Oracle - Derived table, Inserting values into a new table - sql

Question

I have a function that will read through a list of comma-separated values within a field. The result of this function will create a derived table.
Example:
derived Table:
Now, I need to get this data inserted into a new/existing table. I created and object type/package that has all the columns from the derived table, however, I'm not sure how to add/reference this data to be able to insert into the new/existing table.

Answer 1

Functions that return a nested table can be converted into regular rows and columns by selecting from the table function using the TABLE function. (And in recent versions of Oracle, the TABLE operator is optional.)
For example:
SQL> create or replace function get_columns return sys.odcivarchar2list is
2 begin
3 return sys.odcivarchar2list('355352', 'Yes');
4 end;
5 /
Function created.
SQL> select column_value from table(get_columns);
COLUMN_VALUE
--------------------------------------------------------------------------------
355352
Yes
Now the function can be used anywhere, like an insert: insert into some_table select column_value from table(get_columns);
Your specific solution may be more complicated if the results contain multiple columns, or you have multiple result sets, etc. Edit your question with more details about the function and types if you still need help.
Edit: Below is a full example of storing CSV values in a column, creating a function to convert CSV to a VARRAY, calling that function to pivot and then pivot again back into multiple columns:
create table table_1
(
request_id number,
service varchar2(100),
method varchar2(100),
request_date date,
process_date date,
te_suid number,
status varchar2(100),
params varchar2(4000)
);
create or replace type params_varray is varray(48) of varchar2(4000);
--Example function that converts comma-separated lists into VARRAYS.
--(Not tested for nulls, consecutive commas, etc.)
create or replace function get_columns(p_params varchar2, p_delimiter varchar2) return params_varray is
v_array params_varray := params_varray();
begin
for i in 1 .. regexp_count(p_params, p_delimiter) + 1 loop
v_array.extend;
v_array(v_array.count) := regexp_substr(p_params, '[^'||p_delimiter||']+', 1, i);
end loop;
return v_array;
end;
/
insert into table_1 values(1,1,1,sysdate,sysdate,1,1,'1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48');
insert into table_1 values(2,2,2,sysdate,sysdate,2,2,'49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96');
--Convert CSV into multiple columns.
--
--Convert rows back into multiple columns.
select request_id, service, method, request_date, process_date, te_suid, status
,max(case when param_index = 1 then param_value else null end) value1
,max(case when param_index = 2 then param_value else null end) value2
,max(case when param_index = 3 then param_value else null end) value3
--Repeat for 4 through 48.
from
(
--Convert CSV to rows.
select request_id, service, method, request_date, process_date, te_suid, status,
row_number() over (partition by table_1.rowid order by rownum) param_index,
column_value param_value
from table_1
cross join get_columns(table_1.params, ',')
)
group by request_id, service, method, request_date, process_date, te_suid, status;
You can run the example on db<>fiddle.
But if you have any choice in the matter, I strongly suggest you denormalize your table and never store delimited values in a database.