Question

This is a script to produce the dataset of a nested table:

create table descriptions(
id number,
description varchar2(10)
);

insert into descriptions values(1,'abfd');
insert into descriptions values(1,'frte');
insert into descriptions values(2,'eewd');
insert into descriptions values(2,'outr');

create or replace type nt_description as table of varchar2(10);
-- the following query produces a dataset to save as a flat file
select id, cast(collect(description) as nt_description) as nested_column from descriptions group by id;

What I want to achieve is to produce flat output txt file. However, the result of the nested column output as flat file is unexpected when using SQL Developer GUI. The format of the output file is following when exporting data as text:

"ID"    "NESTED_COLUMN"
1   DBUSER.NT_DESCRIPTION('abfd','frte')
2   DBUSER.NT_DESCRIPTION('eewd','outr')

I would like to get the following format instead of preceding format:

"ID"    "NESTED_COLUMN"
1   'abfd','frte'
2   'eewd','outr'

I can achieve it by using text editor and replacing or regexing undesirable data (with the help of notepad++) but is it possible to get the same result without manually replacing “DBUSER.NT_DESCRIPTION(” and “)” by void string?

Answer 1

If you obligatory need use collect, you can make a function as shown in this link.

If not, you can use wm_concat:

select d.id, 
       wm_concat(d.description)
       nested_column 
from 
       descriptions d
group by id;