I have two tables that I want to join into one table and use a TypeID to differentiate them. Let's say the types are A and B. The Tables are A_Level and B_Level
A's Table looks like
Level
Level_ID Description
B's Table looks like
Level
Level_ID Level_Desc
A's Level_ID is referenced from Table C as Level_ID
B's Level_ID is referenced from Table D as Level_ID
I am looking for a script that would merge the two tables into one table (Level_Code) and update the referenced Tables ID's accordingly.
Any help is greatly appreciated.
select a.Level_Id ALevelId, b.Level_Id BLevelId,
case ISNULL(a.Level_Id, 0) when 0 then 'B' else 'A' end AS Type,
case ISNULL(a.Level_Id, 0) when 0 then b.Level_Id else a.Level_Id end AS NewLevel_Id
INTO Dummy
FROM a
FULL JOIN b On (a.Level_ID = b.Level_ID);
UPDATE c
SET c.Level_id = Dummy.NewLevel_Id
from Dummy, c
WHERE c.Level_Id = Dummy.ALevelId
AND Dummy.Type = 'A';
UPDATE d
SET d.Level_id = Dummy.NewLevel_Id
from Dummy, d
WHERE d.Level_Id = Dummy.BLevelId
AND Dummy.Type = 'B';
SELECT Dummy.NewLevel_Id, a.Level, a.LevelDesc As Description
INTO YourNewTable
from Dummy JOIN a ON (Dummy.ALevelId = a.Level_Id)
Where Dummy.Type = 'A'
UNION
SELECT NewLevel_Id, Level, LevelDesc As Description
from Dummy JOIN b ON (Dummy.BLevelId = b.Level_Id)
Where Dummy.Type = 'B'
DROP TAble Dummy;
Related
I have a table Product with 240 000 rows.
I want to select from this table data where idproduct = idproductcomponent
Output makes a table with 3 columns A12345678 35655455952625 9638520963258960, so in different column.
Always idproductcomponent of idproducttype 1 is 0, and idproductcomponent of idproducttype 2,3 are the same of idproduct of idproducttype 1.
Pelase, can you share with me any idea for this select ?
Assuming your database is SQL Server (you don't say which one it is) the query could look like:
select
a.name,
b.name,
c.name
from t a
left join t b on b.idproductcomponent = a.idproduct and b.idproducttype = 2
left join t c on c.idproductcomponent = a.idproduct and c.idproducttype = 3
where a.idproducttype = 1
and a.idproduct = 11163 -- parameter you are searching for
To increase the performance of this query you can add the index:
create index ix1 on t (idproductcomponent);
I have tables:
Result containing 5 columns: result_id, num_1, num_2, num_3, num_4
Ref containing 4 columns: num_1, num_2, num_3, num_4
Columns num contain random int in range of 1-9
Aim of exercise is to display all result_id from Result table which have num values combination present in Ref table and to display result_id which have not met combination criteria.
I've been trying left joining ref to result, but unfortunately no success. Could you please share some light how to deal with it?
If you want the result_id for which combination exists in the ref table then use following JOIN query:
select distinct r.result_id
from results r
join ref on r.num_1 = ref.num_1 and r.num_2 = ref.num_2
and r.num_3 = ref.num_3 and r.num_4 = ref.num_4
If you want the result_id for which combination do not exists in REF table then use the LEFT JOIN as follows:
select r.result_id
from results r
left join ref on r.num_1 = ref.num_1 and r.num_2 = ref.num_2
and r.num_3 = ref.num_3 and r.num_4 = ref.num_4
where ref.num_1 is null -- or use PK / Not nullable column of REF table here
Assuming you want the columns to "line up" and you want to add a flag to the result_id in the first table, then use exists:
select t1.*,
(case when exists (select 1
from table2 t2
where t2.n1 = t1.n1 and t2.n2 = t1.n2 and t2.n3 = t1.n3 and t2.n4
)
then 'present' else 'not present'
end) as flag
from t2;
SELECT A.GRPNO, A.EMPNO, A.DEPNO, A.PENDCD FROM EMPDEP A, EEDPELIG B
WHERE A.GRPNO=B.GRPNO
AND A.EMPNO=B.EMPNO
AND A.DEPNO=B.DEPNO
AND A.GRPNO = 6606 AND A.SPOUSE = 'T'
AND B.ELIGFLAG01 = 'T' AND SNAPTHRUDT ='DEC312999'
Our selection statement has been successful at pulling the information we need however we're new with SQL and are struggling to create an update statement that is replacing the "a.pendcd=0" to "a.pendcd=20" from the information in the select statement. Any help is appreciated, thank you.
update a
a.pendcd=20
FROM EMPDEP A inner join EEDPELIG B
on A.GRPNO=B.GRPNO
AND A.EMPNO=B.EMPNO
AND A.DEPNO=B.DEPNO
AND A.GRPNO = 6606 AND A.SPOUSE = 'T'
AND B.ELIGFLAG01 = 'T' AND SNAPTHRUDT ='DEC312999'
where a.pendcd=0
Oracle does not support FROM or JOIN in UPDATE (under most circumstances).
Just use EXISTS:
UPDATE EMPDEP ed
SET . . .
WHERE EXISTS (SELECT 1
FROM EEDPELIG p
WHERE ed.GRPNO = p.GRPNO AND
ed.EMPNO= p.EMPNO AND
ed.DEPNO= p.DEPNO AND
p.ELIGFLAG01 = 'T'
)
ed.GRPNO = 6606 AND
ed.SPOUSE = 'T' AND
ed.SNAPTHRUDT ='DEC312999';
It is unclear if the condition on SNAPTHRUDT is on the outer table or inner table. If it is on p, then move it to the subquery.
You can use MERGE statement as following:
Lets assume EMPDEP table has primary key which is EMPDEP_UID.
MERGE INTO EMPDEP TRG
USING
(SELECT A.EMPDEP_UID, A.PENDCD
FROM EMPDEP A, EEDPELIG B
WHERE A.GRPNO=B.GRPNO
AND A.EMPNO=B.EMPNO
AND A.DEPNO=B.DEPNO
AND A.GRPNO = 6606
AND A.SPOUSE = 'T'
AND B.ELIGFLAG01 = 'T'
AND SNAPTHRUDT ='DEC312999') SRC
ON (TRG.EMPDEP_UID = SRC.EMPDEP_UID)
WHEN MATCHED THEN
UPDATE SET TRG.PENDCD = 0
WHERE TRG.PENCD = 20;
You can use unique keys instead of primary key to identify the records to be updated. But it is safe to use primary key as unique key can contain nulls which can change the behaviour of our query.
Cheers!!
Hello I have two databases and each has the same tables. For example I have table called world and it has 4 columns: pkey1, pkey2,companyid, company_name
I made a query which searches for rows which have the same pkey1 and pkey2 but one or many of their other property is different in the 2 tables. My question is how can I see only the different properties?
Here is my query it selects the rows which have the same pkey1 and pkey2 how can I upgrade it to see the columns where there is difference in both databases and of course if there is no difference the result of the query should return NULL in the column here is an example what I want to achieve:
in first database (1,1,345,'Ron'), second database (1,1,377,'Ron') the result should be (1,1,345,null)
Basically to reference data in a different database, you'll need a database link, since you are using PostgreSQL, this documentation should help;
https://www.postgresql.org/docs/current/static/dblink.html
You'll need to use this command to create the link (in this case with a name);
SELECT dblink_connect('CONNECTIONNAME', 'REMOTEDBCONNECTIONSTRING');
Then you can use this new connection via a select query;
SELECT *
FROM dblink('CONNECTIONNAME','SELECT * FROM foo') AS t(a int, b text, c text[]);
And replace 'foo' with your 'world' table name, then change the AS into the variables that relate to pkey1, pkey2,companyid, company_name etc.
If your databases are linked, you can join both tables and with "case" statement check if the value has changed:
select a.pkey1, a.pkey2,
case when a.companyid <> b.companyid then a.companyid else null end as companyid,
case when a.company_name <> b.company_name then a.company_name else null end as company_name
from db1.dbo.world a
inner join db2.dbo.world b on a.pkey1 = b.pkey1 and a.pkey2 = b.pkey2
If you want to omit rows with no difference you can use "except":
select a.pkey1, a.pkey2,
case when a.companyid <> b.companyid then a.companyid else null end as companyid,
case when a.company_name <> b.company_name then a.company_name else null end as company_name
from (
select pkey1, pkey2, companyid, company_name
from db1.dbo.world
except
select pkey1, pkey2, companyid, company_name
from db2.dbo.world) a
inner join db2.dbo.world b on a.pkey1 = b.pkey1 and a.pkey2 = b.pkey2
I have two tables RecordMaster and Dummy
Both have columns like Mobile_Number and Insert_Date
I want a row like
1) from Dummy table I want to fetch those rows whose Mobile_Number And Insert_Date are same compared to RecordMaster.
2) from Dummy table I want to fetch those rows whose Mobile_Number And Insert_Date are different compared to RecordMaster.
After that in 1) condition I want to fetch only those rows whose Cpv_Status is not null.
(CPV_STATUS) is one column in the Dummy table..
Help me please ........
To meet your 1) and 3) needs ( optionally include the WHERE as you need).
SELECT d.*
FROM Dummy d
INNER JOIN RecordMaster r
ON r.mobile_number = d.mobile_number
AND r.insert_date = d.insert_date
WHERE d.Cpv_Status IS NOT NULL
2.
SELECT d.*
FROM Dummy d
WHERE NOT EXISTS
(SELECT 1
FROM RecordMaster r
WHERE r.mobile_number = d.mobile_number
AND r.insert_date = d.insert_date
)
To insert these:
INSERT INTO RecordMaster(mobile_number, insert_date)
SELECT d.mobile_number, insert_date
FROM Dummy d
WHERE NOT EXISTS
(SELECT 1
FROM RecordMaster r
WHERE r.mobile_number = d.mobile_number
AND r.insert_date = d.insert_date
)
The following query will give you all records in dummy that match the records in RecordMaster
SELECT a.Mobile_Number,a.Insert_Date ,a.Cpv_Status
FROM RecordMaster a, Dummy b
WHERE a.Mobile_Number = b.Mobile_Number and a.Insert_Date = b.Insert_Date
The following query will give you records in Dummy that don't have matching records in RecordMadter
SELECT a.Mobile_Number,a.Insert_Date ,a.Cpv_Status
FROM Dummy a
WHERE STR(a.Mobile_Number)+STR(a.Mobile_Number) not in
(SELECT STR(Mobile_Number)+STR(Insert_Date) FROM RecordMaster)
if you need both of these results combined in one result set, then use UNION like this
SELECT a.Mobile_Number,a.Insert_Date ,a.Cpv_Status
FROM RecordMaster a, Dummy b
WHERE a.Mobile_Number = b.Mobile_Number and a.Insert_Date = b.Insert_Date
UNION
SELECT a.Mobile_Number,a.Insert_Date ,a.Cpv_Status
FROM Dummy a
WHERE STR(a.Mobile_Number)+STR(a.Mobile_Number) not in
(SELECT STR(Mobile_Number)+STR(Insert_Date) FROM RecordMaster)
Lastly, you can apply any filter you want to the final result set like this:
select * from (
SELECT a.Mobile_Number,a.Insert_Date , a.Cpv_Status,a.Cpv_Status
FROM RecordMaster a, Dummy b
WHERE a.Mobile_Number = b.Mobile_Number and a.Insert_Date = b.Insert_Date
UNION
SELECT a.Mobile_Number,a.Insert_Date ,a.Cpv_Status
FROM Dummy a
WHERE STR(a.Mobile_Number)+STR(a.Mobile_Number) not in
(SELECT STR(Mobile_Number)+STR(Insert_Date) FROM RecordMaster)
) where Cpv_Status is not null