Question

我有以下代码：

WHILE (v_Count > 0) 
LOOP

Select  v_Columns || CHR(10) ||  CHR(9)  || ColumnName || ',' ,
    v_Values || CHR(10) ||  CHR(9)  ||
    CASE ISNUMERIC(Val) 
        WHEN 1 THEN Val 
        WHEN 0 THEN '''' || REPLACE(LTRIM(RTRIM(Val)), '''','''''') || '''' 
    END
    || ',' into v_Columns,v_Values
FROM 

(           
SELECT D.TableName, D.ColumnName, D.Val
FROM            
(         
    SELECT ID, TableName, ColumnName, Val  
    FROM tblTemplates_Load_OtherObjects_Raw OO  
    JOIN tblTemplates_Fields F ON OO.OtherObjectsField = F.FieldName 
    WHERE TemplateType IN ('All', v_TemplateType) 
    AND OO.Val IS NOT NULL 
    AND  (TemplateVersion = p_TemplateVersion or (TemplateVersion is null 
and p_TemplateVersion <> 'V8')) 

    UNION 

    SELECT FieldID, TableName, ColumnName, Val 
    FROM tblTemplates_Fields_OtherDestinations OD 
    JOIN  
    ( 
        SELECT ID, Val 
        FROM tblTemplates_Load_OtherObjects_Raw OO  
        JOIN tblTemplates_Fields F ON OO.OtherObjectsField = F.FieldName 
        WHERE TemplateType IN ('All', v_TemplateType) 
        AND OO.Val IS NOT NULL 
        AND    (TemplateVersion = p_TemplateVersion or (TemplateVersion is 
null and p_TemplateVersion <> 'V8')) 

    ) UsedIDs 
    ON OD.FieldID = UsedIDs.ID


) D


JOIN USER_TAB_COLS C  ON upper(D.TableName) = upper(C.TABLE_NAME) AND 
upper(D.ColumnName) = upper(C.COLUMN_NAME)
WHERE (v_UpdateComp = 0 OR D.TableName <> 'tblComp') 
AND (v_UpdateCompInd = 0 OR D.TableName <> 'tblCompInd') 
ORDER BY D.TableName, D.ColumnName
)

WHERE TableName = v_TableName ;

 v_Count := v_Count -1; 
end loop;

其中v_count是一个数字，范围可以从5到6。

这里的问题是，由于循环的原因，试图将多个值存储到v_Columns和v_Values中，并且获得 exact精确提取将返回比请求的行数更多的错误。我无法弄清楚如何重写这段代码，以使其正常工作。

我该如何重写此代码？

Answer 1

对我来说，最可疑的是df.var1=df.var1.map(df_new.set_index('df2')['val1']) print(df) var1 0 1 1 2 2 1 3 2，如果两个UNION都返回至少一行，则会产生两行。您不能将两行放入这些变量中并得到错误。

该怎么办？如果可能，请将其重写到游标SELECT循环中，该循环将逐行返回。

精确提取返回的行数超过了请求的行数-重写代码

1 个答案: