Question

I’m having a problems filtering the duplicate values from my query. This is my query

SELECT  cid.IDS_NO,cid.SUB_TITLE,cid.E_SUB_NAME, (cid.SUB_TITLE || ' '|| cid.E_SUB_NAME),
(SELECT(INST_BLK_HSE || ' ' || (SELECT E_STREET_NAME FROM street_table where street_code = d.INST_ST_CODE)) 
FROM del d where d.IDS_NO = cid.IDS_NO and ROWNUM = 1), 
(SELECT INST_ST_LEVEL || '-' || INST_STUNIT FROM detail d where d.IDS_NO = cid.IDS_NO and rownum = 1),
('COUNTRY' ||' ' || pl_post),pl_post 
FROM SUBSCRIBER cid where pl_postal_district = 15 and rownum < 3001 
and hi_property_type IN ('CONDO','COMMERCIAL BUILDING');

These are the parts I’m having problems with:

(SELECT(INST_BLK_HSE || ' ' || (SELECT E_STREET_NAME FROM street_table where street_code = d.INST_ST_CODE)) 
FROM del d where d.IDS_NO = cid.IDS_NO and ROWNUM = 1), 
(SELECT INST_ST_LEVEL || '-' || INST_STUNIT FROM detail d where d.IDS_NO = cid.IDS_NO and rownum = 1);

These parts always get multiple records, and I want to get only 1 record with the lowest sequence number, So here’s what I did:

(SELECT(INST_BLK_HSE || ' ' || (SELECT E_STREET_NAME FROM street_table where street_code = d.INST_ST_CODE)) 
FROM del d where d.IDS_NO = cid.IDS_NO and ROWNUM = 1 ORDER BY d.SEQ_NO DESC), 
(SELECT INST_ST_LEVEL || '-' || INST_STUNIT FROM detail d where d.IDS_NO = cid.IDS_NO and rownum = 1 ORDER BY d.SEQ_NO DESC);

But I always get an error saying missing parenthesis, but when I remove the order by, it works fine. What should I do to get my query running correctly?

Answer 1

Formatting your query to make it easier to read goes a long way to begin with.

But you also have a rather tricky problem, since you ..

want to get only 1 record with the lowest sequence number.

Should work like this:

SELECT s.ids_no
      ,s.sub_title
      ,s.e_sub_name
      ,(s.sub_title || ' '|| s.e_sub_name) AS title_sub_name
      ,d.some_col_name
      ,t.inst_name
      ,('COUNTRY' ||' ' || s.pl_post) AS country_post
      ,s.pl_post
FROM   subscriber s

LEFT   JOIN (
   SELECT d.ids_no
         ,(d.inst_blk_hse || ' ' || st.e_street_name) AS some_col_name
         ,ROW_NUMBER() OVER (PARTITION BY d.ids_no ORDER BY d.seq_no) AS rn
   FROM   del d
   LEFT   JOIN e_street_name st ON st.street_code = d.inst_st_code
   ) d ON d.ids_no = s.ids_no AND d.rn = 1

LEFT   JOIN (
   SELECT ids_no
         ,(inst_st_level || '-' || inst_stunit)  AS inst_name
         ,ROW_NUMBER() OVER (PARTITION BY d.ids_no ORDER BY d.seq_no) AS rn
   FROM   detail         
   ) t ON t.ids_no = s.ids_no AND t.rn = 1


WHERE  s.pl_postal_district = 15
AND    s.rownum < 3001 
AND    s.hi_property_type IN ('CONDO','COMMERCIAL BUILDING');

More details on how the subqueries work under this related question:
Select first row in each GROUP BY group?

The core features is ROW_NUMBER() to get the record with the lowest sequence number.
I rewrote your correlated subqueries, because they generally suck performance-wise. They are also hard to read.
I use LEFT JOIN, so you still get a row in case the subquery shouldn’t find anything to match.