Simple Query : Unable to execute

Question

Hi ALL,
I need wnat to extract records from a table where its ID records doesnt match with other Tbale ID values..
TABLEA
ID,NAME
1,A
2,B
TABLEB
ID,AGE
1,60
Now in SQL I can write query as
select TABLEA.ID from TABLEA,TABLEB where TABLEA.ID<>TABLEB.ID
and I can get out as
ID
2
How can I get this in DS??? I am struggleing.. may be its simple ....can any one throw some light?
Thanks
Rajeev

Answer 1

Sorry, I misread your issue and thought you wanted the records where the IDs DID match.
Ok, I know of two options, both quite simple, and both require two steps (two queries):
Option 1) Outer join Tables A and B in a query, with Table A as outer table and B as inner, with join condition A.ID=B.ID.  Then map across the fields you want for your target plus the ID from table B, but rename it on the output schema as 'ID_B'.  Now because this is an outer join the result will be all records from table A with the ID_B value being populated if there was a matching ID in Table B, and if not then it will be NULL. So then link this query output to a second query and just use the filter WHERE ID_B IS NULL (and you don't need to map the ID_B column to the output as you don't need it anymore).  So the second query's output is all the records from Table A that did not have a matching ID in Table B.
If your volume of data is high then I only recommend this option if both tables are in the same datastore (i.e. if the outer join can be pushed down to the database).
Option 2) In the first query link only Table A as the source, then map across all the columns from table A and then create a new function output column called 'ID_B' with a lookup or lookup_ext function which simply performs a lookup of the Table A ID in table B, returning the ID from table B if the lookup finds a match, otherwise NULL.  Then similarly in a second query you filter the records where ID_B value IS NULL (i.e. the lookup DID NOT find a match).
This option is probably going to perform better for high data volumes because Table B can be cached if it can fit into available memory, as you probably know you can choose from different caching options in the lookup function. The default option should be fine unless Table B is huge and your memory is low - check the reference guide.