0

请原谅我问了这么长的问题，但我急需帮助。当我运行下面的查询时，我会得到每日数据。但我想要每月数据，使用 LASTDAY、SUM、GROUP BY 函数。例如，如果我在#startDt# 中选择“20130501”，在#endDt# 中选择“20130701”，它应该给我 3 行累积结果；五月、六月和七月。任何人都可以帮忙吗？

SELECT X.*
 FROM
      ( 
        SELECT 
                    (CASE
                             WHEN DAY.DAY = 'SUM' THEN DAY.DAY
                             ELSE TO_CHAR (TO_DATE (DAY.DAY, 'YYYY-MM-DD'), 'YYYYMMDD') 
                       END
                    ) AS DAY,
                    DAY.WEEK,
                    MNOT.SUM_STD_CNT AS MNOT_CNT,
                    RSC.MOVIE_500K AS MOVIE_500K_CNT,
                    RSC.MOVIE_1M AS MOVIE_1M_CNT,
                    RSC.MOVIE AS MOVIE_CNT,
                    RSC.EXAM AS EXAM_CNT,
                    RSC.WEB AS WEB_CNT,
                    RSC.IMG AS IMG_CNT,
                    RSC.INTERRAC AS INTERRACTIVE_CNT,
                    RSC.DOC AS DOC_CNT,
                    MNOT.SUM_STD_CNT + RSC.MOVIE_500K + RSC.MOVIE_1M + RSC.MOVIE + RSC.EXAM + RSC.WEB + RSC.IMG + RSC.INTERRAC + RSC.DOC TOT_CNT
          FROM 
                    /* DATES */
                   (SELECT 
                                TO_CHAR (TO_DATE (#startDt#,'YYYY-MM-DD')+LEVEL- 1, 'YYYYMMDD') AS DAY

                      FROM DUAL
                      CONNECT BY TO_DATE(#startDt#, 'YYYY-MM-DD') + LEVEL - 1  <![CDATA[<=]]>  TO_DATE(#endDt#, 'YYYY-MM-DD')
                      UNION ALL
                      SELECT 'SUM', '' FROM DUAL
                    ) DAY LEFT OUTER JOIN
                   /* RESOURCE */
                    (
                     SELECT 
                                 NVL(DT_G.COMM_DT, 'SUM') COMM_DT
                                ,NVL(SUM(DT_G.MOVIE_500K), 0) AS MOVIE_500K
                                ,NVL(SUM(DT_G.MOVIE_1M), 0) AS MOVIE_1M
                                ,NVL(SUM(DT_G.MOVIE), 0) AS MOVIE
                                ,NVL(SUM(DT_G.EXAM), 0) AS EXAM
                                ,NVL(SUM(DT_G.DOC), 0) AS DOC 
                                ,NVL(SUM(DT_G.IMG), 0) AS IMG
                                ,NVL(SUM(DT_G.WEB), 0) AS WEB
                                ,NVL(SUM(DT_G.INTERRAC), 0) AS INTERRAC
                       FROM 
                                (
                                 SELECT
                                              COMM_DT
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP10' THEN SUMT.SUM_500K END MOVIE_500K
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP10' THEN SUMT.SUM_1M END AS MOVIE_1M
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP10' THEN SUMT.SUM_500K+SUMT.SUM_1M END AS MOVIE
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP11' THEN SUMT.SUM_STD_CNT END AS EXAM
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP12' THEN SUMT.SUM_STD_CNT END AS DOC
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP13' THEN SUMT.SUM_STD_CNT END AS IMG
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP14' THEN SUMT.SUM_STD_CNT END AS WEB
                                            , CASE SUMT.RSC_TP_DSCD WHEN 'RTP01' THEN SUMT.SUM_STD_CNT END AS INTERRAC
                                  FROM (
                                              SELECT RSC_TP_DSCD, SUM(STDY_CNT) AS SUM_STD_CNT, SUM(MOVIE_STDY_CNT_N1M) AS SUM_1M, SUM(MOVIE_STDY_CNT_N500K) AS SUM_500K, COMM_DT
                                              FROM (
                                                        SELECT RSC_SNO, STDY_CNT, MOVIE_STDY_CNT_N1M, MOVIE_STDY_CNT_N500K, COMM_DT
                                                          FROM LRMS.V_EBSM_PKG_RSC_COMM_CNT
                                                        WHERE PKG_SNO = 0  AND RSC_SNO != 0
                                                        AND COMM_DT BETWEEN TO_CHAR(TO_DATE(#startDt#, 'YYYY-MM-DD'), 'YYYYMMDD') AND TO_CHAR(TO_DATE(#endDt#, 'YYYY-MM-DD'), 'YYYYMMDD')
                                                        ) CNT
                                                   LEFT OUTER JOIN LRMS.V_LRRM_RSC RSC ON CNT.RSC_SNO = RSC.RSC_SNO
                                              GROUP BY RSC_TP_DSCD, COMM_DT
                                            ) SUMT
                                ) DT_G
                         GROUP BY ROLLUP(DT_G.COMM_DT)
                    ) RSC ON DAY.DAY = RSC.COMM_DT

                    LEFT OUTER JOIN
                    (
                      SELECT NVL(SUM(STDY_CNT), 0) AS SUM_STD_CNT, NVL(COMM_DT, 'SUM') COMM_DT
                        FROM LRMS.V_EBSM_PKG_RSC_COMM_CNT
                      WHERE PKG_SNO != 0  AND RSC_SNO = 0
                                AND COMM_DT BETWEEN TO_CHAR(TO_DATE(#startDt#, 'YYYY-MM-DD'), 'YYYYMMDD') AND TO_CHAR(TO_DATE(#endDt#, 'YYYY-MM-DD'), 'YYYYMMDD')
                      GROUP BY ROLLUP(COMM_DT)
                    ) MNOT  ON DAY.DAY = MNOT.COMM_DT
        ) X

 WHERE 1=1
 and X.TOT_CNT IS NOT NULL
4

1 回答 1

1

你可以做类似的事情

SELECT DATENAME(MONTH , DAY) , SUM(COLUMN1) , SUM(COLUMN2) ..
FROM (YOURQUERY) AS A
GROUP BY DATENAME(MONTH , DAY)

我想这会对你有所帮助。

于 2013-07-04T03:14:39.827 回答