable, (SELECT @rn:=0) r
  ) x
  order by plantid, overall_row_num
) src
group by plantid;

参见带有演示的SQL Fiddle

如果值的数量未知,则可以使用准备好的语句来生成此值的动态版本:

SET @sql = NULL;
SELECT
  GROUP_CONCAT(DISTINCT
    CONCAT(
      'max(case when nameRn = ''',
      nameRn,
      ''' then name end) AS ',
      nameRn
    )
  ) INTO @sql
FROM 
(
  select plantid, name,
      concat('name', @num := if(@plantid = `plantid`, @num + 1, 1)) as nameRn,
      @plantid := `plantid` as dummy
  from
  (
    select plantid, name, @rn:=@rn+1 overall_row_num
    from yourtable, (SELECT @rn:=0) r
  ) x
  order by plantid, overall_row_num
) src;


SET @sql = CONCAT('SELECT plantid, ', @sql, ' 
                  FROM 
                  (
                    select plantid, name,
                        concat(''name'', @num := if(@plantid = `plantid`, @num + 1, 1)) as nameRn,
                        @plantid := `plantid` as dummy
                    from
                    (
                      select plantid, name, @rn:=@rn+1 overall_row_num
                      from yourtable, (SELECT @rn:=0) r
                    ) x
                    order by plantid, overall_row_num
                  ) src
                   GROUP BY plantid');

PREPARE stmt FROM @sql;
EXECUTE stmt;
DEALLOCATE PREPARE stmt;

参见带有演示的SQL Fiddle

两者都会产生相同的结果:

| PLANTID | NAME1 |  NAME2 |  NAME3 |
-------------------------------------
|       1 |  tree |   rose | (null) |
|       2 |  bush | (null) | (null) |
|       3 |  tree |   bush |   rose |
2021-04-28