I have a table with the following fields:
ID, VALUES, VARIAB
Im trying to SUM the field VALUES, but it needs to be grouped by ID.
And the Subselect doesnt accept multiple lines. (When i use it with just 1 ID, works fine).
BEGIN
UPDATE TBL2
SET SOMA =
(SELECT SUM(x.VALUES)
FROM TBL3 x INNER JOIN TBL2 y
ON y.ID = x.ID
WHERE y.ID = x.ID AND X.VARIAB = 1
GROUP BY x.ID
);
END;
Im using ORACLE DB, if anyone comment this ill be very gratefull.
Sorry for my english flaws.
To update multiple records at once and/or do an upsert, use MERGE statement like this :
MERGE INTO TBL2 T2 USING (
SELECT x.id,SUM(x.VALUES) total
FROM TBL3 x
WHERE x.VARIAB = 1
GROUP BY x.ID
) T3 ON (T2.ID = T3.ID)
WHEN MATCHED THEN UPDATE
SET T2.SOMA = T3.TOTAL
The issue is that your update statement isn't correlated - ie. your subquery has no reference to the table being updated.
You could change it to:
update tbl2 t2
set soma = (select sum(t3.values)
from tbl3 t3
where t2.id = t3.id
and t3.variab = 1);
Related
I have a query which works fine in MySQL, but when I run it on Oracle I get the following error:
SQL Error: ORA-00933: SQL command not properly ended
00933. 00000 - "SQL command not properly ended"
The query is:
UPDATE table1
INNER JOIN table2 ON table1.value = table2.DESC
SET table1.value = table2.CODE
WHERE table1.UPDATETYPE='blah';
That syntax isn't valid in Oracle. You can do this:
UPDATE table1 SET table1.value = (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC)
WHERE table1.UPDATETYPE='blah'
AND EXISTS (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC);
Or you might be able to do this:
UPDATE
(SELECT table1.value as OLD, table2.CODE as NEW
FROM table1
INNER JOIN table2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) t
SET t.OLD = t.NEW
It depends if the inline view is considered updateable by Oracle
( To be updatable for the second statement depends on some rules listed
here ).
Use this:
MERGE
INTO table1 trg
USING (
SELECT t1.rowid AS rid, t2.code
FROM table1 t1
JOIN table2 t2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) src
ON (trg.rowid = src.rid)
WHEN MATCHED THEN UPDATE
SET trg.value = code;
MERGE with WHERE clause:
MERGE into table1
USING table2
ON (table1.id = table2.id)
WHEN MATCHED THEN UPDATE SET table1.startdate = table2.start_date
WHERE table1.startdate > table2.start_date;
You need the WHERE clause because columns referenced in the ON clause cannot be updated.
Do not use some of the answers above.
Some suggest the use of nested SELECT, don't do that, it is excruciatingly slow. If you have lots of records to update, use join, so something like:
update (select bonus
from employee_bonus b
inner join employees e on b.employee_id = e.employee_id
where e.bonus_eligible = 'N') t
set t.bonus = 0;
See this link for more details.
http://geekswithblogs.net/WillSmith/archive/2008/06/18/oracle-update-with-join-again.aspx.
Also, ensure that there are primary keys on all the tables you are joining.
UPDATE ( SELECT t1.value, t2.CODE
FROM table1 t1
INNER JOIN table2 t2 ON t1.Value = t2.DESC
WHERE t1.UPDATETYPE='blah')
SET t1.Value= t2.CODE
As indicated here, the general syntax for the first solution proposed by Tony Andrews is :
update some_table s
set (s.col1, s.col2) = (select x.col1, x.col2
from other_table x
where x.key_value = s.key_value
)
where exists (select 1
from other_table x
where x.key_value = s.key_value
)
I think this is interesting especially if you want update more than one field.
It works fine oracle
merge into table1 t1
using (select * from table2) t2
on (t1.empid = t2.empid)
when matched then update set t1.salary = t2.salary
This following syntax works for me.
UPDATE
(SELECT A.utl_id,
b.utl1_id
FROM trb_pi_joint A
JOIN trb_tpr B
ON A.tp_id=B.tp_id Where A.pij_type=2 and a.utl_id is null
)
SET utl_id=utl1_id;
Using description instead of desc for table2,
update
table1
set
value = (select code from table2 where description = table1.value)
where
exists (select 1 from table2 where description = table1.value)
and
table1.updatetype = 'blah'
;
UPDATE table1 t1
SET t1.value =
(select t2.CODE from table2 t2
where t1.value = t2.DESC)
WHERE t1.UPDATETYPE='blah';
UPDATE (SELECT T.FIELD A, S.FIELD B
FROM TABLE_T T INNER JOIN TABLE_S S
ON T.ID = S.ID)
SET B = A;
A and B are alias fields, you do not need to point the table.
UPDATE IP_ADMISSION_REQUEST ip1
SET IP1.WRIST_BAND_PRINT_STATUS=0
WHERE IP1.IP_ADM_REQ_ID =
(SELECT IP.IP_ADM_REQ_ID
FROM IP_ADMISSION_REQUEST ip
INNER JOIN VISIT v
ON ip.ip_visit_id=v.visit_id
AND v.pat_id =3702
); `enter code here`
Just as a matter of completeness, and because we're talking Oracle, this could do it as well:
declare
begin
for sel in (
select table2.code, table2.desc
from table1
join table2 on table1.value = table2.desc
where table1.updatetype = 'blah'
) loop
update table1
set table1.value = sel.code
where table1.updatetype = 'blah' and table1.value = sel.desc;
end loop;
end;
/
Oracle base has a good run down on this.
https://oracle-base.com/articles/misc/updates-based-on-queries
From this link - I used a modification of the above query which did not work for me (the answer from mathguy which uses rowid)
MERGE /*+ APPEND PARALLEL(8) */ INTO dest_table tt
USING source_table st
ON (tt.identifier = st.identifier)
WHEN MATCHED THEN
UPDATE SET tt.number = st.number;
Here I have two tables: source and dest. They both have a varchar field in common and I am adding the source identify field (PK) into the dest table.
update table1 a
set a.col1='Y'
where exists(select 1
from table2 b
where a.col1=b.col1
and a.col2=b.col2
)
I have a query which works fine in MySQL, but when I run it on Oracle I get the following error:
SQL Error: ORA-00933: SQL command not properly ended
00933. 00000 - "SQL command not properly ended"
The query is:
UPDATE table1
INNER JOIN table2 ON table1.value = table2.DESC
SET table1.value = table2.CODE
WHERE table1.UPDATETYPE='blah';
That syntax isn't valid in Oracle. You can do this:
UPDATE table1 SET table1.value = (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC)
WHERE table1.UPDATETYPE='blah'
AND EXISTS (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC);
Or you might be able to do this:
UPDATE
(SELECT table1.value as OLD, table2.CODE as NEW
FROM table1
INNER JOIN table2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) t
SET t.OLD = t.NEW
It depends if the inline view is considered updateable by Oracle
( To be updatable for the second statement depends on some rules listed
here ).
Use this:
MERGE
INTO table1 trg
USING (
SELECT t1.rowid AS rid, t2.code
FROM table1 t1
JOIN table2 t2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) src
ON (trg.rowid = src.rid)
WHEN MATCHED THEN UPDATE
SET trg.value = code;
MERGE with WHERE clause:
MERGE into table1
USING table2
ON (table1.id = table2.id)
WHEN MATCHED THEN UPDATE SET table1.startdate = table2.start_date
WHERE table1.startdate > table2.start_date;
You need the WHERE clause because columns referenced in the ON clause cannot be updated.
Do not use some of the answers above.
Some suggest the use of nested SELECT, don't do that, it is excruciatingly slow. If you have lots of records to update, use join, so something like:
update (select bonus
from employee_bonus b
inner join employees e on b.employee_id = e.employee_id
where e.bonus_eligible = 'N') t
set t.bonus = 0;
See this link for more details.
http://geekswithblogs.net/WillSmith/archive/2008/06/18/oracle-update-with-join-again.aspx.
Also, ensure that there are primary keys on all the tables you are joining.
UPDATE ( SELECT t1.value, t2.CODE
FROM table1 t1
INNER JOIN table2 t2 ON t1.Value = t2.DESC
WHERE t1.UPDATETYPE='blah')
SET t1.Value= t2.CODE
As indicated here, the general syntax for the first solution proposed by Tony Andrews is :
update some_table s
set (s.col1, s.col2) = (select x.col1, x.col2
from other_table x
where x.key_value = s.key_value
)
where exists (select 1
from other_table x
where x.key_value = s.key_value
)
I think this is interesting especially if you want update more than one field.
It works fine oracle
merge into table1 t1
using (select * from table2) t2
on (t1.empid = t2.empid)
when matched then update set t1.salary = t2.salary
This following syntax works for me.
UPDATE
(SELECT A.utl_id,
b.utl1_id
FROM trb_pi_joint A
JOIN trb_tpr B
ON A.tp_id=B.tp_id Where A.pij_type=2 and a.utl_id is null
)
SET utl_id=utl1_id;
Using description instead of desc for table2,
update
table1
set
value = (select code from table2 where description = table1.value)
where
exists (select 1 from table2 where description = table1.value)
and
table1.updatetype = 'blah'
;
UPDATE table1 t1
SET t1.value =
(select t2.CODE from table2 t2
where t1.value = t2.DESC)
WHERE t1.UPDATETYPE='blah';
UPDATE (SELECT T.FIELD A, S.FIELD B
FROM TABLE_T T INNER JOIN TABLE_S S
ON T.ID = S.ID)
SET B = A;
A and B are alias fields, you do not need to point the table.
UPDATE IP_ADMISSION_REQUEST ip1
SET IP1.WRIST_BAND_PRINT_STATUS=0
WHERE IP1.IP_ADM_REQ_ID =
(SELECT IP.IP_ADM_REQ_ID
FROM IP_ADMISSION_REQUEST ip
INNER JOIN VISIT v
ON ip.ip_visit_id=v.visit_id
AND v.pat_id =3702
); `enter code here`
Just as a matter of completeness, and because we're talking Oracle, this could do it as well:
declare
begin
for sel in (
select table2.code, table2.desc
from table1
join table2 on table1.value = table2.desc
where table1.updatetype = 'blah'
) loop
update table1
set table1.value = sel.code
where table1.updatetype = 'blah' and table1.value = sel.desc;
end loop;
end;
/
Oracle base has a good run down on this.
https://oracle-base.com/articles/misc/updates-based-on-queries
From this link - I used a modification of the above query which did not work for me (the answer from mathguy which uses rowid)
MERGE /*+ APPEND PARALLEL(8) */ INTO dest_table tt
USING source_table st
ON (tt.identifier = st.identifier)
WHEN MATCHED THEN
UPDATE SET tt.number = st.number;
Here I have two tables: source and dest. They both have a varchar field in common and I am adding the source identify field (PK) into the dest table.
update table1 a
set a.col1='Y'
where exists(select 1
from table2 b
where a.col1=b.col1
and a.col2=b.col2
)
I have a query which works fine in MySQL, but when I run it on Oracle I get the following error:
SQL Error: ORA-00933: SQL command not properly ended
00933. 00000 - "SQL command not properly ended"
The query is:
UPDATE table1
INNER JOIN table2 ON table1.value = table2.DESC
SET table1.value = table2.CODE
WHERE table1.UPDATETYPE='blah';
That syntax isn't valid in Oracle. You can do this:
UPDATE table1 SET table1.value = (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC)
WHERE table1.UPDATETYPE='blah'
AND EXISTS (SELECT table2.CODE
FROM table2
WHERE table1.value = table2.DESC);
Or you might be able to do this:
UPDATE
(SELECT table1.value as OLD, table2.CODE as NEW
FROM table1
INNER JOIN table2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) t
SET t.OLD = t.NEW
It depends if the inline view is considered updateable by Oracle
( To be updatable for the second statement depends on some rules listed
here ).
Use this:
MERGE
INTO table1 trg
USING (
SELECT t1.rowid AS rid, t2.code
FROM table1 t1
JOIN table2 t2
ON table1.value = table2.DESC
WHERE table1.UPDATETYPE='blah'
) src
ON (trg.rowid = src.rid)
WHEN MATCHED THEN UPDATE
SET trg.value = code;
MERGE with WHERE clause:
MERGE into table1
USING table2
ON (table1.id = table2.id)
WHEN MATCHED THEN UPDATE SET table1.startdate = table2.start_date
WHERE table1.startdate > table2.start_date;
You need the WHERE clause because columns referenced in the ON clause cannot be updated.
Do not use some of the answers above.
Some suggest the use of nested SELECT, don't do that, it is excruciatingly slow. If you have lots of records to update, use join, so something like:
update (select bonus
from employee_bonus b
inner join employees e on b.employee_id = e.employee_id
where e.bonus_eligible = 'N') t
set t.bonus = 0;
See this link for more details.
http://geekswithblogs.net/WillSmith/archive/2008/06/18/oracle-update-with-join-again.aspx.
Also, ensure that there are primary keys on all the tables you are joining.
UPDATE ( SELECT t1.value, t2.CODE
FROM table1 t1
INNER JOIN table2 t2 ON t1.Value = t2.DESC
WHERE t1.UPDATETYPE='blah')
SET t1.Value= t2.CODE
As indicated here, the general syntax for the first solution proposed by Tony Andrews is :
update some_table s
set (s.col1, s.col2) = (select x.col1, x.col2
from other_table x
where x.key_value = s.key_value
)
where exists (select 1
from other_table x
where x.key_value = s.key_value
)
I think this is interesting especially if you want update more than one field.
It works fine oracle
merge into table1 t1
using (select * from table2) t2
on (t1.empid = t2.empid)
when matched then update set t1.salary = t2.salary
This following syntax works for me.
UPDATE
(SELECT A.utl_id,
b.utl1_id
FROM trb_pi_joint A
JOIN trb_tpr B
ON A.tp_id=B.tp_id Where A.pij_type=2 and a.utl_id is null
)
SET utl_id=utl1_id;
Using description instead of desc for table2,
update
table1
set
value = (select code from table2 where description = table1.value)
where
exists (select 1 from table2 where description = table1.value)
and
table1.updatetype = 'blah'
;
UPDATE table1 t1
SET t1.value =
(select t2.CODE from table2 t2
where t1.value = t2.DESC)
WHERE t1.UPDATETYPE='blah';
UPDATE (SELECT T.FIELD A, S.FIELD B
FROM TABLE_T T INNER JOIN TABLE_S S
ON T.ID = S.ID)
SET B = A;
A and B are alias fields, you do not need to point the table.
UPDATE IP_ADMISSION_REQUEST ip1
SET IP1.WRIST_BAND_PRINT_STATUS=0
WHERE IP1.IP_ADM_REQ_ID =
(SELECT IP.IP_ADM_REQ_ID
FROM IP_ADMISSION_REQUEST ip
INNER JOIN VISIT v
ON ip.ip_visit_id=v.visit_id
AND v.pat_id =3702
); `enter code here`
Just as a matter of completeness, and because we're talking Oracle, this could do it as well:
declare
begin
for sel in (
select table2.code, table2.desc
from table1
join table2 on table1.value = table2.desc
where table1.updatetype = 'blah'
) loop
update table1
set table1.value = sel.code
where table1.updatetype = 'blah' and table1.value = sel.desc;
end loop;
end;
/
Oracle base has a good run down on this.
https://oracle-base.com/articles/misc/updates-based-on-queries
From this link - I used a modification of the above query which did not work for me (the answer from mathguy which uses rowid)
MERGE /*+ APPEND PARALLEL(8) */ INTO dest_table tt
USING source_table st
ON (tt.identifier = st.identifier)
WHEN MATCHED THEN
UPDATE SET tt.number = st.number;
Here I have two tables: source and dest. They both have a varchar field in common and I am adding the source identify field (PK) into the dest table.
update table1 a
set a.col1='Y'
where exists(select 1
from table2 b
where a.col1=b.col1
and a.col2=b.col2
)
I'm struggling with a join/where clause with what is a rather simple sql select statement.
I am trying to retrieve a list of product information from tb1 with the where condition behind situated in tbl2 but this must be joined by three different columns.
so the SQL would look something along the lines of:
SELECT tb1.*
FROM tb2 INNER JOIN
tb1 ON tb2.Col1 = tb1. Col1 AND tb2.Col2 = tb1. Col2 AND
tb2.Col3 = tb1.Col3
WHERE (tb2.Col1 = col1) AND (tb2.Col2 = col2) AND (tb2.Col4 = string)
ColX is the main where clause with the string to be passed in as parameter; all other columns are within the contexts.
How do you implement multiple joins with a where clause?
And shoves in the right direction, muchly appreciated.
To join on multiple field in LINQ, you have to create a new anonymous type containing the columns you want to compare and then use that anonymous type in the join:
var results = from t1 in context.tb1
join t2 in context.tb2
on new { t1.Col1, t1.Col2, t1.Col3 } equals
new { t2.Col1, t2.Col2, t2.Col3 }
where t2.Col1 == col1 && t2.Col2 == col2 && t2.Col4 == someString
select t1;
And here is the equivalent Lambda Syntax:
var results = context.tb1.Join(
context.tb2,
t1 => new { t1.Col1, t1.Col2, t1.Col3 },
t2 => new { t2.Col1, t2.Col2, t2.Col3 },
(t1, t2) => new { t1, t2 })
.Where(o => o.t2.Col1 == col1
&& o.t2.Col2 == col2
&& o.t2.Col4 == someString)
.Select(o => o.t1);
As you can see, in the case of joins, query syntax usually produces an easier to read statement.
You can also include the WHERE clause in lamda syntax in the reference to the table you're joining on.
var query = from pt in dc.ProjectTasks
join ttab in dc.TimeTaskAssigns on pt.Id equals ttab.ProjectTaskId
join ttb2 in dc.CMS_TAT_TIMEs.Where(a => a.WIP_STATUS == 'B') on ttab.CmsTimeUno equals ttb2.TIME_UNO
select pt;
Seems obvious now, doesn't it? It took me a long time to find that solution.
Also you can group result and use subquery
var innerGroupJoinQuery2 =
from category in categories
join prod in products on category.ID equals prod.CategoryID into prodGroup
from prod2 in prodGroup
where prod2.UnitPrice > 2.50M
select prod2;
ref.: join clause (C# Reference)
In a WHERE part of query we have
SELECT * FROM SomeTable st
WHERE
NOT EXISTS
(SELECT 1 FROM Tab t1 WHERE t1.A = st.A OR t1.B = st.A)
OR EXISTS
(SELECT 1 FROM Tab t2 WHERE (t2.A = st.A OR t2.B = st.A) AND t2.C IS NULL)
Seems like a good candidate for merging...
But I'm staring on that for an hour without any idea.
Would you have some thoughts?
Thanks,
SELECT distinct st.*
FROM SomeTable st
left outer join Tab t
on st.a in (t.a,t.b)
WHERE t.c is null
Sometimes the simplest answer is to use a join rather than an exists.
Try:
NOT EXISTS (SELECT 1 FROM Tab WHERE (Tab.A = #Id OR Tab.B = #ID)
AND (Tab.C is null OR Tab.C != #Var))
I think I've got it
SELECT * FROM SomeTable st
WHERE
0 = (SELECT SUM (NVL(t1.C,0) FROM Tab t1 WHERE t1.A = st.A OR t1.B = st.A)
What do you think?
Of course that will work only for case where 'C' is expected to be NULL.
If we are expecting some value, I think that the check
#SomeValue IN (SELECT t1.C FROM Tab t1 WHERE t1.A = st.A OR t1.B = st.A)
Will do?
It looks as though you're trying to return all SomeTable values where there's no corresponding value on Tab where C is not null - in which case, the following should work:
SELECT * FROM SomeTable st
WHERE NOT EXISTS
(SELECT 1 FROM Tab t1 WHERE (t1.A = st.A OR t1.B = st.A) AND t2.C IS NOT NULL)
However, that's not exactly what your existing query does - in your query, if there are two corresponding records on Tab, where one record has a non-null value for C and the other is null, your query will return a corresponding row, but my query won't.