SQL Server 查询日期范围之间一个月的总天数

Question

我有特定的日期范围，比如

From Date        To Date
---------------------------    
2012-11-10       2012-11-15
2012-11-21       2012-11-22
2012-11-30       2012-12-01

我想编写一个 SQL 查询来计算两个日期之间的总天数和特定月份的总天数

我想要的输出是，

No of days     month
--------------------
   9            11
   1            12

谁能帮我写这个 SQL 查询？

Answer 1

理想情况下，您有一个名为“日期”的表格，其中包含您将使用的所有日期，例如从 1950 年到 2100 年。此查询将为您提供所需的结果：

  select dateadd(m,datediff(m, 0, d.thedate),0) themonth, count(1)
    from dates d
    join ranges r on d.thedate between r.[from date] and r.[to date]
group by datediff(m, 0, d.thedate)
order by themonth;

结果：

|   themonth | COLUMN_1 |
-------------------------
| 2012-11-01 |        9 |
| 2012-12-01 |        1 |

请注意，不要仅将“11”或“12”显示为月份，如果您的范围超过 12 个月，这将无法正常工作，或者在跨过新年时无助于排序，此查询改为显示该月的第一天。

如果没有，您可以根据下面的扩展查询即时虚拟地创建一个dates 表：

;with dates(thedate) as (
  select dateadd(yy,years.number,0)+days.number
    from master..spt_values years
    join master..spt_values days
      on days.type='p' and days.number < datepart(dy,dateadd(yy,years.number+1,0)-1)
   where years.type='p' and years.number between 100 and 150
      -- note: 100-150 creates dates in the year range 2000-2050
      --       adjust as required
)
  select dateadd(m,datediff(m, 0, d.thedate),0) themonth, count(1)
    from dates d
    join ranges r on d.thedate between r.[from date] and r.[to date]
group by datediff(m, 0, d.thedate)
order by themonth;

此处提供了完整的工作示例：SQL Fiddle

Answer 2

试试这个

select ((day(date_to)) - (day(date_from))) as no_of_days,month(date_from)as month from tablename 

Answer 3

我会将其分解为几个步骤（每个步骤都有一个单独的 CTE）：

declare @Ranges table (FromDate date not null,ToDate date not null)
insert into @Ranges (FromDate,ToDate) values
('20121110','20121115'),
('20121121','20121122'),
('20121130','20121201')

;with Months as (
    select
        DATEADD(month,DATEDIFF(month,'20010101',FromDate),'20010101') as MonthStart,
        DATEADD(month,DATEDIFF(month,'20010101',FromDate),'20010131') as MonthEnd
    from @Ranges
    union /* not all */
    select
        DATEADD(month,DATEDIFF(month,'20010101',ToDate),'20010101') as MonthStart,
        DATEADD(month,DATEDIFF(month,'20010101',ToDate),'20010131') as MonthEnd
    from @Ranges
), MonthRanges as (
    select
        CASE WHEN r.FromDate > m.MonthStart then r.FromDate ELSE m.MonthStart END as StartRange,
        CASE WHEN r.ToDate < m.MonthEnd then r.ToDate ELSE m.MonthEnd END as EndRange
    from
        @Ranges r
            inner join
        Months m
            on
                r.ToDate >= m.MonthStart and
                r.FromDate <= m.MonthEnd
)
select
    DATEPART(month,StartRange),
    SUM(DATEDIFF(day,StartRange,EndRange)+1) /* Inclusive */
from
    MonthRanges
group by
    DATEPART(month,StartRange)

首先，Months CTE 查找我们可能感兴趣的每个月的第一天和最后几天 (*)。然后，MonthRanges 将此数据与原始范围重新组合，并根据需要对其进行拆分，以便我们处理的每个时间段仅代表一个月中的几天。然后我们可以只使用DATEDIFF 来计算每个范围跨越的天数（并添加 1，因为我们正在处理日期并想要包含值）

(*) Months CTE 将起作用，前提是我们不处理跨越多个月的任何范围，并且在中间几个月内没有其他范围开始或结束。如果您需要应对这种情况，我需要修改Months CTE。例如。如果我们在上面的示例中添加('20120115','20120315')（并且没有其他范围），我们将不会使用上面的方法得到二月份的结果。我们需要应对这种情况吗？

---

为了应对 (*) 中指出的情况，我们可以将上述查询中的 Months CTE 替换为：

;With LastMonth as (
    select MAX(ToDate) as Mx from @Ranges
), MultiMonths as (
    select
        DATEADD(month,DATEDIFF(month,'20010101',FromDate),'20010101') as MonthStart,
        DATEADD(month,DATEDIFF(month,'20010101',FromDate),'20010131') as MonthEnd
    from @Ranges
    union all
    select
        DATEADD(month,1,MonthStart),
        DATEADD(month,1,MonthEnd)
    from MultiMonths
    where MonthStart <= (select Mx from LastMonth)
), Months as (
    select distinct MonthStart,MonthEnd from MultiMonths 
)

Answer 4

请原谅我写得不好的 SQL。
假设是月份差异。 fromdate 和 todate 之间是 1。

架构

CREATE TABLE dateData
    (fromdate datetime, todate datetime)
;

INSERT INTO dateData
    (fromdate, todate)
VALUES
    ('2012-11-10', '2012-11-15'),
    ('2012-11-21', '2012-11-22'),
    ('2012-11-30', '2012-12-01')
;

SQL

select mth, sum(days) as daysInMth
from
(
select month(fromdate) as mth,
sum(case 
when month(fromdate) = month(todate) then datediff(dd, fromdate, todate)+1
else datediff(dd, fromdate, dateadd(mm, 1, fromdate) - day(fromdate)) + 1 end)
as days 
from dateData
group by month(fromdate)
union
select month(todate) as mth,
sum(case when month(todate) <> month(fromdate) then
datediff(dd, fromdate, dateadd(mm, 1, fromdate) - day(fromdate)) + 1
else 
case when month(todate) = month(fromdate) then 0 else
datediff(dd, convert(datetime, year(todate) + '-' + month(todate) + '-1'), todate) 
end
end) as days
from dateData
group by month(todate)
) aggregated
group by mth

在 SQLFiddle 上查看：http://www.sqlfiddle.com/#!3/9f7da/56