[sql] Can I use CASE statement in a JOIN condition?

The following image is a part of Microsoft SQL Server 2008 R2 System Views. From the image we can see that the relationship between sys.partitions and sys.allocation_units depends on the value of sys.allocation_units.type. So to join them together I would write something similar to this:

SELECT  *
FROM    sys.indexes i
        JOIN sys.partitions p
            ON i.index_id = p.index_id 
        JOIN sys.allocation_units a
            ON CASE
               WHEN a.type IN (1, 3)
                   THEN a.container_id = p.hobt_id 
               WHEN a.type IN (2)
                   THEN a.container_id = p.partition_id
               END 

But the upper code gives a syntax error. I guess that's because of the CASE statement. Can anyone help to explain a little?


Add error message:

Msg 102, Level 15, State 1, Line 6 Incorrect syntax near '='.

this is the image

This question is related to sql sql-server join case

The answer is


Instead, you simply JOIN to both tables, and in your SELECT clause, return data from the one that matches:

I suggest you to go through this link Conditional Joins in SQL Server and T-SQL Case Statement in a JOIN ON Clause

e.g.

    SELECT  *
FROM    sys.indexes i
        JOIN sys.partitions p
            ON i.index_id = p.index_id 
        JOIN sys.allocation_units a
            ON a.container_id =
            CASE
               WHEN a.type IN (1, 3)
                   THEN  p.hobt_id 
               WHEN a.type IN (2)
                   THEN p.partition_id
               END 

Edit: As per comments.

You can not specify the join condition as you are doing.. Check the query above that have no error. I have take out the common column up and the right column value will be evaluated on condition.


Took DonkeyKong's example.

The issue is I needed to use a declared variable. This allows for stating your left and right-hand side of what you need to compare. This is for supporting an SSRS report where different fields must be linked based on the selection by the user.

The initial case sets the field choice based on the selection and then I can set the field I need to match on for the join.

A second case statement could be added for the right-hand side if the variable is needed to choose from different fields

LEFT OUTER JOIN Dashboard_Group_Level_Matching ON
       case
         when @Level  = 'lvl1' then  cw.Lvl1
         when @Level  = 'lvl2' then  cw.Lvl2
         when @Level  = 'lvl3' then  cw.Lvl3
       end
    = Dashboard_Group_Level_Matching.Dashboard_Level_Name

Yes, you can. Here is an example.

SELECT a.*
FROM TableA a
LEFT OUTER JOIN TableB j1 ON  (CASE WHEN LEN(COALESCE(a.NoBatiment, '')) = 3 
                                THEN RTRIM(a.NoBatiment) + '0' 
                                ELSE a.NoBatiment END ) = j1.ColumnName 

Here I have compared the difference in two different result sets:

SELECT main.ColumnName, compare.Value PreviousValue,  main.Value CurrentValue
FROM 
(
    SELECT 'Name' AS ColumnName, 'John' as Value UNION ALL
    SELECT 'UserName' AS ColumnName, 'jh001' as Value UNION ALL
    SELECT 'Department' AS ColumnName, 'HR' as Value UNION ALL
    SELECT 'Phone' AS ColumnName, NULL as Value UNION ALL
    SELECT 'DOB' AS ColumnName, '1993-01-01' as Value UNION ALL
    SELECT 'CreateDate' AS ColumnName, '2017-01-01' as Value UNION ALL
    SELECT 'IsActive' AS ColumnName, '1' as Value
) main
INNER JOIN
(
    SELECT 'Name' AS ColumnName, 'Rahul' as Value UNION ALL
    SELECT 'UserName' AS ColumnName, 'rh001' as Value UNION ALL
    SELECT 'Department' AS ColumnName, 'HR' as Value UNION ALL
    SELECT 'Phone' AS ColumnName, '01722112233' as Value UNION ALL
    SELECT 'DOB' AS ColumnName, '1993-01-01' as Value UNION ALL
    SELECT 'CreateDate' AS ColumnName, '2017-01-01' as Value UNION ALL
    SELECT 'IsActive' AS ColumnName, '1' as Value
) compare
ON main.ColumnName = compare.ColumnName AND
CASE 
    WHEN main.Value IS NULL AND compare.Value IS NULL THEN 0
    WHEN main.Value IS NULL AND compare.Value IS NOT NULL THEN 1
    WHEN main.Value IS NOT NULL AND compare.Value IS NULL THEN 1
    WHEN main.Value <> compare.Value THEN 1
END = 1 

Try this:

...JOIN sys.allocation_units a ON 
  (a.type=2 AND a.container_id = p.partition_id)
  OR (a.type IN (1, 3) AND a.container_id = p.hobt_id)

This seems nice

https://bytes.com/topic/sql-server/answers/881862-joining-different-tables-based-condition

FROM YourMainTable
LEFT JOIN AirportCity DepCity ON @TravelType = 'A' and DepFrom =  DepCity.Code
LEFT JOIN AirportCity DepCity ON @TravelType = 'B' and SomeOtherColumn = SomeOtherColumnFromSomeOtherTable

I took your example and edited it:

SELECT  *
FROM    sys.indexes i
    JOIN sys.partitions p
        ON i.index_id = p.index_id 
    JOIN sys.allocation_units a
        ON a.container_id = (CASE
           WHEN a.type IN (1, 3)
               THEN p.hobt_id 
           WHEN a.type IN (2)
               THEN p.partition_id
           ELSE NULL
           END)

I think you need two case statements:

SELECT  *
FROM    sys.indexes i
    JOIN sys.partitions p
        ON i.index_id = p.index_id 
    JOIN sys.allocation_units a
        ON 
        -- left side of join on statement
            CASE
               WHEN a.type IN (1, 3)
                   THEN a.container_id
               WHEN a.type IN (2)
                   THEN a.container_id
            END 
        = 
        -- right side of join on statement
            CASE
               WHEN a.type IN (1, 3)
                   THEN p.hobt_id
               WHEN a.type IN (2)
                   THEN p.partition_id
            END             

This is because:

  • the CASE statement returns a single value at the END
  • the ON statement compares two values
  • your CASE statement was doing the comparison inside of the CASE statement. I would guess that if you put your CASE statement in your SELECT you would get a boolean '1' or '0' indicating whether the CASE statement evaluated to True or False

Examples related to sql

Passing multiple values for same variable in stored procedure SQL permissions for roles Generic XSLT Search and Replace template Access And/Or exclusions Pyspark: Filter dataframe based on multiple conditions Subtracting 1 day from a timestamp date PYODBC--Data source name not found and no default driver specified select rows in sql with latest date for each ID repeated multiple times ALTER TABLE DROP COLUMN failed because one or more objects access this column Create Local SQL Server database

Examples related to sql-server

Passing multiple values for same variable in stored procedure SQL permissions for roles Count the Number of Tables in a SQL Server Database Visual Studio 2017 does not have Business Intelligence Integration Services/Projects ALTER TABLE DROP COLUMN failed because one or more objects access this column Create Local SQL Server database How to create temp table using Create statement in SQL Server? SQL Query Where Date = Today Minus 7 Days How do I pass a list as a parameter in a stored procedure? SQL Server date format yyyymmdd

Examples related to join

Pandas Merging 101 pandas: merge (join) two data frames on multiple columns How to use the COLLATE in a JOIN in SQL Server? How to join multiple collections with $lookup in mongodb How to join on multiple columns in Pyspark? Pandas join issue: columns overlap but no suffix specified MySQL select rows where left join is null How to return rows from left table not found in right table? Why do multiple-table joins produce duplicate rows? pandas three-way joining multiple dataframes on columns

Examples related to case

PostgreSQL CASE ... END with multiple conditions SQL Server: use CASE with LIKE SQL Server IIF vs CASE SELECT using 'CASE' in SQL SELECT query with CASE condition and SUM() GROUP BY + CASE statement SQL use CASE statement in WHERE IN clause SQL Server - Case Statement where to place CASE WHEN column IS NULL in this query Regarding Java switch statements - using return and omitting breaks in each case