需要像UNION的查询，但在第二的相应列的别名和第n个查询，而不是所述第一查询的列/别名

Question

返回数据考虑此UNION两个（或更多）的查询：

SELECT activity AS activity1, completion AS status1, date_end AS date1 
    FROM activities 
    WHERE (
     matter LIKE '%JR161167' 
     AND activity LIKE '1.1 %' 
     AND DATEDIFF(CURDATE() , date_end) <120 
    ) 
UNION 
    SELECT activity AS activity2, completion AS status2, date_end AS date2 
    FROM activities 
    WHERE (
     matter LIKE '%JR161167' 
     AND activity LIKE '1.2 %' 
     AND DATEDIFF(CURDATE() , date_end) <120 
    ) 

活性列将包含一个可变字符串值，但该字符串将可靠地与1.1或1.2或1.3一起前置至1.9。

问题栏也是一个字符串，但会可靠地追加一个类似于JR161167的帐号。

我想查询已经从内部过去的120天（这就是DATEFIFF部分）DATE_END所有记录，符合指定帐号（JR161167），但在活动场中包含的任何和所有子“1 ñ“。

使用UNION可从单个表“activities”中返回正确的数据，但似乎没有办法在第一个查询的列或别名下标题结果集。相反，我想是这样的：

activity1    status1 date1  activity2 status2   date2 
-------------------------------------------------------------------------------- 
1.1 Step 1 begins... Incomplete 2017-4-16 1.2  Step 2 begins... Incomplete 2017-4-30 

相反，我的查询栈的所有结果，不管是否含有1.1，1.2，1.3，等等，一个别名列标题，activity1下。与日期和完成字段数据相同的东西，它并不会很好地分类到相应的日期1，日期2等或状态1，状态2等。

我最近读了一个看起来像这样的例子，它貌似我要实现的逻辑，我不知道是否有人能理解其意图，并帮助我制定正确的查询：

SELECT date_end 
IF(WHERE description LIKE '1.1 %') AS date1, 
IF(WHERE description LIKE '1.2 %') AS date2` 
FROM activities 

Answer 1

SQL Fiddle

查询1：

SELECT a1.*, a2.* 
FROM (
    SELECT matter, activity, completion, date_end 
    FROM activities 
    WHERE 
     matter LIKE '%JR16116' 
     AND activity = '1.1' 
     AND DATEDIFF(CURDATE(), date_end) <120 
) a1 
JOIN (
    SELECT matter, activity, completion, date_end 
    FROM activities 
    WHERE 
     matter LIKE '%JR16116' 
     AND activity = '1.2' 
     AND DATEDIFF(CURDATE(), date_end) <120 
) a2 
    ON a1.matter = a2.matter 

Results：

| matter | activity | completion | date_end | matter | activity | completion | date_end | 
|---------|----------|------------|------------|---------|----------|------------|------------| 
| JR16116 |  1.1 |   1 | 2017-04-22 | JR16116 |  1.2 |   1 | 2017-04-23 | 

查询2：

-- other way without sub request but clean join 

SELECT 
    a.matter, a.activity, a.completion, a.date_end 
    ,a2.activity, a2.completion, a2.date_end 
    -- ,a3.activity, a3.completion, a3.date_end 
FROM activities a 
JOIN activities a2 
    ON (
     a.matter = a2.matter 
     AND a2.activity = '1.2' 
     AND DATEDIFF(CURDATE(), a2.date_end) <120 
    ) 
-- JOIN activities a3 
-- ON (
--  a.matter = a3.matter 
--  AND a3.activity = '1.3' 
--  AND DATEDIFF(CURDATE(), a3.date_end) <120 
-- ) 

WHERE 
    a.matter LIKE '%JR16116' 
    AND a.activity = '1.1' 
    AND DATEDIFF(CURDATE(), a.date_end) <120 

Results：

| matter | activity | completion | date_end | activity | completion | date_end | 
|---------|----------|------------|------------|----------|------------|------------| 
| JR16116 |  1.1 |   1 | 2017-04-22 |  1.2 |   1 | 2017-04-23 |