ORDER BY上PLSQL

Question

問題

我已經設置的數據應取出來，應該按以下順序

Last Name  First Name  Location 
Blake   James   101007 
fusio   Wiko   101008 
        100 Total 

FAROMOJU  BANKOLE   303315 
Gibbs   Rhonda   303315 
Greene   Leette   303315 
        331 Total 

...... 
............ 

在該準備一份報告，我們可以看到，3字母代碼 [從提取位置代碼 - CONCAT（SUBSTR（COLUMN3,3,3），'Total'）]在所有位置設置後。我只是想出報告，除了地點代碼降序的情況外，它工作得很好。它使用'位置'進行排序後，將其轉換爲3個字母的代碼，而不是使用完整的數字代碼。所以在結果中降序不能很好地工作，因爲我需要。請給我一個解決方案，

我附上我的代碼，

****************************************************** 

FUNCTION GET_ACTY_SUMMARY(V_STARTDATE IN DATE,V_ENDDATE IN DATE) 
RETURN TEMP_OUTPUT_TABLE PIPELINED IS 
CURSOR ACTY_SUMMARY IS 

/* The query fetches the report field grouped by First Name,Last Name 
and Location.*/ 

    SELECT 
      COLUMN1 AS LASTNAME, 
      COLUMN2 AS FIRSTNAME, 
      COLUMN3 AS LOCATION, 
      SUM(COLUMN4) AS ACTIVITYNM1, 
      SUM(COLUMN5) AS ACTIVITYNM2, 
      SUM(COLUMN6) AS ACTIVITYNM3, 
      SUM(COLUMN7) AS ACTIVITYNM4, 
      SUM(COLUMN8) AS ACTIVITYNM5, 
      SUM(COLUMN9) AS REG_HRS, 
      SUM(COLUMN10) AS OT_HRS, 
      SUM(COLUMN11) AS TOTAL_HRS, 
      SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 

    FROM 
    TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
    GROUP BY 
    COLUMN1, 
    COLUMN2, 
    COLUMN3 


    UNION 

    SELECT NULL, 
     NULL, 
     CONCAT(SUBSTR(COLUMN3,3,3),'Total'), 
     SUM(COLUMN4) AS ACTIVITYNM1, 
     SUM(COLUMN5) AS ACTIVITYNM2, 
     SUM(COLUMN6) AS ACTIVITYNM3, 
     SUM(COLUMN7) AS ACTIVITYNM4, 
     SUM(COLUMN8) AS ACTIVITYNM5, 
     SUM(COLUMN9) AS REG_HRS, 
     SUM(COLUMN10) AS OT_HRS, 
     SUM(COLUMN11) AS TOTAL_HRS, 
     SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 

    FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
    GROUP BY COLUMN3 
    UNION 

    SELECT NULL, 
     NULL, 
     'Total', 
     SUM(COLUMN4) AS ACTIVITYNM1, 
     SUM(COLUMN5) AS ACTIVITYNM2, 
     SUM(COLUMN6) AS ACTIVITYNM3, 
     SUM(COLUMN7) AS ACTIVITYNM4, 
     SUM(COLUMN8) AS ACTIVITYNM5, 
     SUM(COLUMN9) AS REG_HRS, 
     SUM(COLUMN10) AS OT_HRS, 
     SUM(COLUMN11) AS TOTAL_HRS, 
     SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 

    FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 


    ORDER BY 3 desc ; 

GAS ACTY_SUMMARY%ROWTYPE; 

/*Intialize the table output format with NULL.*/ 

TT TEMP_OUTPUT_FORMAT := INITIALIZE_TABLE_FORMAT(); 
    BEGIN 
    OPEN ACTY_SUMMARY; 
    FETCH ACTY_SUMMARY INTO GAS; 
    WHILE ACTY_SUMMARY%FOUND LOOP 
     BEGIN 
     TT.COLUMN1 := GAS.LASTNAME; 
     TT.COLUMN2 := GAS.FIRSTNAME; 
     TT.COLUMN3 := GAS.LOCATION; 
     TT.COLUMN4 := GAS.ACTIVITYNM1; 
     TT.COLUMN5 := GAS.ACTIVITYNM2; 
     TT.COLUMN6 := GAS.ACTIVITYNM3; 
     TT.COLUMN7 := GAS.ACTIVITYNM4; 
     TT.COLUMN8 := GAS.ACTIVITYNM5; 
     TT.COLUMN9 := GAS.REG_HRS; 
     TT.COLUMN10 := GAS.OT_HRS; 
     TT.COLUMN11 := GAS.TOTAL_HRS; 
     TT.COLUMN12 := GAS.PRODUCTIVITY_PERCENTAGE; 
     PIPE ROW(TT); 
    END; 
    FETCH ACTY_SUMMARY INTO GAS; 
    END LOOP; 
    CLOSE ACTY_SUMMARY; 
    RETURN; 
END GET_ACTY_SUMMARY; 

END GEHC_ACTY_REPT_PKG; 


****************************************************** 

感謝 阿南德

Answer 1

我建議你更改查詢是

SELECT 
      COLUMN1 AS LASTNAME, 
      COLUMN2 AS FIRSTNAME, 
      COLUMN3 AS LOCATION, 
      COLUMN3 AS FULL_LOCATION, 
      SUM(COLUMN4) AS ACTIVITYNM1, 
      SUM(COLUMN5) AS ACTIVITYNM2, 
      SUM(COLUMN6) AS ACTIVITYNM3, 
      SUM(COLUMN7) AS ACTIVITYNM4, 
      SUM(COLUMN8) AS ACTIVITYNM5, 
      SUM(COLUMN9) AS REG_HRS, 
      SUM(COLUMN10) AS OT_HRS, 
      SUM(COLUMN11) AS TOTAL_HRS, 
      SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 
    FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
    GROUP BY 
    COLUMN1, 
    COLUMN2, 
    COLUMN3 
UNION ALL 
    SELECT NULL, 
     NULL, 
     CONCAT(SUBSTR(COLUMN3,3,3),'Total'), 
     COLUMN3 AS FULL_LOCATION, 
     SUM(COLUMN4) AS ACTIVITYNM1, 
     SUM(COLUMN5) AS ACTIVITYNM2, 
     SUM(COLUMN6) AS ACTIVITYNM3, 
     SUM(COLUMN7) AS ACTIVITYNM4, 
     SUM(COLUMN8) AS ACTIVITYNM5, 
     SUM(COLUMN9) AS REG_HRS, 
     SUM(COLUMN10) AS OT_HRS, 
     SUM(COLUMN11) AS TOTAL_HRS, 
     SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 
    FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
    GROUP BY COLUMN3 
UNION ALL 
    SELECT NULL, 
     NULL, 
     'Total', 
     'Total' AS FULL_LOCATION, 
     SUM(COLUMN4) AS ACTIVITYNM1, 
     SUM(COLUMN5) AS ACTIVITYNM2, 
     SUM(COLUMN6) AS ACTIVITYNM3, 
     SUM(COLUMN7) AS ACTIVITYNM4, 
     SUM(COLUMN8) AS ACTIVITYNM5, 
     SUM(COLUMN9) AS REG_HRS, 
     SUM(COLUMN10) AS OT_HRS, 
     SUM(COLUMN11) AS TOTAL_HRS, 
     SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 
    FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
    ORDER BY 4 DESC; 

我們正在試圖這裏要做的是在查詢中添加完整的位置（數據中的COLUMN3，如果我正在讀取原始查詢），以便排序可以基於它。您可能需要更改GROUP BY的 - 我不確定，也沒有可供您測試的數據。但這是基本的想法。

分享和享受。

Answer 2

我會建議看看分組集，而不是使用聯合數據在一起。分組集（與CUBE和ROLLUP一起）是彙總多個級別的數據的方法，這是您似乎在做的事情。使用分組集合將獲得您正在查找的數據。

爲了排序數據，您需要向order by子句添加其他列。我使用的grouping_id（）函數返回一個數字（0或1），這取決於它是否是給定表達式的'superaggregate'行。這些'superaggregate'行是與總計和小計相關的附加行。我正在對數據進行排序以及該列是否是超級聚集的。

認爲你應該能夠做到像：

SELECT 
    COLUMN1 AS LASTNAME, 
    COLUMN2 AS FIRSTNAME, 
    case 
    when grouping_id(column3, substr(column3,3,3)) = 3 then 'Total' 
    when grouping_id(column3, substr(column3,3,3)) = 2 
     then substr(column3,3,3) ||' Total' 
    else column3 end as location, 
    SUM(COLUMN4) AS ACTIVITYNM1, 
    SUM(COLUMN5) AS ACTIVITYNM2, 
    SUM(COLUMN6) AS ACTIVITYNM3, 
    SUM(COLUMN7) AS ACTIVITYNM4, 
    SUM(COLUMN8) AS ACTIVITYNM5, 
    SUM(COLUMN9) AS REG_HRS, 
    SUM(COLUMN10) AS OT_HRS, 
    SUM(COLUMN11) AS TOTAL_HRS, 
    SUM(COLUMN12) AS PRODUCTIVITY_PERCENTAGE 
FROM TABLE(ACTY_COLUMN(V_STARTDATE,V_ENDDATE)) 
GROUP BY grouping sets ((), (substr(column3,3,3)), 
    (COLUMN1,COLUMN2,COLUMN3, substr(column3,3,3))) 
order by 
    grouping_id(substr(column3,3,3)), 
    substr(column3,3,3) desc, 
    grouping_id(column3, substr(column3,3,3)), 
    column3 desc 

有望運行更快，做你想做的。我可能已經使order by子句比必要的更復雜，但它應該做你需要的。