T-SQL JOIN table on condition - tsql

Question

I have a procedure with select from 4 tables. In dependence on input variable, I want to join and show 2 or all 4. So if the #var = 1 only table1 and table2 will be output, if #var = 2 join from all 4 tables is output.
What I have tried is this:
SELECT * FROM t1
JOIN t2 ON t1.pk1 = t2.pk1
LEFT JOIN t3 ON t2.pk2 = t3.pk2 AND CASE #var = 2 THEN 1 ELSE = 0 END = 1
LEFT JOIN t4 ON t3.pk3 = t4.pk3 AND CASE #var = 2 THEN 1 ELSE = 0 END = 1
But the output is shown only for variable = 2 otherwise the output is empty.
In reality I have a lot more tables connected together and these 2 are needed only for WHERE part (more filtering). When I don't get any columns to filter from t3 and t4 I want to show everything without any connection with these tables. When I have data to filter from t3 and t4 I need to join them and use them in WHERE clause.
EDIT:
So the final solution is like this:
IF (#var = 2)
BEGIN
SELECT * FROM t1
JOIN t2 ON t1.pk1 = t2.pk1
LEFT JOIN t3 ON t2.pk2 = t3.pk2
LEFT JOIN t4 ON t3.pk3 = t4.pk3
END
ELSE
BEGIN
SELECT * FROM t1
JOIN t2 ON t1.pk1 = t2.pk1
END

Answer 1

Are you looking for this?
Let's create a mockup scenario to simulate your issue:
DECLARE #t1 TABLE(ID INT IDENTITY,SomeValue VARCHAR(100));
DECLARE #t2 TABLE(ID INT IDENTITY,t1ID INT /*FK to t1*/,SomeValue VARCHAR(100));
DECLARE #t3 TABLE(ID INT IDENTITY,t2ID INT /*FK to t2*/,SomeValue VARCHAR(100));
DECLARE #t4 TABLE(ID INT IDENTITY,t3ID INT /*FK to t3*/,SomeValue VARCHAR(100));
INSERT INTO #t1 VALUES('Row 1'),('Row 2');
INSERT INTO #t2 VALUES(1,'Row 1.1'),(1,'Row 1.2'),(2,'Row 2.1');
INSERT INTO #t3 VALUES(1,'Row 1.1.1'),(1,'Row 1.1.2'),(2,'Row 1.2.1'),(3,'Row 2.1.1');
INSERT INTO #t4 VALUES(1,'Row 1.1.1.1'),(1,'Row 1.1.1.2'),(4,'Row 2.1.1.1'),(4,'Row 2.1.1.2');
--Now we need your variable #var
DECLARE #var INT;
--We can set this to either 1 or 2 (or leave it unset, which is NULL)
--SET #var=1; --show all
SET #var=2; --just t1 and t2
--Check the query's result with various input:
SELECT *
FROM #t1 t1
INNER JOIN #t2 t2 ON t1.ID = t2.t1ID
LEFT JOIN #t3 t3 ON t2.ID = t3.t2ID AND ISNULL(#var,1)=1
LEFT JOIN #t4 t4 ON t3.ID = t4.t3ID AND ISNULL(#var,1)=1;
The idea in short:
The ON-clause will return FALSE due to the AND in any case, where #var is neiter NULL nor 1.
Both values 1 and NULL will be taken as 1. Any other value will not be equal to 1.
But I'm not fully sure, if I really got what you need...