我想允许我的用户为他们的活动设置时间表。它可能是一天，或者为方便起见，我想让他们指定一个重复发生的事件（类似于Outlook约会）。
对于单个事件，它似乎很容易（伪代码）：
只要有一个DateOfEvent列，其上有日期。
抓住未来事件：
从DateOfEvent＆gt;的事件中选择* {DateTime.Now}
但是我如何存储和查询重复出现的事件？我不需要做任何时间，因为我只是单独存储，如果他们需要不同的时间，我只是让他们创建另一个事件。所以没有：每个星期三的5点和星期四的3点。
示例：
每个星期一，星期二，星期五，星期四，星期五，每周
每周结婚
每个月的第二个星期二
我以前查询的内容
我添加了开始和结束日期。如果用户选择了一个日期，我将开始日期和结束日期都设置为所选日期。我只需稍微修改答案的代码。
DECLARE
@StartDate SMALLDATETIME,
@EndDate SMALLDATETIME;
SELECT
@StartDate = '20091129',
@EndDate = '20101220';
SELECT
d.CurrentDate,
m.*
FROM
Calendar AS d
INNER JOIN Meet AS m
ON
(
(d.CurrentDate = m.StartDate AND d.CurrentDate = m.EndDate)
OR d.DaysOfTheMonth = m.DayOfTheMonth
OR (d.DaysOfTheWeek = m.DayOfTheWeek AND COALESCE(m.WeekOfTheMonth, d.WeekOfTheMonth) = d.WeekOfTheMonth)
OR d.DaysOfTheWeek IN (1,7) AND m.OnWeekends = 1
OR d.DaysOfTheWeek BETWEEN 2 AND 6 AND m.OnWeekDays = 1
)
-- now you can inner join to the event table
-- to list the name, other details, etc.
WHERE
d.CurrentDate BETWEEN @StartDate AND @EndDate
AND d.CurrentDate BETWEEN m.StartDate AND m.EndDate
ORDER BY d.CurrentDate;
GO
填写Calandar表：
INSERT dbo.Calendar(CurrentDate, DaysOfTheMonth, DaysOfTheWeek, WeekOfTheMonth, IsWeekDay, Even)
SELECT
CurrentDate, DaysOfTheMonth, DaysOfTheWeek,
WeekOfTheMonth = DATEDIFF(WEEK, DATEADD(DAY, 1-DaysOfTheMonth, CurrentDate), CurrentDate) + 1,
CASE WHEN DaysOfTheWeek IN (1, 7) THEN 0 ELSE 1 END AS IsWeekDay,
CASE WHEN w % 2 = 1 THEN 0 ELSE 1 END AS Even
FROM
(
SELECT
CurrentDate,
DaysOfTheMonth = DATEPART(DAY, CurrentDate),
DaysOfTheWeek = DATEPART(WEEKDAY, CurrentDate),
w = DATEPART(WEEK, CurrentDate)
FROM
(
SELECT CurrentDate = DATEADD(DAY, n-1, '20141001')
FROM
(
SELECT TOP 900
n = ROW_NUMBER() OVER (ORDER BY c1.[object_id])
FROM sys.all_columns AS c1
CROSS JOIN sys.all_columns AS c2
ORDER BY n
) AS x) AS y) AS z;
答案 0 :(得分：4)
你可以这样做：
CREATE TABLE dbo.EventSchedule
(
EventID INT, -- FOREIGN KEY to event details
EventDate SMALLDATETIME, -- if one-time event
DayOfMonth TINYINT, -- if once a month, e.g. 3rd of every month
DayOfWeek TINYINT, -- if once a week, e.g. every Tuesday = 3
WeekDays BIT, -- if only on weekdays, e.g. 1 = mon-fri
Weekends BIT, -- if only on weekends, e.g. 1 = sat-sun
-- the next two are combined, e.g. 2/2 = 2nd Monday of each month
MonthlyInstance TINYINT,
MonthlyWeekday TINYINT
);
那么如果你想找到计划在给定日期发生的所有事件，你可以这样做：
DECLARE
@dt SMALLDATETIME,
@dm TINYINT,
@dw TINYINT,
@inst TINYINT;
SELECT
@dt = '20091201',
@dm = DATEPART(DAY, @dt)
@dw = DATEPART(WEEKDAY, @dt),
@inst = DATEDIFF(WEEK, DATEADD(DAY, 1-@dm, @dt), @dt) + 1;
SELECT EventID
FROM dbo.EventSchedule
WHERE EventDate = @dt
OR DayOfMonth = @dm
OR DayOfWeek = @dw
OR (Weekdays = 1 AND @dw BETWEEN 2 AND 6)
OR (Weekends = 1 AND @dw IN (1,7))
OR (MonthlyInstance = @inst AND MonthlyWeekday = @dw);
这是未经测试的（毕竟我在感恩节时从Mac发布），并依赖于SET DATEFIRST作为默认值（其中Sunday = 1，Monday = 2等）。我对第二个星期二的部分也没有100％的自信，因为它可能需要一些额外的数学，这取决于当月第一天的工作日。但是我认为这可以为你提供其他部分的开始，当我有机会时，我会回来重温一下。
有关更复杂的选项，请参阅：http://www.codeproject.com/KB/database/sqlscheduleselector.aspx
这是一个更完整的示例，包含一个填充的日历表，一些不同类型的示例事件，以及一个查询给定日期范围的所有预期事件的查询。
CREATE DATABASE test;
GO
USE test;
GO
SET NOCOUNT ON;
GO
CREATE TABLE dbo.Calendar
(
dt SMALLDATETIME PRIMARY KEY,
dm TINYINT,
dw TINYINT,
mw TINYINT,
wd BIT
);
GO
-- populate the table with the days from 2009
-- (you will want more obviously; change TOP or add WHERE)
INSERT dbo.Calendar(dt, dm, dw, mw, wd)
SELECT
dt, dm, dw,
mw = DATEDIFF(WEEK, DATEADD(DAY, 1-dm, dt), dt) + 1,
CASE WHEN dw IN (1, 7) THEN 0 ELSE 1 END
FROM
(
SELECT
dt,
dm = DATEPART(DAY, dt),
dw = DATEPART(WEEKDAY, dt)
FROM
(
SELECT dt = DATEADD(DAY, n-1, '20090101')
FROM
(
SELECT TOP 365
n = ROW_NUMBER() OVER (ORDER BY c1.[object_id])
FROM sys.all_columns AS c1
CROSS JOIN sys.all_columns AS c2
ORDER BY n
) AS x) AS y) AS z;
GO
-- create your schedule table:
CREATE TABLE dbo.EventSchedule
(
EventID INT, -- FOREIGN KEY to event details
EventDate SMALLDATETIME, -- if one-time event
[DayOfMonth] TINYINT, -- if once a month, e.g. 3rd of every month
[DayOfWeek] TINYINT, -- if once a week, e.g. every Tuesday = 3
Weekdays BIT, -- if only on weekdays, e.g. 1 = mon-fri
Weekends BIT, -- if only on weekends, e.g. 1 = sat-sun
-- if you want every day, set Weekdays+Weekends = 1
WeekOfMonth TINYINT -- if only the nth Tuesday etc.
);
-- create some events:
INSERT dbo.EventSchedule
(
EventID,
EventDate,
[DayOfMonth],
[DayOfWeek],
Weekdays,
Weekends,
WeekOfMonth
)
-- one on Jan 5th:
SELECT 1, '20090105', NULL, NULL, NULL, NULL, NULL
-- one on the 3rd of every month:
UNION ALL SELECT 2, NULL, 3, NULL, NULL, NULL, NULL
-- one every Tuesday:
UNION ALL SELECT 3, NULL, NULL, 3, NULL, NULL, NULL
-- one the 2nd Wednesday of each month:
UNION ALL SELECT 4, NULL, NULL, 4, NULL, NULL, 2
-- one only on weekends:
UNION ALL SELECT 5, NULL, NULL, NULL, NULL, 1, NULL
-- one only on weekdays:
UNION ALL SELECT 6, NULL, NULL, NULL, 1, NULL, NULL;
-- now, we have 6 events, some of which will happen
-- multiple times in any given date range. So let's
-- try it:
DECLARE
@StartDate SMALLDATETIME,
@EndDate SMALLDATETIME;
SELECT
@StartDate = '20090101',
@EndDate = '20090108';
SELECT
d.dt,
e.EventID
FROM
dbo.Calendar AS d
INNER JOIN dbo.EventSchedule AS e
ON
(
d.dt = e.EventDate
OR d.dm = e.[DayOfMonth]
OR (d.dw = e.[DayOfWeek] AND COALESCE(e.WeekOfMonth, d.mw) = d.mw)
OR d.dw IN (1,7) AND e.Weekends = 1
OR d.dw BETWEEN 2 AND 6 AND e.Weekdays = 1
)
-- now you can inner join to the event table
-- to list the name, other details, etc.
WHERE
d.dt BETWEEN @StartDate AND @EndDate
ORDER BY d.dt, e.EventID;
GO
DROP TABLE dbo.Calendar, dbo.EventSchedule;
GO
USE [master];
GO
DROP DATABASE test;
GO
答案 1 :(得分：3)
要存储时间表，请查看我在这个问题中的答案
What is the best way to represent "Recurring Events" in database?
这里我将介绍SQL Server如何在其sysschedules及相关表中存储计划信息。他们有一个非常好的架构设计，支持您询问的所有调度案例。 MSDN上的文档也非常有用。
要获得展望类型显示的日期范围之间的所有计划日期，请查看Peter Larsson的博客，他创建了一个优秀的计划来计算这些值。
CREATE PROCEDURE dbo.uspGetScheduleTimes
(
@startDate DATETIME,
@endDate DATETIME
)
AS
/*
This code is blogged here
http://weblogs.sqlteam.com/peterl/archive/2008/10/10/Keep-track-of-all-your-jobs-schedules.aspx
*/
SET NOCOUNT ON
-- Create a tally table. If you already have one of your own please use that instead.
CREATE TABLE #tallyNumbers
(
num SMALLINT PRIMARY KEY CLUSTERED
)
DECLARE @index SMALLINT
SET @index = 1
WHILE @index <= 8640
BEGIN
INSERT #tallyNumbers
(
num
)
VALUES (
@index
)
SET @index = @index + 1
END
-- Create a staging table for jobschedules
CREATE TABLE #jobSchedules
(
rowID INT IDENTITY(1, 1) PRIMARY KEY CLUSTERED,
serverName SYSNAME NOT NULL,
jobName SYSNAME NOT NULL,
jobDescription NVARCHAR(512) NOT NULL,
scheduleName SYSNAME NOT NULL,
scheduleID INT NOT NULL,
categoryName SYSNAME NOT NULL,
freq_type INT NOT NULL,
freq_interval INT NOT NULL,
freq_subday_type INT NOT NULL,
freq_subday_interval INT NOT NULL,
freq_relative_interval INT NOT NULL,
freq_recurrence_factor INT NOT NULL,
startDate DATETIME NOT NULL,
startTime DATETIME NOT NULL,
endDate DATETIME NOT NULL,
endTime DATETIME NOT NULL,
jobEnabled INT NOT NULL,
scheduleEnabled INT NOT NULL
)
/*
-- Popoulate the staging table for JobSchedules with SQL Server 2000
INSERT #jobSchedules
(
serverName,
jobName,
jobDescription,
scheduleName,
scheduleID,
categoryName,
freq_type,
freq_interval,
freq_subday_type,
freq_subday_interval,
freq_relative_interval,
freq_recurrence_factor,
startDate,
startTime,
endDate,
endTime,
jobEnabled,
scheduleEnabled
)
SELECT sj.originating_server,
sj.name,
COALESCE(sj.description, ''),
sjs.name,
sjs.schedule_id,
sc.name,
sjs.freq_type,
sjs.freq_interval,
sjs.freq_subday_type,
sjs.freq_subday_interval,
sjs.freq_relative_interval,
sjs.freq_recurrence_factor,
COALESCE(STR(sjs.active_start_date, 8), CONVERT(CHAR(8), GETDATE(), 112)),
STUFF(STUFF(REPLACE(STR(sjs.active_start_time, 6), ' ', '0'), 3, 0, ':'), 6, 0, ':'),
STR(sjs.active_end_date, 8),
STUFF(STUFF(REPLACE(STR(sjs.active_end_time, 6), ' ', '0'), 3, 0, ':'), 6, 0, ':'),
sj.enabled,
sjs.enabled
FROM msdb..sysjobschedules AS sjs
INNER JOIN msdb..sysjobs AS sj ON sj.job_id = sjs.job_id
INNER JOIN msdb..syscategories AS sc ON sc.category_id = sj.category_id
WHERE sjs.freq_type IN (1, 4, 8, 16, 32)
ORDER BY sj.originating_server,
sj.name,
sjs.name
*/
-- Popoulate the staging table for JobSchedules with SQL Server 2005 and SQL Server 2008
INSERT #JobSchedules
(
serverName,
jobName,
jobDescription,
scheduleName,
scheduleID,
categoryName,
freq_type,
freq_interval,
freq_subday_type,
freq_subday_interval,
freq_relative_interval,
freq_recurrence_factor,
startDate,
startTime,
endDate,
endTime,
jobEnabled,
scheduleEnabled
)
SELECT srv.srvname,
sj.name,
COALESCE(sj.description, ''),
ss.name,
ss.schedule_id,
sc.name,
ss.freq_type,
ss.freq_interval,
ss.freq_subday_type,
ss.freq_subday_interval,
ss.freq_relative_interval,
ss.freq_recurrence_factor,
COALESCE(STR(ss.active_start_date, 8), CONVERT(CHAR(8), GETDATE(), 112)),
STUFF(STUFF(REPLACE(STR(ss.active_start_time, 6), ' ', '0'), 3, 0, ':'), 6, 0, ':'),
STR(ss.active_end_date, 8),
STUFF(STUFF(REPLACE(STR(ss.active_end_time, 6), ' ', '0'), 3, 0, ':'), 6, 0, ':'),
sj.enabled,
ss.enabled
FROM msdb..sysschedules AS ss
INNER JOIN msdb..sysjobschedules AS sjs ON sjs.schedule_id = ss.schedule_id
INNER JOIN msdb..sysjobs AS sj ON sj.job_id = sjs.job_id
INNER JOIN sys.sysservers AS srv ON srv.srvid = sj.originating_server_id
INNER JOIN msdb..syscategories AS sc ON sc.category_id = sj.category_id
WHERE ss.freq_type IN(1, 4, 8, 16, 32)
ORDER BY srv.srvname,
sj.name,
ss.name
-- Only deal with jobs that has active start date before @endDate
DELETE
FROM #JobSchedules
WHERE startDate > @endDate
-- Only deal with jobs that has active end date after @startDate
DELETE
FROM #JobSchedules
WHERE endDate < @startDate
-- Deal with first, second, third, fourth and last occurence
DECLARE @tempStart DATETIME,
@tempEnd DATETIME
SELECT @tempStart = DATEADD(MONTH, DATEDIFF(MONTH, '19000101', @startDate), '19000101'),
@TempEnd = DATEADD(MONTH, DATEDIFF(MONTH, '18991231', @endDate), '18991231')
CREATE TABLE #dayInformation
(
infoDate DATETIME PRIMARY KEY CLUSTERED,
weekdayName VARCHAR(9) NOT NULL,
statusCode INT NOT NULL,
lastDay TINYINT DEFAULT 0
)
WHILE @tempStart <= @tempEnd
BEGIN
INSERT #dayInformation
(
infoDate,
weekdayName,
statusCode
)
SELECT @tempStart,
DATENAME(WEEKDAY, @tempStart),
CASE
WHEN DATEPART(DAY, @tempStart) BETWEEN 1 AND 7 THEN 1
WHEN DATEPART(DAY, @tempStart) BETWEEN 8 AND 14 THEN 2
WHEN DATEPART(DAY, @tempStart) BETWEEN 15 AND 21 THEN 4
WHEN DATEPART(DAY, @tempStart) BETWEEN 22 AND 28 THEN 8
ELSE 0
END
SET @tempStart = DATEADD(DAY, 1, @tempStart)
END
UPDATE di
SET di.statusCode = di.statusCode + 16
FROM #dayInformation AS di
INNER JOIN (
SELECT DATEDIFF(MONTH, '19000101', infoDate) AS theMonth,
DATEPART(DAY, MAX(infoDate)) - 6 AS theDay
FROM #dayInformation
GROUP BY DATEDIFF(MONTH, '19000101', infoDate)
) AS x ON x.theMonth = DATEDIFF(MONTH, '19000101', di.infoDate)
WHERE DATEPART(DAY, di.infoDate) >= x.theDay
UPDATE di
SET di.lastDay = 16
FROM #dayInformation AS di
INNER JOIN (
SELECT DATEDIFF(MONTH, '19000101', infoDate) AS theMonth,
MAX(infoDate) AS theDay
FROM #dayInformation
GROUP BY DATEDIFF(MONTH, '19000101', infoDate)
) AS x ON x.theMonth = DATEDIFF(MONTH, '19000101', di.infoDate)
WHERE di.infoDate = x.theDay
UPDATE #dayInformation
SET lastDay = DATEPART(DAY, infoDate)
WHERE DATEPART(DAY, infoDate) BETWEEN 1 AND 4
-- Stage all individual schedule times
CREATE TABLE #scheduleTimes
(
rowID INT NOT NULL,
infoDate DATETIME NOT NULL,
startTime DATETIME NOT NULL,
endTime DATETIME NOT NULL,
waitSeconds INT DEFAULT 0
)
CREATE CLUSTERED INDEX IX_rowID ON #scheduleTimes(rowID)
-- Insert one time only schedules
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime
)
SELECT rowID,
startDate,
startTime,
endTime
FROM #jobSchedules
WHERE freq_type = 1
AND startDate >= @StartDate
AND startDate <= @EndDate
-- Insert daily schedules
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime,
waitSeconds
)
SELECT js.rowID,
di.infoDate,
js.startTime,
js.endTime,
CASE js.freq_subday_type
WHEN 1 THEN 0
WHEN 2 THEN js.freq_subday_interval
WHEN 4 THEN 60 * js.freq_subday_interval
WHEN 8 THEN 3600 * js.freq_subday_interval
END
FROM #jobSchedules AS js
INNER JOIN #dayInformation AS di ON di.infoDate >= @startDate
AND di.infoDate <= @endDate
WHERE js.freq_type = 4
AND DATEDIFF(DAY, js.startDate, di.infoDate) % js.freq_interval = 0
-- Insert weekly schedules
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime,
waitSeconds
)
SELECT js.rowID,
di.infoDate,
js.startTime,
js.endTime,
CASE js.freq_subday_type
WHEN 1 THEN 0
WHEN 2 THEN js.freq_subday_interval
WHEN 4 THEN 60 * js.freq_subday_interval
WHEN 8 THEN 3600 * js.freq_subday_interval
END
FROM #jobSchedules AS js
INNER JOIN #dayInformation AS di ON di.infoDate >= @startDate
AND di.infoDate <= @endDate
WHERE js.freq_type = 8
AND 1 = CASE
WHEN js.freq_interval & 1 = 1 AND di.weekdayName = 'Sunday' THEN 1
WHEN js.freq_interval & 2 = 2 AND di.weekdayName = 'Monday' THEN 1
WHEN js.freq_interval & 4 = 4 AND di.weekdayName = 'Tuesday' THEN 1
WHEN js.freq_interval & 8 = 8 AND di.weekdayName = 'Wednesday' THEN 1
WHEN js.freq_interval & 16 = 16 AND di.weekdayName = 'Thursday' THEN 1
WHEN js.freq_interval & 32 = 32 AND di.weekdayName = 'Friday' THEN 1
WHEN js.freq_interval & 64 = 64 AND di.weekdayName = 'Saturday' THEN 1
ELSE 0
END
AND(DATEDIFF(DAY, js.startDate, di.infoDate) / 7) % js.freq_recurrence_factor = 0
-- Insert monthly schedules
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime,
waitSeconds
)
SELECT js.rowID,
di.infoDate,
js.startTime,
js.endTime,
CASE js.freq_subday_type
WHEN 1 THEN 0
WHEN 2 THEN js.freq_subday_interval
WHEN 4 THEN 60 * js.freq_subday_interval
WHEN 8 THEN 3600 * js.freq_subday_interval
END
FROM #jobSchedules AS js
INNER JOIN #dayInformation AS di ON di.infoDate >= @startDate
AND di.infoDate <= @endDate
WHERE js.freq_type = 16
AND DATEPART(DAY, di.infoDate) = js.freq_interval
AND DATEDIFF(MONTH, js.startDate, di.infoDate) % js.freq_recurrence_factor = 0
-- Insert monthly relative schedules
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime,
waitSeconds
)
SELECT js.rowID,
di.infoDate,
js.startTime,
js.endTime,
CASE js.freq_subday_type
WHEN 1 THEN 0
WHEN 2 THEN js.freq_subday_interval
WHEN 4 THEN 60 * js.freq_subday_interval
WHEN 8 THEN 3600 * js.freq_subday_interval
END
FROM #jobSchedules AS js
INNER JOIN #dayInformation AS di ON di.infoDate >= @startDate
AND di.infoDate <= @endDate
WHERE js.freq_type = 32
AND 1 = CASE
WHEN js.freq_interval = 1 AND di.weekdayName = 'Sunday' THEN 1
WHEN js.freq_interval = 2 AND di.weekdayName = 'Monday' THEN 1
WHEN js.freq_interval = 3 AND di.weekdayName = 'Tuesday' THEN 1
WHEN js.freq_interval = 4 AND di.weekdayName = 'Wednesday' THEN 1
WHEN js.freq_interval = 5 AND di.weekdayName = 'Thursday' THEN 1
WHEN js.freq_interval = 6 AND di.weekdayName = 'Friday' THEN 1
WHEN js.freq_interval = 7 AND di.weekdayName = 'Saturday' THEN 1
WHEN js.freq_interval = 8 AND js.freq_relative_interval = di.lastDay THEN 1
WHEN js.freq_interval = 9 AND di.weekdayName NOT IN('Sunday', 'Saturday') THEN 1
WHEN js.freq_interval = 10 AND di.weekdayName IN('Sunday', 'Saturday') THEN 1
ELSE 0
END
AND di.statusCode & js.freq_relative_interval = js.freq_relative_interval
AND DATEDIFF(MONTH, js.startDate, di.infoDate) % js.freq_recurrence_factor = 0
-- Get the daily recurring schedule times
INSERT #scheduleTimes
(
rowID,
infoDate,
startTime,
endTime,
waitSeconds
)
SELECT st.rowID,
st.infoDate,
DATEADD(SECOND, tn.num * st.waitSeconds, st.startTime),
st.endTime,
st.waitSeconds
FROM #scheduleTimes AS st
CROSS JOIN #tallyNumbers AS tn
WHERE tn.num * st.waitSeconds <= DATEDIFF(SECOND, st.startTime, st.endTime)
AND st.waitSeconds > 0
-- Present the result
SELECT js.scheduleID,
js.serverName,
js.jobName,
js.jobDescription,
js.scheduleName,
js.categoryName,
st.infoDate,
st.startTime,
st.endTime,
js.jobEnabled,
js.scheduleEnabled
FROM #scheduleTimes AS st
INNER JOIN #jobSchedules AS js ON js.rowID = st.rowID
-- Clean up
DROP TABLE #jobSchedules,
#dayInformation,
#scheduleTimes,
#tallyNumbers
要获得有关时间表的有意义的简短描述，例如“每周一下午2:00”，请查看sp_get_schedule_description sproc。
答案 2 :(得分：1)
我们创建了sql'server作业，定期（看着你每小时声称就足够了）称为一些存储过程。另一方面，这个程序选择做两件事：