I have checked the forum here and understand that t-SQL does not have FOR loops. However, I'm wondering the best way to do what I need to do here.
Here is my table (mytable) simplified:
id Code1 Code2 Code3 CodeDesc1 CodeDesc2 CodeDesk3
__ _____ _____ _____ _________ _________ _________
1 ABC DEF ZYX
2 DEF ABC ZYX
3 ZYX GHJ ABC
There are many more rows and many more Code and CodeDesc columns in the real situation. I am converting SAS code to T-SQL and the SAS code has this (simplified):
DO x = 1 to 4 ;
IF code0(x) = 'ABC' THEN CodeDesc0(x) = 'ABC Description'
ELSE
IF code0(x) = 'DEF' THEN CodeDesc0(x) = 'DEF Description'
ELSE
IF code0(x) = 'ZYX' THEN CodeDesc0(x) = 'ZYX Description'
My question is how can I do something similar in T-SQL.
Thanks for your suggestions.
Adding more info...
If I were to do what I need to do using straight T-SQL, without any programming it would look like this:
UPDATE mytable SET CodeDesc1 = 'ABC Description' WHERE Code1 = 'ABC'
UPDATE mytable SET CodeDesc2 = 'ABC Description' WHERE Code2 = 'ABC'
UPDATE mytable SET CodeDesc3 = 'ABC Description' WHERE Code3 = 'ABC'
UPDATE mytable SET CodeDesc1 = 'DEF Description' WHERE Code1 = 'DEF'
UPDATE mytable SET CodeDesc2 = 'DEF Description' WHERE Code2 = 'DEF'
UPDATE mytable SET CodeDesc3 = 'DEF Description' WHERE Code3 = 'DEF'
So, as you can see with several different codes and descriptions I could easily be writing 100's of update statements to cover all the possibilities.
I hope that helps to shed a bit more light.
Do not use a loop. Upddates should never be done in loops!!!!
What I would do is create a table that maps the abreviations to the long descriptions.
Then you can add new ones easily and you can write the update this way:
Update a
set description = b.description
from tableA a
join tableb b on a.abbreviation = b.abbrevation
where a.description <> b.descrioption or a.description is null
Here's an example of what you're looking for:
/********** example 1 **********/
declare #au_id char( 11 )
set rowcount 0
select * into #mytemp from authors
set rowcount 1
select #au_id = au_id from #mytemp
while ##rowcount <> 0
begin
set rowcount 0
select * from #mytemp where au_id = #au_id
delete #mytemp where au_id = #au_id
set rowcount 1
select #au_id = au_id from #mytemp<BR/>
end
set rowcount 0
Taken from here: http://support.microsoft.com/kb/111401
Related
Thaks #Tim Biegeleisen but ,at the end of line is not needed. In this code first part(email) works fine but second part(mobileno) is not working. I have tried it with both , after end and without ,.
UPDATE table_name
SET
email = CASE id WHEN 1 THEN email1
WHEN 2 THEN email2
...
WHEN 22 THEN email22 END
mobileno = CASE id WHEN 1 THEN 00000
WHEN 2 THEN 11111
...
WHEN 22 THEN 2222222222 END
WHERE
id BETWEEN 1 AND 22;
I would use a VALUES clause (similar to an INSERT statement):
UPDATE table_name t
SET email = v.email,
mobileno = v.mobileno
from (
values
(1, 'email1', '00001'),
(2, 'email2', '00002'),
(3, 'email3', '00003'),
....
(22, 'email22', '00022')
) as v(id, email, mobileno)
where t.id = v.id;
The best option might be to create a separate table which contains the desired email and mobile values, based on some id. Lacking this, we could use a big CASE expression here:
UPDATE table_name
SET
email = CASE id WHEN 1 THEN email1,
WHEN 2 THEN email2,
...
WHEN 22 THEN email22 END,
mobileno = CASE id WHEN 1 THEN 00000,
WHEN 2 THEN 11111,
...
WHEN 22 THEN 2222222222 END
WHERE
id BETWEEN 1 AND 22;
Assuming you did have a table with the id values and emails/mobile numbers, you could do an update join here:
UPDATE table_name AS t1
SET email = t2.email, mobileno = t2.mobileno
FROM temp_table_name t2
WHERE t1.id = t2.id;
I'm trying to INSERT into a table a column that is part of another column in another table using TSQL, but I get the error stating that there is more than one value returned when I used that subquery as an expression. I understand what causes the error, but I can't seem to think of a way to make it produce what I want.
I'm trying to do something similar to:
A.Base B.Reference C.Wanted
--- ---- ----
abcdaa aa abcdaa
bcdeab bb cdefbb
cdefbb cc efghcc
defgbc ddd fghddd
efghcc
fghddd
So I'm using the code:
INSERT INTO C ( [Some other column], Wanted )
SELECT
A.[Some other column],
, CASE
WHEN LEN( B.Reference ) = 2 THEN
( SELECT A.Base FROM A WHERE RIGHT( A.Base, 2 ) =
( SELECT B.Reference FROM B WHERE LEN( B.Reference ) = 2 )
)
WHEN LEN( B.Reference ) = 3 THEN
( SELECT A.Base FROM A WHERE RIGHT( A.Base, 3 ) =
( SELECT B.Reference FROM B WHERE LEN( B.Reference ) = 3 )
)
END
FROM
A
, B
Which will return me the "more than 1 value" error. Honestly, I'm probably making this way more convoluted than it needs to be, but I've been staring at these tables for a while now.
I hope I'm getting the idea across as to what I'm trying to do.
If you know the records aren't duplicate, and you are sure your JOIN between A and B works (as Martin mentioned) can't you just select distinct to return just the unique records?
I'd try it like this:
--Create a mockup with declared table variables and test data
DECLARE #tblA TABLE(someColumnInA VARCHAR(100));
DECLARE #tblB TABLE(someColumnInB VARCHAR(100));
DECLARE #tblC TABLE(someColumnInC VARCHAR(100));
INSERT INTO #tblA VALUES
('abcdaa')
,('bcdeab')
,('cdefbb')
,('defgbc')
,('efghcc')
,('fghddd')
INSERT INTO #tblB VALUES
('aa')
,('bb')
,('cc')
,('ddd');
--The query
INSERT INTO #tblC(someColumnInC)
SELECT SomeColumnInA
FROM #tblA a
WHERE EXISTS(SELECT 1 FROM #tblB b WHERE a.someColumnInA LIKE '%' + b.SomeColumnInB + '%');
SELECT * FROM #tblC;
The idea in short:
After creating a mockup (please do this next time in advance) we use a query to insert all values from #tblA into #tblC as long as there exists any value in #tblB, which is part of the current value in #tblA.
How about doing something like this?
select *
from A
where RIGHT(A.Base,2) IN (select B.Reference FROM B WHERE LEN(B.Reference) = 2)
UNION ALL
select *
from A
where RIGHT(A.Base,3) IN (select B.Reference FROM B WHERE LEN(B.Reference) = 3)
Second select (from linked server) does not return any values.. Object_ID doesnt work. Is any workaround?
select '', name
FROM sys.databases
WHERE 1 = 1
AND NAME <> db_name() -- exclude current database
AND CASE
WHEN STATE = 0
THEN CASE
WHEN OBJECT_ID(NAME + '.dbo.tPA_SysParam', 'U') IS NOT NULL
THEN 1
END
END = 1
union
select '[LINKED]', name
FROM [LINKED].master.sys.databases
WHERE 1 = 1
AND CASE
WHEN STATE = 0
THEN CASE
WHEN OBJECT_ID('[LINKED].'+NAME + '.dbo.tPA_SysParam', 'U') IS NOT NULL
THEN 1
END
END = 1
You can also mimic OBJECT_ID with a little help from the PARSENAME function:
Declare #FullTableName nvarchar(max) = '[dbo].[MyTable]';
Select t.object_id
From [LINKED].MyDatabase.sys.tables As t
Inner Join [LINKED].MyDatabase.sys.schemas As s On t.schema_id = s.schema_id
Where t.[name] = PARSENAME(#FullTableName, 1)
And s.[name] = PARSENAME(#FullTableName, 2)
I need a query based on an exclusive Or statement for this I try using the case but I won't get a result in case of Null...
...
and b.[U_Periode] = CASE
when (b.U_Periode= #period) then #period
when (b.U_Periode is NULL ) then null
end
...
The Case that won't be catched is... if B.U_Status is null and b.U_Periode is null.
If the var Periode match the Value and the U_Status = 0 or 1
the only way getting this working for me was this:
...
and
ISNULL( b.[U_Status],'0') = CASE
when (b.U_Status= '1') then '1'
when (isnull( b.U_Status,'0')= '0') then '0'
end
and
ISNULL (b.[U_Periode],'01.01.1901') = CASE
when (b.U_Periode= #period) then #period
when (ISNULL (b.U_Periode,'01.01.1901') = '01.01.1901' ) then '01.01.1901'
end
are there any other better solutions for this?
Best regards
Oliver
Okay... here is my Problem
Table1
InsID ContractID
1 1
2 1
Table2
ID insid Period Status Count
1 1 null null 100
2 1 30.09.2015 1 500
3 2 null null 100
4 2 30.09.2015 1 500
Case '31.08.2015'
in total Value should be 200
in case of '30.09.2015'
the Value should be 1.000
XOR /OR will do the same in this case.
Value case '31.08.2015' = 200
value Case ' 30.09.2015 = 2200
So this is somesing like a subquery
left join (
[dbo].[Table2]b
inner join [dbo].[Table 3]K on k.DocEntry = b.DocEntry and CAST( k.U_CSetID as int) >0
)
on b.[U_InsID] in(select... but here I should have an if statement...
If there are results matching the Date join this
If not than join the result NULL is matching to periode...
Okay.. here is the complete query
the Table2 with the Date of 31.08.2015 should have one Record that include
B_STATUS = Null and B.Preiode = null and there is no available record with the U_Periode '31.08.2015' and a Staus ...
with a date of 30.09.2015
there is a Record matching U_Period = '30.09.2015' in this case the Record with U_Period=null should not effect the result...
Declare #period as varchar(20)= '31-08-2015 00:00:00'
declare #Customer as Varchar(15)='12345'
declare #Contract as varchar(30) = '123'
declare #test as varchar(1) = null
select SUM(cast(K.U_Count as decimal))as counter, K.U_CounterTyp
from [dbo].[Table1] a
left join (
[dbo].[Table2]b
inner join [dbo].[Table 3]K on k.DocEntry = b.DocEntry and CAST( k.U_CSetID as int) >0
)
on b.[U_InsID]=a.[insID] and b.[U_ObjectType]in ('5','1') and
ISNULL( b.[U_Status],'0') = CASE
when (b.U_Status= '1') then '1'
when (isnull( b.U_Status,'0')= '0') then '0'
end
and
ISNULL (b.[U_Periode],'01.01.1901') = CASE
when (b.U_Periode= #period) then #period
when (ISNULL (b.U_Periode,'01.01.1901') = '01.01.1901' ) then '01.01.1901'
end
where a.[customer] =#Customer and a.[Status]='A' and a.[U_ContrCount]='1'
and a.[manufSN] in(
select c.[ManufSN] from [dbo].[Table4] c
inner join [dbo].[OCTR]d on d.[ContractID] = c.[ContractID]
where c.[ManufSN]=a.[manufSN]
and d.[CstmrCode] = a.[customer] and d.[ContractID]=#Contract
)
group by K.U_CounterTyp
you must use the "^" operand, this is XOR operand in TSQL
expression ^ expression
the expression must return or 0 or 1...
stupid example:
WHERE (name like "stackoverflow") ^ (age > 10)
font: https://msdn.microsoft.com/en-us/library/ms190277(v=sql.105).aspx
Update for your check
WHERE (CONVERT(VARCHAR(10), a.[U_Periode], 104) = '30.08.2015') != (a.U_Periode IS NULL)
Olay here is my funktion that chek either a Date Result is there or not.
The only thing is, that the performance is not the best if I run hundreths of sentences ... foreach record (up to app. 1000) I need to query each subtable... and there are many many records in...
Anyway this is the function
CREATE FUNCTION fn_GetInsCounters(#InsId as Varchar(30), #Date as datetime)
returns datetime
as
begin
declare #count as int = 0
declare #Retruns as Datetime
set #count =
(
select count( b.Docentry)
from [dbo].[Table2]b
where b.U_Insid =#InsID
and b.U_Status <> '2'
and b.[U_ObjectType]in ('5','1')
and b.U_Periode=#Date
)
if(#count>0)
begin
set #Retruns = #date
end
else
begin
set #Retruns = '01.01.1901'
end
return #Retruns
end
if anyone gets a better idea???
Best regards
Oliver
I'm writing a query with some CASE expressions and it outputs helper-data columns which help me determine whether or not a specific action is required. I would like to know if I can somehow use the result of a subquery as the output without having to perform the same query twice (between WHEN (subquery) THEN and as the result after THEN)
The dummy code below describes what I'm after. Can this be done? I'm querying a MS2005 SQL database.
SELECT 'Hello StackOverflow'
,'Thanks for reading this question'
,CASE
WHEN
(
SELECT count(*)
FROM sometable
WHERE condition = 1
AND somethingelse = 'value'
) > 0 THEN
-- run the query again to get the number of rows
(
SELECT count(*)
FROM sometable
WHERE condition = 1
AND somethingelse = 'value'
)
ELSE 0
END
SELECT 'Hello StackOverflow'
,'Thanks for reading this question'
,CASE
WHEN
(
SELECT count(*)
FROM sometable
WHERE condition = 1
AND somethingelse = 'value'
) AS subqry_count > 0 THEN
-- use the subqry_count, which fails... "Incorrect syntax near the keyword 'AS'"
subqry_count
ELSE 0
END
Just use the subquery as the source you are selecting from:
SELECT 'Hello StackOverflow'
,'Thanks for reading this question'
,CASE subqry_count.Cnt
WHEN 0 THEN 0
ELSE subqry_count.Cnt
END
FROM ( SELECT count(*) AS Cnt
FROM sometable
WHERE condition = 1
AND somethingelse = 'value'
) subqry_count
As an aside, if you are just going to return 0 if the output from COUNT is 0, then you don't even need to use a CASE statement.