How to fetch the data from pl/sql table dynamically

Question

Hi All, I have the requirement of comparing two db views data in pl/sql. So, I have bulk collect the view into pl/sql table. But, the issue is, It is expecting to provide the column name for comparison. But, in my case, column name is dynamic. So, i cannot provide the column name directly for comparison.
For eg: In my view t1_VW, i have 4 columns. stid, c1,c2,c3,c4 and similar structure for t2_vw
my code
TYPE v1_type IS TABLE OF t1_vw%ROWTYPE;
l_data v1_type;
TYPE v1_type1 IS TABLE OF t2_vw%ROWTYPE;
l_data1 v1_type1;
test varchar2(1000);
test1 varchar2(1000);
temp1 number;
begin
SELECT * Bulk collect into l_data
FROM T1_VW;
SELECT * Bulk collect into l_data1
FROM T2_VW;
select l_data(1).stid into temp1 from dual; -- It is working fine and gives me the value properly
-- But, in my case, we are reading the column names from array, i am constructing the query dynamically and execute it.
test :='select l_data(1).stid into temp1 from dual';
execute immediate test into temp1;
-- I am getting error as follows:
Error report:
ORA-00904: "L_DATA": invalid identifier
ORA-06512: at "SYSTEM.BULKCOMPARISON", line 93
ORA-06512: at line 2
00904. 00000 - "%s: invalid identifier"
*Cause:   
*Action
end;
- Please help me to get rid of this issue. Is it possible to construct the query dynamically and fetch the data?. If not, is there any other better approach to compare the data between two views?.

Answer 1

Output should display what are all columns changed and its old value and new value.
For eg., output should be
COLUMNNAME OLD_VALUE NEW_VALUE STID
C1 20 10 1
C2 50 40 2
C3 60 70 2
C2 80 90 3Why no do this only via a simple sql ?
create table a (STID number, C1 number,  C2 number, C3 number);
insert into a values (1, 20, 30, 40)
insert into a values (2, 40, 50, 60);
insert into a values (3, 90, 80, 100);
create table b as select *
from a where 1 = 0;
insert into b values (1, 10, 30, 40)
insert into b values (2, 40, 40, 70);
insert into b values (3, 90, 90, 100);
commit;And now you can issue such a kind of select
SELECT stid , c1, c2, c3                      
   FROM
  ( SELECT a.*,
         1 src1,
         to_number(null) src2        
   FROM  a   
   UNION ALL
   SELECT b.*,
         to_number(null) src1,
         2  src2        
    FROM b
   GROUP BY stid , c1, c2, c3
   HAVING count(src1) <> count(src2)
   order by stid;I would then create a new table a_b_difference having the same structure as a or b and insert into it like this
create table a_b_diff as select * from a where 1 = 0;
insert into a_b_diff
SELECT stid , c1, c2, c3                      
   FROM
  ( SELECT a.*,
         1 src1,
         to_number(null) src2        
   FROM  a   
   UNION ALL
   SELECT b.*,
         to_number(null) src1,
         2  src2        
    FROM b
   GROUP BY stid , c1, c2, c3
   HAVING count(src1) <> count(src2)
   order by stid
   ;Then each time there is a difference between a column in a and its equivalente one in b (per unique stid ) a record will be inserted in this table.
You can do more by adding the name of the table in front of each record in this table to see exactly where the data comes from
Best Regards
Mohamed Houri