Question

我有几位顾问记录他们的工作开始/结束时间。因为他们有两小时的最低费用，我需要确保我正确计算他们的工作时间。当他们在同一站点上有连续的工作时，问题就出现了，工作之间的空闲时间在15到60分钟之间。

以下是按cid，job_date，start_time排序的示例表的查询结果。对于此特定示例，所有作业时间条目都在同一天：

 id  job_num    cid job_date    start_time           end_time
 19  130513-20  6   05/13/2013  2013-05-13 13:30:00  2013-05-13 16:30:00
 10  130513-10  7   05/13/2013  2013-05-13 11:00:00  2013-05-13 13:00:00
 16  130513-17  13  05/13/2013  2013-05-13 14:45:00  2013-05-13 15:30:00
 15  130513-16  16  05/13/2013  2013-05-13 14:45:00  2013-05-13 16:15:00
 18  130513-19  34  05/13/2013  2013-05-13 11:15:00  2013-05-13 12:15:00
 8   130513-08  44  05/13/2013  2013-05-13 12:00:00  2013-05-13 14:30:00
 6   130513-06  47  05/13/2013  2013-05-13 11:00:00  2013-05-13 12:00:00
 9   130513-09  47  05/13/2013  2013-05-13 12:00:00  2013-05-13 12:30:00
 14  130513-15  47  05/13/2013  2013-05-13 14:30:00  2013-05-13 15:15:00
 3   130513-03  50  05/13/2013  2013-05-13 09:00:00  2013-05-13 10:45:00
 7   130513-07  50  05/13/2013  2013-05-13 11:00:00  2013-05-13 12:00:00
 11  130513-11  50  05/13/2013  2013-05-13 13:00:00  2013-05-13 14:30:00
 17  130513-18  50  05/13/2013  2013-05-13 15:00:00  2013-05-13 16:15:00
 2   130513-02  52  05/13/2013  2013-05-13 09:30:00  2013-05-13 10:30:00
 4   130513-04  52  05/13/2013  2013-05-13 10:30:00  2013-05-13 11:00:00
 5   130513-05  52  05/13/2013  2013-05-13 11:00:00  2013-05-13 13:00:00
 12  130513-12  52  05/13/2013  2013-05-13 14:00:00  2013-05-13 14:45:00
 13  130513-13  52  05/13/2013  2013-05-13 15:00:00  2013-05-13 17:00:00
 20  130513-21  61  05/13/2013  2013-05-13 11:00:00  2013-05-13 12:00:00
 1   130513-01  70  05/13/2013  2013-05-13 14:30:00  2013-05-13 15:15:00

我想要完成的是更新特定行（job_num）的end_time，使其等于查询中同一顾问（cid）的下一个job_num的end_time，但仅限于差异（空闲时间）在第一个作业的end_time和后续作业的start_time之间<= 30分钟。基本上，我想将空闲时间添加到比较中第一个作业的end_time。

例如：

 id  job_num    cid job_date    start                end                
 3   130513-03  50  05/13/2013  2013-05-13 09:00:00  2013-05-13 10:45:00  <-- original
 7   130513-07  50  05/13/2013  2013-05-13 11:00:00  2013-05-13 12:00:00 
 11  130513-11  50  05/13/2013  2013-05-13 13:00:00  2013-05-13 14:30:00  <-- original
 17  130513-18  50  05/13/2013  2013-05-13 15:00:00  2013-05-13 16:15:00 
...
 12  130513-12  52  05/13/2013  2013-05-13 14:00:00  2013-05-13 14:45:00  <-- original
 13  130513-13  52  05/13/2013  2013-05-13 15:00:00  2013-05-13 17:00:00

成为：

 3   130513-03  50  05/13/2013  2013-05-13 09:00:00  2013-05-13 11:00:00  <-- changed
 7   130513-07  50  05/13/2013  2013-05-13 11:00:00  2013-05-13 12:00:00 
 11  130513-11  50  05/13/2013  2013-05-13 13:00:00  2013-05-13 15:00:00  <-- changed
 17  130513-18  50  05/13/2013  2013-05-13 15:00:00  2013-05-13 16:15:00 
...
 12  130513-12  52  05/13/2013  2013-05-13 14:00:00  2013-05-13 15:00:00  <-- changed
 13  130513-13  52  05/13/2013  2013-05-13 15:00:00  2013-05-13 17:00:00

我将如何进行此更改？我已经看到了使用临时表和变量的类似问题的解决方案，这些表和变量非常适合检查时间差异但是没有一个可以根据我的特定问题进行调整（在我对MySQL查询的有限知识中）。我还将这个问题标记为PHP，因为在PHP中执行此操作可能会更简单。

任何帮助将不胜感激！

Answer 1

我认为id是表格的Primary Key，而且是AUTO-INCREMENT。

我有这个UPDATE查询：

UPDATE table T1
    SET T1.end_time = (
            SELECT T2.end_time
                FROM table T2
                WHERE T2.cid = T1.cid
                    AND T2.job_date = T1.job_date
                    AND T2.id > T1.id
                    AND T2.start_time <= DATE_ADD(T1.end_time, INTERVAL 30 MINUTE)
                ORDER BY T2.id ASC
                LIMIT 1
        )
    WHERE id IN (
        SELECT DISTINCT T1.Id
            FROM table T1
            INNER JOIN table T2 ON T2.cid = T1.cid
                AND T2.job_date = T1.job_date
                AND T2.id > T1.id
                AND T2.start_time <= DATE_ADD(T1.end_time, INTERVAL 30 MINUTE)
            WHERE T1.job_date = @date
        );

虽然有点复杂，但并不那么有效。但由于这只是一个查询，我认为您需要每次更新end_time（每日/每周/等）。

并确保设置@date变量的值，或者您可以将其替换为所需的日期。

我认为使用CURSOR（这很慢）会比我上面的查询更有效。

以下是CURSOR查询示例：

DECLARE id INT(11);
DECLARE newEndDate DATETIME;

DECLARE noMoreRows BIT(1);
DECLARE customCursor CURSOR FOR
    SELECT T1.Id
        , CAST(SUBSTRING_INDEX(GROUP_CONCAT(T2.start_date ORDER BY T2.Id ASC), ',', 1) AS DATETIME) AS NewEndDate
        FROM table T1
        INNER JOIN table T2 ON T2.cid = T1.cid
             AND T2.job_date = T1.job_date
             AND T2.id > T1.id
             AND T2.start_time <= DATE_ADD(T1.end_time, INTERVAL 30 MINUTE)
        WHERE T1.job_date = @date
        GROUP BY T1.Id
        HAVING NewEndDate <> '' OR NewEndDate IS NOT NULL;
DECLARE CONTINUE HANDLER FOR NOT FOUND SET noMoreRows = 1;

OPEN customCursor;
customLoop: LOOP

    FETCH customCursor INTO id, newEndDate;

    IF noMoreRows THEN
      LEAVE customLoop;
   END IF;

    UPDATE table
        SET end_date = newEndDate
        WHERE id = id;

END LOOP customLoop;
CLOSE customCursor;

请注意，CURSOR应位于STORED PROCEDURE内。以下是CURSORS的{{3}}。

Answer 2

我假设您使用php格式化开始时间和结束时间，然后插入数据库表。

$start = date("Y-m-d H:i:s",time());
$query = "select * from table name where cid = $cid and  UNIX_TIMESTAMP($start) - UNIX_TIMESTAMP(end_time) <= 1800";
mysqli_query($con,$query);

if ($result=mysqli_query($con,$query))
  {
    $rowcount=mysqli_num_rows($result);

    if($rowcount > 1)
    { 
     $row = mysqli_fetch_array($result, MYSQLI_ASSOC);

     query2 = "update tablename set end_time = '$start' where id = '$row['id']'";

     query3 = "insert into tablename(cid,date,start_time,end_time) values($cid,$date,$start,$end)";


    }else{

    //do the insert as such
    query2 = "insert into tablename(cid,date,start_time,end_time) values($cid,$date,$start,$end)";

    }

    mysqli_free_result($result);
  }

MySQL - 更改一行的时间值以匹配同一表的另一行的时间值

2 个答案: