Match any substring of 4 consecutive characters - sql-server

Question

I'm looking for a join expression for matching strings from two different tables which both contain the same sub-string of 4 consective characters.
For example, the following should match:
String1 String2
-------- -----------
xxjohnyy abcjohnabc [common substring: "john"]
xxjohnyy johnny [common substring: "john"]
birdsings ravenbird [common substring: "bird"]
singbird a singer [common substring: "sing"]

Answer 1

This problem is very similar to finding the Longest Common Substring problem. You find the Longest Common Substring and then you pick those with common strings of 4. You will definitely find this link and this link helpful for you.

Answer 2

This is a very good exercise. Here is my attempt using Tally Table.
SQL Fiddle
;WITH E1(N) AS(
SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL
SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1 UNION ALL SELECT 1
),
E2(N) AS(SELECT 1 FROM E1 a CROSS JOIN E1 b),
E4(N) AS(SELECT 1 FROM E2 a CROSS JOIN E2 b),
E8(N) AS(SELECT 1 FROM E4 a CROSS JOIN E4 b),
Tally(N) AS(
SELECT TOP (
SELECT
CASE
WHEN MAX(LEN(String1)) > MAX(LEN(String2)) THEN MAX(LEN(String1))
ELSE MAX(LEN(String2))
END
FROM TestTable
)
ROW_NUMBER() OVER(ORDER BY (SELECT NULL))
FROM E8
),
CteTable AS( -- Added an ID to uniquely identify each row
SELECT *, Id = ROW_NUMBER() OVER(ORDER BY (SELECT NULL)) FROM TestTable
),
CteSubStr1 AS(
SELECT
ct.*,
substr = SUBSTRING(ct.String1, t.N, 4)
FROM CteTable ct
CROSS APPLY(
SELECT N FROM Tally
WHERE N <= LEN(ct.String1) - 3
)t
),
CteSubStr2 AS(
SELECT
ct.*,
substr = SUBSTRING(ct.String2, t.N, 4)
FROM CteTable ct
CROSS APPLY(
SELECT N FROM Tally
WHERE N <= LEN(ct.String2) - 3
)t
),
CteCommon AS(
SELECT * FROM CteSubStr1 c1
WHERE EXISTS(
SELECT 1 FROM CteSubStr2
WHERE
Id = c1.Id
AND substr = c1.substr
)
)
SELECT
String1, String2, substr
FROM (
SELECT *, RN = ROW_NUMBER() OVER(PARTITION BY Id ORDER BY LEN(substr) DESC)
FROM CteCommon
)t
WHERE RN = 1
Result
| String1 | String2 | substr |
|-----------|------------|--------|
| xxjohnyy | abcjohnabc | john |
| xxjohnyy | johnny | john |
| birdsings | ravenbird | bird |
| singbird | a singer | sing |
This part looks for the longest common substring.
SELECT
String1, String2, substr
FROM (
SELECT *, RN = ROW_NUMBER() OVER(PARTITION BY Id ORDER BY LEN(substr) DESC)
FROM CteCommon
)t
WHERE RN = 1
To get all the common substrings, use this instead:
SELECT * FROM CteCommon

Answer 3

;with pos as(select 1 as p
union all
select p + 1 from pos where p < 100),
uni as(select *, row_number() over(order by (select null)) id from t)
select t1.s1, t1.s2, ca.s
from uni t1
cross apply(select substring(t2.s2, p, 4) s
from uni t2
cross join pos
where t1.id = t2.id and
len(substring(t2.s2, p, 4)) = 4 and
t1.s1 like '%' + substring(t2.s2, p, 4) + '%')ca
Fiddlee http://sqlfiddle.com/#!3/bd4dd/16
Just change 100 to actual length of your columns...