Search code examples
sqlt-sqlsql-server-2008output-clause

Output Clause: The multi-part identifier could not be bound


I am running in to the dreaded "The multi-part identifier could not be bound" error on a stored procedure I am currently working on. I have a few questions in regards to the query below.

  1. Why am I getting this error?
  2. Why would this error occur on ImportFundingDateTime instead of FloorplanId given that they both come from the same query, but FloorplanId is listed first in the output clause?
  3. Can I adjust this query to not get the error while still keeping the general structure the same?

.

DECLARE @Results                Table(
    [FloorPlanId]               UNIQUEIDENTIFIER,
    [ImportFundingDateTime]     DATETIME,
    [TimeStamp]                 VARBINARY(8),
    [BusinessId]                UNIQUEIDENTIFIER
    )

UPDATE CacRecord 
    SET MatchFound = 1
    OUTPUT  fp.[FloorplanId], cr.[ImportFundingDateTime],
            fp.[TimeStamp], buyer.[BusinessId]
    INTO @Results(  [FloorplanId], [ImportFundingDateTime], 
                    [TimeStamp], [BusinessId])
    FROM CacRecord cr WITH (NOLOCK)
    INNER JOIN CacBatch cb WITH (NOLOCK)
        ON cr.CacBatchId = cb.CacBatchId
    INNER JOIN Floorplan fp WITH (NOLOCK)
        ON fp.UnitVIN = cr.ImportVin
        AND COALESCE(fp.UnitVIN, '') <> ''
    INNER JOIN Business buyer WITH (NOLOCK)
        ON buyer.BusinessId = fp.BuyerBusinessId
    LEFT OUTER JOIN BusinessContact bc WITH (NOLOCK)
        ON bc.BusinessId = buyer.BusinessId
    LEFT OUTER JOIN Contact c WITH (NOLOCK)
        ON c.ContactId = bc.ContactId
    WHERE cb.CacJobInstanceId = @cacJobInstanceId
        AND fp.FloorplanStatusId = 1 --Approved
        AND COALESCE(cr.ImportVin, '') <> ''
        AND 1 = 
            CASE
                WHEN cr.ImportFein = buyer.FederalTaxID  
                    AND COALESCE(cr.ImportFein, '') <> '' THEN 1
                WHEN cr.ImportSsn = c.Ssn 
                    AND COALESCE(cr.ImportSsn, '') <> '' THEN 1
                ELSE 0
            END;

Solution

  • Please recheck the syntax of the OUTPUT clause OUTPUT on MSDN

    Syntax
    
    <column_name> ::=
    { DELETED | INSERTED | from_table_name } . { * | column_name }
    
    from_table_name
    
    Is a column prefix that specifies a table included in the FROM clause
    of a DELETE or UPDATE statement that is used tospecify the rows to
    update or delete.
    

    It looks like you have aliased CacRecord in the FROM clause as "cr", but have not correlated that with the UPDATE clause.

    Note: Even with it aliases in the FROM clause and NOT aliased in the UPDATE cause, SQL Server appears to recognize CacRecord as the UPDATE table, requiring you to use INSERTED instead of cr as the virtual table name.

    UPDATE cr
    SET MatchFound = 1
    OUTPUT fp.[FloorplanId], INSERTED.[ImportFundingDateTime],
      fp.[TimeStamp], buyer.[BusinessId]
    INTO @Results( [FloorplanId], [ImportFundingDateTime], 
        [TimeStamp], [BusinessId])
    FROM CacRecord cr WITH (NOLOCK)
    INNER JOIN CacBatch cb WITH (NOLOCK)
     ON cr.CacBatchId = cb.CacBatchId
    INNER JOIN Floorplan fp WITH (NOLOCK)
     ON fp.UnitVIN = cr.ImportVin
     AND COALESCE(fp.UnitVIN, '') <> ''
    INNER JOIN Business buyer WITH (NOLOCK)
     ON buyer.BusinessId = fp.BuyerBusinessId
    LEFT OUTER JOIN BusinessContact bc WITH (NOLOCK)
     ON bc.BusinessId = buyer.BusinessId
    LEFT OUTER JOIN Contact c WITH (NOLOCK)
     ON c.ContactId = bc.ContactId
    WHERE cb.CacJobInstanceId = @cacJobInstanceId
     AND fp.FloorplanStatusId = 1 --Approved
     AND COALESCE(cr.ImportVin, '') <> ''
     AND 1 = 
      CASE
       WHEN cr.ImportFein = buyer.FederalTaxID  
        AND COALESCE(cr.ImportFein, '') <> '' THEN 1
       WHEN cr.ImportSsn = c.Ssn 
        AND COALESCE(cr.ImportSsn, '') <> '' THEN 1
       ELSE 0
      END;
    

    For visitors to this question, this code block shows multiple tables being referenced in the OUTPUT clause correctly.

    create table TO1 (id int, a int);
    create table TO2 (id int, b int);
    create table TO3 (id int, c int);
    insert into TO1 select 1,1;
    insert into TO2 select 1,2;
    insert into TO3 select 1,3;
    insert into TO3 select 1,4;
    
    declare @catch table (a int, b int, c int)
    update c
    set c = a.a
    output a.a, b.b, INSERTED.c
    into @catch(a,b,c)
    from TO1 a
    inner join TO2 b on a.id=b.id
    inner join TO3 c on a.id=c.id