SQL NOT IN failed - sql

I am working on a query that will check the temp table if there is a record that do not exist on the main table. My query looks like this
SELECT * FROM [Telemarketing].[dbo].[PDCampaignBatch_temp]
WHERE [StartDateTime] NOT IN (SELECT [StartDateTime] FROM [Telemarketing].[dbo].PDCampaignBatch GROUP BY [StartDateTime])
but the problem is it does not display this row
even if that data does not exist in my main table. What seems to be the problem?

NOT IN has strange semantics. If any values in the subquery are NULL, then the query returns no rows at all. For this reason, I strongly recommend using NOT EXISTS instead:
SELECT t.*
FROM [Telemarketing].[dbo].[PDCampaignBatch_temp] t
WHERE NOT EXISTS (SELECT 1
FROM [Telemarketing].[dbo].PDCampaignBatch cb
WHERE t.StartDateTime = cb.StartDateTime
);

If the set is evaluated by the SQL NOT IN condition contains any values that are null, then the outer query here will return an empty set, even if there are many [StartDateTime]s that match [StartDateTime]s in the PDCampaignBatch table.
To avoid such issue,
SELECT *
FROM [Telemarketing].[dbo].[PDCampaignBatch_temp]
WHERE [StartDateTime] NOT IN (
SELECT DISTINCT [StartDateTime]
FROM [Telemarketing].[dbo].PDCampaignBatch
WHERE [StartDateTime] IS NOT NULL
);

Let's say PDCampaignBatch_temp and PDCampaignBatch happen to have the same structure (same columns in the same order) and you're tasked with getting the set of all rows in PDCampaignBatch_temp that aren't in PDCampaignBatch. The most effective way to do that is to make use of the EXCEPT operator, which will deal with NULL in the expected way as well:
SELECT * FROM [Telemarketing].[dbo].[PDCampaignBatch_temp]
EXCEPT
SELECT * FROM [Telemarketing].[dbo].[PDCampaignBatch]
In production code that is not a one-off, don't use SELECT *, write out the column names instead.

Most likely your issue is with the datetime. You may be only displaying a certain degree of percision like the year/month/date. The data may be stored as year/month/date/hour/minute/second/milisecond. If so you have to match down the the most granluar measurement of the data. If one field is a date and the other is a date time they also will likely never match up. Thus you always get no responses.

Related

Sub-query result set not read by main query

I have written a query like this :
select *
from DATASYNCH_HA_TO_TRG_AUDIT_HIST
where PSX_BATCH_ID IN (select PSX_BATCH_ID
from DATASYNCH_HA_TO_TRG_AUDIT_T
);
Here,when I execute the sub-query alone, it results some values and when I put those values in the place of sub-query the main query also returns some values.But,when I use this whole query ,it does not result any values.How is it possible?
Hope PSX_BATCH_ID column datatype is integer. if it is varchar filed, then trim the value.
SELECT *
FROM DATASYNCH_HA_TO_TRG_AUDIT_HIST
WHERE TRIM(PSX_BATCH_ID) IN
(SELECT TRIM(PSX_BATCH_ID) FROM DATASYNCH_HA_TO_TRG_AUDIT_T);
Instead of using IN query, use JOIN
SELECT *
FROM DATASYNCH_HA_TO_TRG_AUDIT_HIST A
INNER JOIN DATASYNCH_HA_TO_TRG_AUDIT_T B
ON (A.PSX_BATCH_ID = B.PSX_BATCH_ID)
Really Sorry guys,I found out it was actually a database issue.I ran the query recently in a procedure,it is working fine.

SQL Server where column in where clause is null

Let's say that we have a table named Data with Id and Weather columns. Other columns in that table are not important to this problem. The Weather column can be null.
I want to display all rows where Weather fits a condition, but if there is a null value in weather then display null value.
My SQL so far:
SELECT *
FROM Data d
WHERE (d.Weather LIKE '%'+COALESCE(NULLIF('',''),'sunny')+'%' OR d.Weather IS NULL)
My results are wrong, because that statement also shows values where Weather is null if condition is not correct (let's say that users mistyped wrong).
I found similar topic, but there I do not find appropriate answer.
SQL WHERE clause not returning rows when field has NULL value
Please help me out.
Your query is correct for the general task of treating NULLs as a match. If you wish to suppress NULLs when there are no other results, you can add an AND EXISTS ... condition to your query, like this:
SELECT *
FROM Data d
WHERE d.Weather LIKE '%'+COALESCE(NULLIF('',''),'sunny')+'%'
OR (d.Weather IS NULL AND EXISTS (SELECT * FROM Data dd WHERE dd.Weather LIKE '%'+COALESCE(NULLIF('',''),'sunny')+'%'))
The additional condition ensures that NULLs are treated as matches only if other matching records exist.
You can also use a common table expression to avoid duplicating the query, like this:
WITH cte (id, weather) AS
(
SELECT *
FROM Data d
WHERE d.Weather LIKE '%'+COALESCE(NULLIF('',''),'sunny')+'%'
)
SELECT * FROM cte
UNION ALL
SELECT * FROM Data WHERE weather is NULL AND EXISTS (SELECT * FROM cte)
statement show also values where Wether is null if condition is not correct (let say that users typed wrong sunny).
This suggests that the constant 'sunny' is coming from end-user's input. If that is the case, you need to parameterize your query to avoid SQL injection attacks.

mismatch not picked up when one value is null

I have a simple SQL query where a comparison is done between two tables for mismatching value.
Yesterday, we picked up an issue where one field was null and the other wasn't, but a mismatch was not detected.
As far as I can determine,the logic has been working all along until yesterday.
Here is the logic of the SQL:
CREATE TABLE Table1
(UserID INT,PlayDate DATETIME)
CREATE TABLE Table2
(UserID INT,PlayDate DATETIME)
INSERT INTO Table1 (UserID)
SELECT 5346
INSERT INTO Table2 (UserID,PlayDate)
SELECT 5346,'2012-11-01'
SELECT a.UserID
FROM Table1 a
INNER JOIN
Table2 b
ON a.UserID = b.UserID
WHERE a.PlayDate <> b.PlayDate
No values are returned even though the PlayDate values are different.
I have now updated the WHERE to read:
WHERE ISNULL(a.PlayDate,'') <> ISNULL(b.PlayDate,'')
Is there a setting in SQL which someone could have changed to cause the original code to no longer pick up the difference in fields?
Thanks
NULL <> anything
is unknown not true. SQL uses three valued logic (false/true/unknown) and the predicate needs to evaluate to true in a where clause for the row to be returned.
In fact in standard SQL any comparison with NULL except for IS [NOT] NULL yields unknown. Including WHERE NULL = NULL
You don't state RDBMS but if it supports IS DISTINCT FROM you could use that or if you are using MySQL it has a null safe equality operator <=> you could negate.
You say you think it previously behaved differently. If you are on SQL Server you might be using a different setting for ANSI_NULLS somehow but this setting is deprecated and you should rewrite any code that depends on it anyway.
You can simulate IS DISTINCT FROM in SQL Server with WHERE EXISTS (SELECT a.PlayDate EXCEPT SELECT b.PlayDate)
Not even a NULL can be equal to NULL.
Here are two common queries that just don’t work:
select * from table where column = null;
select * from table where column <> null;
there is no concept of equality or inequality, greater than or less
than with NULLs. Instead, one can only say “is” or “is not”
(without the word “equal”) when discussing NULLs.
- The correct way to write the queries
select * from table where column IS NULL;
select * from table where column IS NOT NULL;

sql server - how to execute the second half of 'or' clause only when first one fails

Suppose I have a table with following records
value text
company/about about Us
company company
company/contactus company contact
I have a very simple query in sql server as below. I am having problem with the 'or' condition. In below query, I am trying to find text for value 'company/about'. If it is not found, then only I want to run the other side of 'or'. The below query returns two records as below
value text
company/about about Us
company company
Query
select
*
from
tbl
where
value='company/about' or
value=substring('company/about',0,charindex('/','company/about'))
How can I modify the query so the result set looks like
value text
company/about about Us
A bit roundabout, but you can check for the existence of results from the first where clause:
select
*
from
tbl
where
value='company/about' or
(
not exists (select * from tbl where value='company/about')
and
value=substring('company/about',0,charindex('/','company/about'))
)
Since your second condition can be re-written as value = 'company' this would work (at least for the data and query you've presented):
select top(1) [value], [text]
from dbo.MyTable
where value in ('company/about', 'company')
order by len(value) desc
The TOP() ignores the second row if both are found, and the ORDER BY ensures that the first row is always the one with 'company/about', if it exists.

TSQL NOT EXISTS Why is this query so slow?

Debugging an app which queries SQL Server 05, can't change the query but need to optimise things.
Running all the selects seperately are quick <1sec, eg: select * from acscard, select id from employee... When joined together it takes 50 seconds.
Is it better to set uninteresting accesscardid fields to null or to '' when using EXISTS?
SELECT * FROM ACSCard
WHERE NOT EXISTS
( SELECT Id FROM Employee
WHERE Employee.AccessCardId = ACSCard.acs_card_number )
AND NOT EXISTS
( SELECT Id FROM Visit
WHERE Visit.AccessCardId = ACSCard.acs_card_number )
ORDER by acs_card_id
Do you have indexes on Employee.AccessCardId, Visit.AccessCardId, and ACSCard.acs_card_number?
The SELECT clause is not evaluated in an EXISTS clause. This:
WHERE EXISTS(SELECT 1/0
FROM EMPLOYEE)
...should raise an error for dividing by zero, but it won't. But you need to put something in the SELECT clause for it to be a valid query - it doesn't matter if it's NULL or a zero length string.
In SQL Server, NOT EXISTS (and NOT IN) are better than the LEFT JOIN/IS NULL approach if the columns being compared are not nullable (the values on either side can not be NULL). The columns compared should be indexed, if they aren't already.