Reputation: 182
Lets say I have following tables:
Table1
ID Number
1 2
2 34
3 1 <---- Input (ID = 3) ==> (Number = 1)
4 6
5 5
*6* 7 <---- Want to find (Number = 6) because match in Table2
7 22
and Table2
Number Code Att1 Att2 Att3
1 1 1 <-----|
1 2 1 2 <-----|
6 2 f 2 |
6 3 4 3 2 |
2 4 6 |---Match
22 5 2 2 2 |
5 2 h 3 b |
7 1 1 <-----|
7 2 1 2 <-----|
7 h 5 r
So here is my Problem: I want the IDs from Table1 that have all Code and Attributes from Table2 that a given (variable) input ID has. At the end I want to create a stored procedure/function that gives me all IDs meeting that condition.
As an Example: Input-ID: 3. Would return ID 6 because Number 7 (mapped from ID 6 in Table1) has the rows Number 1 (mapped from ID 3 in Table1) has. It has more but that doesn't matter, its just important it has all rows the input one has.
(I can't find a solution to comparing a set of rows to another set of rows that is not known before.)
Thanks for any help!
Edit: To make it more understandible, here what I want in words step-by-step.
Map input ID to Number in Table1
Get All Rows from Table2 having Number from Step 1
Get all Number that have the same (can have more) Rows as from Step 2
Get IDs for that Numbers (and return them)
Upvotes: 0
Views: 120
Reputation: 350
select t11.ID as Id_To_Find,t12.ID as Id_Found
from Table1 t11
join (
select t21.Number as Found,t22.Number as ToFind from Table2 t21
left join Table2 t22 on t21.Code = t22.Code
and t21.Att1 = t22.Att1
and t21.Att2 = t22.Att2
and t21.Att3 = t22.Att3
and t21.Number <> t22.Number
group by t21.Number,t22.Number
having COUNT(*) = (select COUNT(*) from Table2 where Number = t22.Number))
as FindMatches
on t11.Number = FindMatches.ToFind
join Table1 t12 on t12.Number = FindMatches.Found
Upvotes: 1
Reputation: 1218
Try something like this. Haven't tested it, but basically you inner join on all of the attributes that need to match. The HAVING clause is a crude check to make sure that it matched all the rows. Edit: Forgot to add the input ID WHERE clause.
SELECT t1b.ID FROM
Table1 t1a
INNER JOIN Table2 t2a ON t1a.Number = t2a.Number
INNER JOIN Table2 t2b ON t2a.Number <> t2b.Number AND t2a.Code = t2b.CODE AND t2a.Att1 = t2b.Att1 AND t2a.Att2 = t2b.Att2 AND t2a.Att3 = t2b.Att3
INNER JOIN Table1 t1b ON t1b.Number = t2b.Number
WHERE t1a.ID = 3
GROUP BY t1b.ID
HAVING COUNT(*) = (SELECT COUNT(*) FROM Table1 WHERE ID = t1a.ID)
Upvotes: 2
Reputation: 849
Kind of hard to understand what you're trying to acheive. As i understood from your example, you want to match the Number
for the input ID
in Table1
with any column (correct?) in Table2
.
With input ID=3
, the SELECT
will return Number=7
. In the IN (...)
-condition, you can specify whichever columns in Table2
you want to match to Table1.Number
.
DECLARE @Input INT = 3 -- Your input
SELECT DISTINCT t1.Number
FROM Table1 t
INNER JOIN Table2 t2 ON t.Number IN (t2.Number, t2.Code, t2.Att1, t2.Att2, t2.Att3)
INNER JOIN Table1 t1 ON t2.Number = t1.Number AND t.ID <> t1.ID
WHERE t.ID = @Input
Upvotes: 0