Question

我正在使用SQL Server来管理网站的数据。我刚刚决定向数据库添加两个新列，我想为表中的每一行填充这两列。要填充每个列，我使用另一个表中的信息。我已经创建了一个在C＃中为我做的方法，但它需要很长时间。它大约35,000行。这是我的C＃方法所做的，但我想知道将它作为存储过程是否更好，如果它更好，我怎么能去做呢？

    public void UpdateAssignedCountAndLastReponseDate()
    {
        //Loop through each task in the database 
        foreach (var task in IssueDb.AP_Tasks.ToList())
        {
            // Check to see if there is both not a null and a last response date
            if (task.AssignedCount != null && task.LastReponseDate != null) continue;
            //Check the task history for records
            var taskhistory = from x in IssueDb.AP_TaskHistory
                              where x.TaskID == task.TaskID && x.AssignDate != null
                              select x;
            //get last response date
            var maxAssignDate = (from th in IssueDb.AP_TaskHistory where task.TaskID == th.TaskID orderby th.AssignDate descending select th.AssignDate).FirstOrDefault();
            IssueDb.Entry(task).State = EntityState.Modified;
            task.LastReponseDate = maxAssignDate;
            //get count
            task.AssignedCount = taskhistory.Count();
            if (task.LastReponseDate == null)
                task.LastReponseDate = task.TaskDate;
            if (task.AssignedCount == null)
                task.AssignedCount = 0;
            //save these numbers to task columns in the DB
            IssueDb.SaveChanges();
        }
    }

我目前在ASP的EF5中这样做。 NET应用程序MVC

这是我到目前为止所做的：

   UPDATE [dbo].[AP_Tasks]
   SET [AssignedCount] =
   (
        CASE [AssignedCount] IS NULL THEN (SELECT *, COUNT(*) FROM [dbo].[AP_TaskHistory] INNER JOIN [AP_TaskHistory] WHERE [AP_TaskHistory].[TaskID] == [AP_Tasks].[TaskID] AND [AP_TaskHistory].[AssignDate] IS NOT NULL)

   );
   SET [LastReponseDate] =
   (
        CASE [AP_Tasks].[LastReponseDate] IS NULL
        THEN (SELECT FIRST(AssignDate) FROM [dbo].[AP_TaskHistory] INNER JOIN [AP_TaskHistory] WHERE [AP_TaskHistory].[TaskID] == [AP_Tasks].[TaskID] AND [AP_TaskHistory].[AssignDate] IS NOT NULL ORDER BY [AP_TaskHistory].[AssignDate] DESC)
   );

以下是我的错误：

Msg 156, Level 15, State 1, Line 5
Incorrect syntax near the keyword 'IS'.
Msg 156, Level 15, State 1, Line 5
Incorrect syntax near the keyword 'WHERE'.
Msg 195, Level 15, State 10, Line 11
'FIRST' is not a recognized built-in function name.

Answer 1

在SQL Server中，您可以执行连接其他表的更新语句。这是一个例子

UPDATE u
SET u.FirstName = t2.SomeColumn
FROM Users u
     INNER JOIN Table2 t2 ON (u.UserID = t2.UserID)

由于您使用的是EF，因此您可以将其放入迁移脚本中。

这种方法会快得多！

Answer 2

我拿了你的更新声明并清理了它：

UPDATE  [dbo].[AP_Tasks]
SET     [AssignedCount] = ( SELECT  COUNT(*)
                            FROM    [dbo].[AP_TaskHistory]
                            WHERE  [AP_TaskHistory].[TaskID] = [AP_Tasks].[TaskID] AND  [AP_TaskHistory].[AssignDate] IS NOT NULL
                          ) ,
        [LastReponseDate] = ( SELECT    MAX(AssignDate)
                              FROM      [dbo].[AP_TaskHistory]
                              WHERE  [AP_TaskHistory].[TaskID] = [AP_Tasks].[TaskID] AND  [AP_TaskHistory].[AssignDate] IS NOT NULL
                            )

修正语法
删除了CASE声明，如果这些是新列，则不必担心现有数据

Answer 3

当然，使用存储过程会更快。

您的存储过程的代码看起来与您的C＃代码非常相似，您可以使用while，if / else语句，...

将C＃方法转换为存储过程

3 个答案: