Join table using columns which contains other table's id's separated by commas - sql

Question

I have FACULTY table which contain the column with other table's id's separated by commas. I want to join those with respective table.
faculty table:
id | name | course_id | subject_id
a | smith | 2,3 | 1,2
course table:
id | name
1 | bcom
2 | mcom
3 | bba
subject table:
id | name
1 | account
2 | state
3 | economics
I want to get result from these table like..
faculty.id, faculty.name, course.name(using faculty.course_id), subject.name(using faculty.subject_id)
I have tried a lot of queries and also finds from Google but it didn't gave me proper result.

Answer 1

I do not think the performance will be too nice but worths trying. This solution would work in SQL SERVER:
SELECT *
FROM faculty F
JOIN course C
ON ','+F.course_id+',' LIKE '%,'+CONVERT(VARCHAR,C.ID) +',%'
JOIN subject S
ON ','+F.subject_id_id+',' LIKE '%,'+CONVERT(VARCHAR,S.ID) +',%'
Based on Albin Sunnanbo's comment i would also sugget you add some many too many tables:
fcourses
facultyId
courseId
and
fsubjects
facultyId
subjectId
That way you could do a proper join :
SELECT *
FROM faculty F
JOIN fcourses FC
ON F.Id = FC.facultyId
JOIN course C
ON FC.courseId = C.ID
JOIN fsubjects FS
ON F.Id = FS.facultyId
JOIN subject S
ON FS.courseId = S.ID

Answer 2

You can do the following query
select * from faculty F
JOIN course C
on CHARINDEX((','+CAST(c.id as varchar(10))+','), (','+f.courseid+',')) > 0
JOIN subject s
on CHARINDEX((','+CAST(s.id as varchar(10))+','), (','+f.subjectid+',')) > 0

Answer 3

I've done something similar like this:
select f.id, f.lname, f.fname, U.useridlist
from TABLE1 F, TABLE2 U
where ',' || U.useridlist || ',' like '%,' || f.id || ',%'

Answer 4

If you can create a 'string to int table' function, I would look at the following:
Create the function
CREATE FUNCTION [dbo].[udf_ConvertIntListToTable] (#list varchar(MAX))
RETURNS #tbl TABLE (val int) AS
BEGIN
DECLARE #ix int,
#pos int,
#str varchar(MAX),
#num int
SET #pos = 1
SET #ix = 1
WHILE #ix > 0
BEGIN
SET #ix = charindex(',', #list, #pos)
IF #ix > 0
SET #str = substring(#list, #pos, #ix - #pos)
ELSE
SET #str = substring(#list, #pos, len(#list))
SET #str = ltrim(rtrim(#str))
IF #str LIKE '%[0-9]%' AND
(#str NOT LIKE '%[^0-9]%' OR
#str LIKE '[-+]%' AND
substring(#str, 2, len(#str)) NOT LIKE '[-+]%[^0-9]%')
BEGIN
SET #num = convert(int, #str)
INSERT #tbl (val) VALUES(#num)
END
SET #pos = #ix + 1
END
RETURN
END
Then query using CROSS APPLY
declare #FacultyTable table(id int PRIMARY KEY, name nvarchar(50), course_id varchar(50))
declare #CourseTable table(id int PRIMARY KEY, name nvarchar(50))
insert into #FacultyTable values(1, 'Peter Sagal', '11,22')
insert into #FacultyTable values(2, 'Carl Kasell', '22,33')
insert into #CourseTable values(11,'News')
insert into #CourseTable values(22,'News')
insert into #CourseTable values(33,'News')
insert into #CourseTable values(44,'News')
select *
from #FacultyTable f
CROSS APPLY(
SELECT *
FROM #CourseTable c
WHERE
c.id IN (SELECT * FROM dbo.udf_ConvertIntListToTable(f.course_id))
) tCourses