Updating earliest record in SQL? - sql

Question

I have a problem at the moment where I would like to update the earliest entry for a job in my table called Service_Log_Temp with duplicate = 'No' and the rest with duplicate = 'Yes'
Here is my table at the moment:
So I would like a script which looks at the K_B_Job_No (you'll see I have 2 different jobs listed here) and updates the earliest entry based on L1_GA_SR_Creation_Date & L2_B_Start_time_Of_Activity with Duplicate = 'No' and the rest with Duplicate = 'Yes'.
This is how it should look:
Any ideas?

Answer 1

try this
select key_service,k_r_job_no,L1_GA_SR_Creation_Date,L2_B_Start_time_Of_Activity,
case
when (select min(L1_GA_SR_Creation_Date) and min(L2_B_Start_time_Of_Activity)
then 'No'
else 'Yes'
end as Duplicate
from Service_Log_Temp
group by k_r_job_no

Answer 2

;with a as (select K_B_Job_No, min(L1_GA_SR_Creation_Date + cast(L2_B_Start_time_Of_Activity as datetime)) as dttime from Service_Log_Temp
group by K_B_Job_No)
select key_service,K_B_Job_No,L1_GA_SR_Creation_Date,L2_B_Start_time_Of_Activity,
(case when
(select dttime from a where a.K_B_Job_No=b.K_B_Job_No)=(L1_GA_SR_Creation_Date + cast(L2_B_Start_time_Of_Activity as datetime))
then 'NO' else 'YES' end) as Duplicate
from Service_Log_Temp b

Answer 3

this should work in postgres, where the concatenation is || - you must change that to whatever is the concatenation in your system
background: To find the time for the first run of each job :
select k_r_job_no,min( L1_GA_SR_Creation_Date+L2_B_Start_time_Of_Activity) from
service_log_temp
group by k_r_job_no
The problem is just that we cannot get the ids out from this query...
but we can use this as a subquery. As it is normally not possible to get more than one row out of a subquery, we have to do a concatenation to get the right time for each job: (you may need to do a type cast to get the concatenation to work)
select key_service from service_log_temp where
k_r_job_no||L1_GA_SR_Creation_Date+L2_B_Start_time_Of_Activity in(
select k_r_job_no||min( L1_GA_SR_Creation_Date+L2_B_Start_time_Of_Activity)
from service_log_temp
group by k_r_job_no
)
This can be rewritten as an update:
update service_log_temp set Duplicate='Yes' where
k_r_job_no||L1_GA_SR_Creation_Date+L2_B_Start_time_Of_Activity in(
select k_r_job_no||min( L1_GA_SR_Creation_Date+L2_B_Start_time_Of_Activity)
from service_log_temp
group by k_r_job_no
)
And then afterwards
update service_log_temp set Duplicate='No' where Duplicate is null
(or, use "case" as SRIRAM proposed if you want to have it in one query)