Question

I have two tables emp_master and transfer. for each employee_id
in emp_master there can be multiple entries in transfer table. I want to get the latest
record from the transfer table based on date_of_tansfer and display it with data from
emp_master. I have written a query like this which is not working. I am getting a error
“EMP.EMPLOYEE_ID: invalid identifier”. I can get the details for single employee if i hardcode
it with that employee_id.. but how can i modify it to get data for all the employees..

select distinct emp.employee_id,trnsf.OU
from emp_master emp 
left join (select * from (select * from transfer where employee_id = emp.employee_id 
and date_of_transfer <= SYSDATE order by date_of_transfer desc) where rownum = 1) trnsfr
on trnsfr.EMPLOYEE_ID = emp.employee_id


CREATE TABLE "EMP_MASTER" 
   ( "EMPLOYEE_ID" NUMBER(10,0), 
     "FIRST_NAME" VARCHAR2(30 BYTE), 
     "MIDDLE_NAME" VARCHAR2(30 BYTE), 
     "SURNAME" VARCHAR2(30 BYTE), 
   ) 

 CREATE TABLE "TRANSFER" 
   (    "EMPLOYEE_SR_NO" NUMBER, 
        "EMPLOYEE_ID" NUMBER(10,0), 
        "DATE_OF_TRANSFER" DATE, 
        "OU" VARCHAR2(30 BYTE)
 ) 

Answer 1

The reason you are getting that error is the nesting level of your derived tables (sub-queries). A sub-query can “see” only the columns/tables from the immediate “parent” query (which in your case is the outer select that limits the result to one row).

The following should do what you want:

select distinct emp.employee_id, trnsfr.ou
from emp_master emp 
left join (
      select transfer.employee_id,
             transfer.ou,
             row_number() over (order by date_of_transfer desc) as rn
      from transfer 
      where date_of_transfer <= SYSDATE 
    ) trnsfr
    on trnsfr.employee_id = emp.employee_id and trnsfr.rn = 1;

If date_of_transfer is really a “DATE” (i.e. no time involved) than you might consider using date_of_transfer <= trunc(SYSDATE) instead in order to “remove” the time from the SYSDATE result.