SQL Server: Compare two columns in two tables -


i have 2 tables (from 2 different systems) track employees' hours. in both tables, employees enter date , hours. need create audit report shows discrepancy. report needs show columns , display null/no match if there mismatch. 1 table might have more/less entries other table or duplicate entries , need catch (two entries on same day same amount of hours in 1 table). both tables have userid can joined on.

if there match based on date , hours, show values. if there mismatch based on hours, show null or no match when there mismatch. if there duplicate entry, image below, match first entry , report second 1 null or no match.

i tried joining tables based userid, date, , hours not able tell mismatch came from.

table a:

enter image description here

table b:

enter image description here

left join on userid, date, , hours

enter image description here

set nocount on;  declare @table1 table (     userid int,     entry_date date,     [hours] varchar(10) )  declare @table2 table (     userid int,     entry_date datetime,     [hours] varchar(10) )  insert @table1 select 1,'8/14/2017','10:00' insert @table1 select 2,'8/14/2017','5:00' insert @table1 select 2,'8/14/2017','5:00' insert @table1 select 2,'8/14/2017','5:00' insert @table1 select 2,'8/14/2017','5:00' insert @table1 select 3,'8/14/2017','5:00' insert @table1 select 3,'8/14/2017','6:00' insert @table1 select 3,'8/14/2017','6:00' insert @table1 select 3,'8/14/2017','6:00'  insert @table2 select 1,'8/14/2017','10:00' insert @table2 select 2,'8/14/2017','8:00' insert @table2 select 3,'8/14/2017','6:00' insert @table2 select 4,'8/14/2017','2:00' insert @table2 select 1,'8/14/2017','10:00' insert @table2 select 3,'8/14/2017','6:00'  ;with cte_table1 (     select  t.userid userid, cast(t.entry_date date) entry_date,   t.[hours] [hours],              row_number() over(partition t.userid, t.entry_date, t.[hours] order t.[hours]) rnk     @table1 t ), cte_table2 (     select  t.userid userid, cast(t.entry_date date) entry_date,   t.[hours] [hours],              row_number() over(partition t.userid, t.entry_date, t.[hours] order t.[hours]) rnk     @table2 t ), cte_matches (     select  t1.userid userid, t1.entry_date entry_date, t1.[hours] [hours], t1.rnk     cte_table1 t1     inner join cte_table2 t2         on (t1.userid = t2.userid , t1.entry_date = t2.entry_date , t1.[hours] = t2.[hours] , t1.rnk = t2.rnk) ),cte_match_duplicates (     select 'table1matchduplicate' errortype, *         (         select t.*         (select userid, entry_date, [hours], max(rnk) rnk cte_matches group userid, entry_date, [hours]) m         inner join cte_table1 t             on (t.userid = m.userid , t.entry_date = m.entry_date , t.[hours] = m.[hours] , t.rnk > m.rnk)     )q     union     select 'table2matchduplicate' errortype, *         (         select t.*          (select userid, entry_date, [hours], max(rnk) rnk cte_matches group userid, entry_date, [hours]) m         inner join cte_table2 t             on (t.userid = m.userid , t.entry_date = m.entry_date , t.[hours] = m.[hours] , t.rnk > m.rnk)     )q ) , cte_table1_unmatched (        select t.userid, t.entry_date, t.[hours]      @table1 t     left outer join cte_matches m         on (t.userid = m.userid , cast(t.entry_date date) = m.entry_date , t.[hours] = m.[hours])     m.userid null ), cte_table2_unmatched (        select t.userid, t.entry_date, t.[hours]      @table2 t     left outer join cte_matches m         on (t.userid = m.userid , cast(t.entry_date date) = m.entry_date , t.[hours] = m.[hours])     m.userid null ) select null errortype, userid, entry_date, [hours] cte_matches union select 'table1mismatch' errortype, userid, entry_date, [hours] cte_table1_unmatched union select 'table2mismatch' errortype, userid, entry_date, [hours] cte_table2_unmatched union select errortype, userid, entry_date, [hours] cte_match_duplicates order errortype 

http://rextester.com/udg95824

if need find duplicates had no matches well:

,cte_table1_unmatched_duplicates (     select userid, entry_date, [hours]     cte_table1_unmatched     group userid, entry_date, [hours]     having count(*) > 1 ),cte_table2_unmatched_duplicates (     select userid, entry_date, [hours]     cte_table2_unmatched     group userid, entry_date, [hours]     having count(*) > 1 ) ... union select 'table1unmatchedduplicates' errortype, userid, entry_date, [hours] cte_table1_unmatched_duplicates union select 'table2unmatchedduplicates' errortype, userid, entry_date, [hours] cte_table2_unmatched_duplicates 

http://rextester.com/kejjf79330


Comments

Popular posts from this blog

python Tkinter Capturing keyboard events save as one single string -

android - InAppBilling registering BroadcastReceiver in AndroidManifest -

javascript - Z-index in d3.js -