SQL inner join if not found omit condition - sql

Question

SQL inner join if not found omit condition
Suppose I have a SQL query like:
SELECT A.* FROM Table_A A
INNER JOIN Table_B B ON A.Field1 = B.Field1 AND A.Field2 = B.Field2
WHERE A.Field3 > 10
What I'd like to achieve is, we first inner join by Field1/Field2, if no data found only join by Field1.
What's the proper way to this?
Table A:
Id Field1 Field2 Field3
1 a b 12
2 a c 13
3 e f 14
2 d c 15
Table B:
Id Field1 Field2
1 a b
2 e g
Result should be:
Id Field1 Field2 Field3
1 a b 12
3 e f 14

Answer 1

First part of the union is your original. Second part joins only on the first set of fields in cases where there is no match for both sets of fields.
SELECT * FROM Table_A A
INNER JOIN Table_B B ON A.Field1 = B.Field1 AND A.Field2 = B.Field2
WHERE A.Field3 > 10
UNION
SELECT * FROM Table_A A
INNER JOIN Table_B B ON A.Field1 = B.Field1
WHERE A.Field3 > 10
AND NOT EXISTS
(SELECT 1 FROM TABLE_B B2
WHERE A.Field1 = B2.Field1 AND A.Field2 = B2.Field2);

Answer 2

You can join on the same table twice, one with inner join and one with left join and put the conditional logic on the left join.
select distinct t1.*
from table1 t1
left join table2 t2 on t1.f2 = t2.f2
inner join table2 t3 on t1.f1 = t3.f1

Answer 3

You can use CASE in INNER JOIN. Just check this :
declare #a table (col1 int, col2 int, col3 int)
declare #b table (col1 int, col2 int, col3 int)
Insert into #a values(1,1,11),(2,2,15),(3,4,20),(4,5,9)
Insert into #b values(1,1,5),(2,2,2),(3,3,20),(4,5,20)
select A.* from #a A inner join #b B on
(CASE
WHEN A.col2 = B.col2 then 1
WHEN A.col2 <> B.col2 then 1
ELSE 1
END) = 1 and A.col1 = B.col1
where A.col3 > 10

Answer 4

You could try the following query:
SELECT ta.*
FROM #Table_A ta
INNER JOIN #Table_B tb ON ta.Field1 = tb.Field1
WHERE ta.Field3 > 10
AND
(
ta.Field2 = tb.Field2
OR
NOT EXISTS (
SELECT 1 FROM #Table_A ta2
INNER JOIN #Table_B tb2 ON ta2.Field1 = tb2.Field1 AND ta2.Field2 = tb2.Field2
WHERE ta2.Field1 = ta.Field1
)
)
Demo link: http://rextester.com/DKRG30176

Answer 5

;With Tab1(Id,Field1, Field2, Field3)
AS
(
SELECT 1,'a','b',12 Union all
SELECT 2,'a','c',13 Union all
SELECT 3,'e','f',14 Union all
SELECT 2,'d','c',15
)
,Tab2 (Id, Field1, Field2)
AS
(
SELECT 1,'a','b' Union ALL
SELECT 2,'e','g'
)
SELECT Id,Field1, Field2, Field3 From Tab1
EXCEPT
SELECT Id
,Field1
,Field2
,Field3
FROM (
SELECT I.Id
,i.Field1
,i.Field2
,i.Field3
,COUNT(i.Id) OVER (
PARTITION BY i.Id ,i.Field1, i.Field2 ORDER BY i.Id
) Seq
FROM Tab1 i
LEFT JOIN Tab2 o ON i.Id = o.Id
WHERE i.Field1 !=o.Field1 AND i.Field2 !=o.Field2
) Dt
WHERE Dt.Seq = 1 AND Dt.Field3 > 10
OutPut
Id Field1 Field2 Field3
---------------------------
1 a b 12
3 e f 14