2つの日付間の1か月あたりの日数をカウントする - SQL

Question

私は間違った用語を使用していて、私が書いたコードが恐ろしい場合、私は非常にSQLに慣れています。

私は1年あたりの月数を出力するクエリを作成しようとしています。私が作成したテストテーブルはdatetestです：

startdate enddate 
2105-12-16 2016-02-15 
2017-01-01 2017-01-02 

と私の下に貼り付けたクエリを使用して、次の結果を得る：

Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 
2015 31 15 0 0 0 0 0 0 0 0 0 16 
2017 4 0 0 0 0 0 0 0 0 0 0 0 

日付範囲は、2年間で行くとき私がいる問題があります結果は実際の年ではなく開始日の年になります。

誰もがそう、私はあなたがそう、この概念はうまくいくかもしれない、あなたの構文を見て、SQLサーバーを使用していない知っている私はそれを本当に感謝 おかげ

---

クエリ

SELECT 
year(startdate) AS Year, 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-01-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-01-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-01-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-01-01') AS DATETIME),startdate))+1,0)) AS "Jan", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-02-28') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-02-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-02-28') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-02-01') AS DATETIME),startdate))+1,0)) AS "Feb", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-03-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-03-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-03-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-03-01') AS DATETIME),startdate))+1,0)) AS "Mar", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-04-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-04-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-04-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-04-01') AS DATETIME),startdate))+1,0)) AS "Apr", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-05-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-05-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-05-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-05-01') AS DATETIME),startdate))+1,0)) AS "May", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-06-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-06-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-06-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-06-01') AS DATETIME),startdate))+1,0)) AS "Jun", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-07-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-07-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-07-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-07-01') AS DATETIME),startdate))+1,0)) AS "Jul", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-08-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-08-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-08-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-08-01') AS DATETIME),startdate))+1,0)) AS "Aug", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-09-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-09-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-09-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-09-01') AS DATETIME),startdate))+1,0)) AS "Sep", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-10-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-10-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-10-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-10-01') AS DATETIME),startdate))+1,0)) AS "Oct", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-11-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-11-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-11-30') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-11-01') AS DATETIME),startdate))+1,0)) AS "Nov", 

SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(startdate), '-12-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(startdate), '-12-01') AS DATETIME),startdate))+1,0))+SUM(GREATEST(datediff(least(CAST(CONCAT(YEAR(enddate), '-12-31') AS DATETIME),enddate),greatest(CAST(CONCAT(YEAR(enddate), '-12-01') AS DATETIME),startdate))+1,0)) AS "Dec" 

from datetest 

GROUP BY year(startdate) 

Answer 1

を助けることができる場合あなたが持っているrdbmsに依存するかどうかは分かりません。しかし、コンセプトは、2つの日付の差が最も大きい集計表を作成し、元の表に結合して日付範囲を個々の行に均等化し、その情報を戻すだけです。

DECLARE @Table AS TABLE (startdate DATETIME, enddate DATETIME) 
INSERT INTO @Table (startdate, enddate) VALUES ('2015-12-16','2016-02-15'),('2017-01-01','2017-01-02') 

;WITH cteTally as (
    SELECT MAX(DATEDIFF(day,startdate,enddate)) - 1 as Tally 
    FROM 
     @Table 

    UNION ALL 

    SELECT Tally - 1 
    FROM 
     cteTally 
    WHERE 
     Tally - 1 >= 0 
) 

, cteDates AS (
    SELECT 
     DATEADD(day,c.Tally,t.startdate) as date 
     ,YEAR(DATEADD(day,c.Tally,t.startdate)) as [Year] 
     ,LEFT(DATENAME(month,DATEADD(day,c.Tally,t.startdate)),3) as [Month] 
    FROM 
     @Table t 
     INNER JOIN cteTally c 
     ON DATEDIFF(day,t.startdate,t.enddate) - 1 >= c.Tally 
) 

SELECT * 
FROM 
    cteDates 
    PIVOT (
     COUNT([date]) 
     for [Month] IN ([Jan],[Feb],[Mar],[Apr],[May],[Jun],[Jul],[Aug],[Sep],[Oct],[Nov],[Dec]) 
    ) p 

oracle、postgressql、およびsql-serverは、このタイプの構造をサポートしています。この場合、mysqlはそうしません。