QMF Turning Sequence Number Results into their own columns without PIVOT - sql

Question

I'm sorry, I'm new and have almost no training. I've been searching for a few days on this and maybe I'm just not using the correct terms...
Using QMF for Windows.
I have 3 columns in my ADDRESSTABLE table - address identifier codes, address line sequence numbers and their corresponding address lines. ADR_CODE, SEQ_NO, ADRS_LINE.
Each address record has between 3 and 5 lines, and thusly, 3 to 5 sequence numbers. So, when I pull a query for address identifier codes, I get 3-5 repetitions of the address identifier code. Like so:
SELECT DISTINCT A.ADR_CODE, A.SEQ_NO, A.ADRS_LINE
FROM ADDRESSTABLE A
WHERE (A.ADR_CODE LIKE 'A%')
And I get:
ADR_CODE SEQ_NO ADRS_LINE
AAAA 1 JOHN DOE
AAAA 2 123 HAPPY STREET
AAAA 3 ANYWHERE, NY
AAAA 4 12345
AABB 1 234 MAIN STREET
AABB 2 SOMEWHERE, MN
AABB 3 34567
ACDE 1 MR PINK
ACDE 2 21 RESERVOIR RD
ACDE 3 APT 4
ACDE 4 LOS ANGELES
ACDE 5 90210
And I figured out that if I do:
SELECT DISTINCT A.ADR_CODE, MIN(A.SEQ_NO CONCAT A.ADRS_LINE) AS
"FIRST ADDRESS LINE", MAX(A.SEQ_NO CONCAT A.ADRS_LINE) AS
"LAST ADDRESS LINE"
FROM ADDRESSTABLE A
WHERE (A.ADR_CODE LIKE 'A%')
ORDER BY A.ADR_CODE ASC
GROUP BY A.ADR_CODE
I get:
ADR_CODE FIRST ADDRESS LINE LAST ADDRESS LINE
AAAA 1JOHN DOE 412345
AABB 1234 MAIN STREET 334567
ACDE 1MR PINK 590210
My question is, how do I get the rest of those in between lines? MIN+1 and MAX-1 is illegal, MIN(A.SEQ_NO +1... and MAX(A.SEQ_NO-1... is illegal. I'm stuck and I don't want to use PIVOT because I want this whole thing to be part of a larger query. In short, My query should end up with about 7000 rows of freight records - each with their own address in a line - instead of 7000 rows*(3 to 5 address lines per record).Thank you, James

Answer 1

Although not exactly a PIVOT, will a multiple join to same address table work for you as ...
SELECT
A.ADR_CODE,
A.ADRS_LINE as AdrLine1,
A2.ADRS_LINE as AdrLine2,
A3.ADRS_LINE as AdrLine3,
COALESCE( A4.ADRS_LINE, "" ) as AdrLine4,
COALESCE( A5.ADRS_LINE, "" ) as AdrLine5
FROM
ADDRESSTABLE A
JOIN ADDRESSTABLE A2
ON A.ADR_CODE = A2.ADR_CODE
AND A2.SEQ_NO = 2
JOIN ADDRESSTABLE A3
ON A.ADR_CODE = A3.ADR_CODE
AND A3.SEQ_NO = 3
LEFT JOIN ADDRESSTABLE A4
ON A.ADR_CODE = A4.ADR_CODE
AND A4.SEQ_NO = 4
LEFT JOIN ADDRESSTABLE A5
ON A.ADR_CODE = A5.ADR_CODE
AND A5.SEQ_NO = 5
WHERE
A.ADR_CODE LIKE 'A%'
AND A.SEQ_NO = 1
ORDER BY
A.ADR_CODE
Since the query is applied to the address table for only sequence #1, that will keep distinct adr_code entires for you since you are not getting the possible 2-5 address lines.
NOW comes the JOINs. I rejoin to same address table on the same adr_code key, but for each one additionally only for that particular address line... applying LEFT-JOIN for 4 and 5 since you stated 1-3 are always and 4-5 are only POSSIBLE (also applied coalesce() to prevent NULLs).