Pages

22 December 2013

SQL- Checking Duplicate Records in a Table

Q. How to check duplicate records in a table using SQL Server ?

        Suppose we have a table without a primary key. We want to find whether duplicate records are present in the table or not. For this scenario, we have to select all columns name and count from a table and group it by all columns name in a table and then filter based on count greater than 1. Lets see the below query.

 SELECT OrderNo,Customer,Amount,Location,Item,COUNT(*)AS TotCount 
 FROM OrderDetails 
 GROUP BY OrderNo,Customer,Amount,Location,Item
HAVING COUNT(*)>1



We see that records which are marked with red is the duplicate record. Lets find out with the help of query.



Suppose we have a scenario where we need to find all the records based on the duplicate value present in a column. Lets assume we have to find all the records from the table where column  'Location' has duplicate value.

SELECT OD.* FROM OrderDetails OD 
INNER JOIN 
(SELECT Location
FROM OrderDetails
GROUP BY Location
HAVING COUNT(*)>1)A
 ON A.Location=OD.Location





Thus we see that all the records related to duplicates present in the column 'Location' has been filtered out.

SQL- NULL Functions

1. ISNULL() : 

Syntax:
ISNULL ( check_expression , replacement_value )
It replaces the NULL value with replacement value. If the Check_expression evaluates to NULL then it is replaced with the replacement value else the value of Check_expression is returned.
ISNULL function can be used both in SQL SERVER and MSAccess
Let's understand with examples: Consider a below table SalesInfo for illustrations:
SELECT * FROM SalesInfo


We see that some of the products has not been sold . This is is evident as we can see there are NULL in QuantitySold column. Now,for better understanding we want  to display zero insteadof  NULL in QuantitySold column,we can do this by using ISNULL function.

SELECT ProductId,ProductName,Price,
ISNULL(QuantitySold,0)
FROM SalesInfo





Suppose we want to calculate, sales per item. For this we need to multiply the unit price and QuantitySold. If the Quantity sold is NULL, we can not get the correct result because multiplication of any number with NULL is always NULL. In such scenario, ISNULL() function is very useful. We can easily replace all NULL values with zero using ISNULL() function.

SELECT ProductName,
Price*(ISNULL(QuantitySold,0))AS SalePerItem
FROM SalesInfo




2. NVL() :

Syntax:
NVL( check_expression , replacement_value )
Just like ISNULL function in SQL SERVER, we have NVL function in ORACLE. It replaces the NULL value with replacement value. If the Check_expression evaluates to NULL then it is replaced with the replacement value else the value of Check_expression is returned.
This function holds the same explaination as that of ISNULL function.
SQL> SELECT ProductId,ProductName,Price,NVL(QuantitySold,0)
   2      FROM SalesInfo;

NOTE: IFNULL performs the same function as ISNULL () and NVL()
in Mysql database.

3. COALESCE():
 Syntax:
 COALESCE(Expression1, Expression2[,Expression3.........ExpressionN)
It returns the first non null value among its arguments.
Let us understand from the 'TelephoneRecords' Table.
SELECT * FROM TelephoneRecords

We want to display the Name and Mobile number of each person. For displaying the mobile number, Primary No should be displayed. If the Primary number is not available then Alternate No should be displayed.If Alternate No is not available then Office No should be displayed. For such a scenario, COALESCE function will serve our purpose as it always returns the first non null value.


SELECT Name,
COALESCE(PrimaryNo,AlternateNo,OfficeNo)
AS MobileNo
FROM TelephoneRecords





NOTE: If all the arguments in COALESCE function returns NULL then output will also be NULL.

It can also work similar to ISNULL function.

SELECT Name,COALESCE(PrimaryNo,0) 
AS MobileNo
FROM TelephoneRecords


We see that all NULL values replaced with zero.

18 November 2013

SQL- Scenario Based Questions PART- 1

SQL- Scenario based questions Part 1.

                                   Now the recruiters are setting a new trend by asking a scenario based questions in an interview.  Keeping in this mind, we have come up with many standard scenario based questions.

Scenario: You have been given the following Table "Annualhike". This table contains the name of employee, salary, designation, CRR and the increment. You have to test the increment column of the table based on the below logic:

  • Employee gets an increment of 15% of Salary if he/she gets CRR 1+.
  • Employee gets an increment of 10% of Salary if he/she gets CRR 1.
  • Employee gets an increment of 5% of Salary if he/she gets CRR 2.

You have to find out the records if there are any incorrect population in Increment field.

SELECT * FROM Annualhike


Solution

Method1

SELECT * FROM Annualhike A 
WHERE Increment <> (SELECT CASE WHEN A.CRR='2' 
                      THEN .05* SalaryPackage
      WHEN A.CRR='1'  THEN .1* SalaryPackage
      WHEN A.CRR='1+' THEN .15* SalaryPackage  
      END AS Increment)



Method 2

SELECT * FROM Annualhike
EXCEPT
SELECT Employee,SalaryPackage,Designation,CRR ,
              CASE WHEN CRR='2' THEN .05* SalaryPackage
                         WHEN CRR='1' THEN .1* SalaryPackage
                         WHEN CRR='1+' THEN .15* SalaryPackage  
              END AS Increment
 FROM Annualhike



For other SQL Scenario Based Questions, please visit below link
http://itjunction4all.blogspot.com/2015/05/sql-scenario-based-questions-part-3.html

6 November 2013

SQL- Displaying all records from a table for last one year

Q. How to find all the records from a table for last one year ?

       A table named "Transactions" has the following transaction details.

Transactions
Suppose, we have to find all the records which has been transacted in the last one year. For this scenario we will make use of date field which is "DateOfTransaction" .The following below query in SQL Server will fetch the required results.


SELECT * FROM TRANSACTIONS 

WHERE DateOfTransaction > DATEADD(YYYY,-1,GETDATE())





Suppose we have to get total amount spent by each ID in last one year. The following below query will fetch the desired results.

SELECT ID,SUM(Amount) AS TotAmount FROM TRANSACTIONS 
WHERE DateOfTransaction > DATEADD(YYYY,-1,GETDATE())
GROUP BY ID


Similar scenario can be handled in ORACLE with slight changes in query.We will have to to use ADD_MONTHS instead of DATEADD and SYSDATE instead of GETDATE().

SELECT * FROM TRANSACTIONS
WHERE dateOftransaction > ADD_MONTHS(SYSDATE,-12);

SELECT ID,SUM(Amount) TotAmount FROM TRANSACTIONS
WHERE DateOfTransaction > ADD_MONTHS(SYSDATE,-12)
GROUP BY ID;








9 September 2013

SQL- String Functions

  
SQL String functions are built in functions. It is deterministic in nature which means that it always  produces the same result when particular or specific input is provided under same state of database. String functions helps in manipulating string operations.

1.ASCII: It returns the ASCII code of the extreme left character in a character expression or string.

Syntax:
ASCII( Character expression )

Examples:
SELECT ASCII('A') AS
--65
SELECT ASCII('B'AS B
--66
SELECT ASCII('S'AS S
--83
SELECT ASCII('SUNIL'AS SUNIL
--83


ASCII Illustrations
2.CHAR: It converts ASCII code value to character.


Syntax:
CHAR( numeric expression )

Examples:
SELECT CHAR('65'AS 
--A
SELECT CHAR('66'AS B
--B
SELECT CHAR('83'AS S
--S





NOTE: Numeric expression should be in between 0 and 255. If it is not within this range, NULL value will be returned.

SELECT CHAR(257)
-- NULL

3.CHARINDEX: It returns the starting position of the character expression present in another character expression. Basically it takes three arguments. The first argument is the character expression which need to be searched ,The second argument is the character expression in which the expression to be searched. The third argument is optional. It tells the starting position at which the search will begin.

Syntax:
CHARINDEX( character expression1, character expression2 [, starting position])

Examples:

SELECT CHARINDEX('IT','IT JUNCTION 4 ALL IT PEOPLE')
--1

SELECT CHARINDEX('IT','IT JUNCTION 4 ALL IT PEOPLE', 2 )
--19





4. LEFT: It returns specified number of characters from left part of the character expression. It takes two arguments. First argument is the string or character expression. Second argument is numeric expression which tells how many characters to be extracted from left.

Syntax:
LEFT( character expression , numeric expression )

Examples:

SELECT LEFT('Sunil',3) AS LEFT_Demo
-- Sun

--Declaring local variable of VARCHAR type
DECLARE @Name VARCHAR(50)
--Assigning value to the local variaible
SET @Name = 'Sunil Kumar'
--Using LEFT function to extract First Name
SELECT LEFT(@Name, CHARINDEX(' ', @Name) - 1) AS [FirstName]
--Sunil



5. RIGHT : It returns specified number of characters from right part of the character expression. It takes two arguments. First argument is the string or character expression. Second argument is numeric expression which tells how many characters to be extracted from right.

Syntax:
RIGHT( character expression , numeric expression )

Examples:

SELECT RIGHT('Sunil',3) AS RIGHT_Demo
-- nil

--Declaring local variable of VARCHAR type
DECLARE @Name VARCHAR(50)
--Assigning value to the local variaible
SET @Name = 'Sunil Kumar'
--Using RIGHT function to extract Last Name
SELECT RIGHT(@Name, CHARINDEX(' ', @Name) -1) AS [LastName]
--Kumar





6.REVERSE: REVERSE function reverse the string. It takes only one argument which is the string which need to be reversed.

Syntax:
REVERSE(String )


Examples:
SELECT REVERSE('IT JUNCTION 4 ALL')
--LLA 4 NOITCNUJ TI

DECLARE @Name VARCHAR(100) = 'Sunil Kumar Gupta'
SELECT RIGHT(@Name, CHARINDEX(' ', REVERSE(@Name)) - 1) AS [LastName]
GO
--Gupta

DECLARE @CheckPalindrome VARCHAR(100) = 'Malayalam'
SELECT CASE WHEN @CheckPalindrome=REVERSE(@CheckPalindrome)
       THEN 'YES'
       ELSE 'NO' END AS CheckPalindrome
GO
--YES





7. LEN: It returns the number of characters in a character expression. It takes character expression or string as an argument. Leading space (space that appears at the start of the string ) is taken into consideration while trailing space (space that appears at the end of the string )is ignored.

Syntax:
LEN( character expression )

Examples:
SELECT LEN('Sunil') AS LEN_Demo1
--5
SELECT LEN('     Sunil') AS LEN_Demo2
--10
SELECT LEN('Sunil     ') AS LEN_Demo3
--5
SELECT LEN('Sunil Kumar') AS LEN_Demo4
--11




8. REPLACE: It replaces all the occurrences of a specified pattern of string with desired string  in another string. It takes three arguments. The first argument  is the source string. The second argument is the string which is to searched in source string and the third is the replacement string .


Syntax:
REPLACE( source string expression , search string , replacement string)

Examples:
SELECT REPLACE('XYZABCXYZABC','XYZ','ABC') AS REPLACE_Demo1
--ABCABCABCABC

SELECT REPLACE('Sea fish are found in sea','sea','River') AS REPLACE_Demo2
--River fish are found in River




9. STUFF: It deletes specified length of characters in the source string at the specified  position and inserts another string at the specified position.


Syntax:
STUFF( source string expression ,length, start position, string)

Examples:

SELECT STUFF('Sunil Kumar',3,3,'man') AS STUFF_DEMO
--Suman Kumar

DECLARE @Date VARCHAR(20)
SET @Date = '20071988'
SELECT STUFF(STUFF(@Date,3,0,'/'),6,0,'/') AS [DD/MM/YYYY]
--20/07/1988

DECLARE @Time VARCHAR(20)
SET @Time = '1120'
SELECT STUFF(@Time,3,0,':') AS [HH:MM]
--11:20




10. SUBSTRING: It returns the part of the string. It takes three arguments. First argument is the source string or character or text. Second argument is the start position from where string will be returned. Third argument is the length which tells how many character to be returned from start position.


Syntax:
SUBSTRING( source string expression , start position, length)

SELECT SUBSTRING('sunil',3,5)
--nil SELECT SUBSTRING('SQLServer2008',10,4)
--2008 DECLARE @Name VARCHAR(20)
SET @Name='Sunil Kumar'
SELECT 
SUBSTRING(@Name,1,CHARINDEX(' ',@Name)-1)AS FirstName,
SUBSTRING(@Name,CHARINDEX(' ',@Name)+1,LEN(@Name))AS LastName




11. RTRIM: RTRIM function removes the trailing spaces or blanks from the  string or character expression.

Syntax:
RTRIM( string expression )

Examples:
SELECT RTRIM('Sunil Kumar   ')
--Sunil Kumar

DECLARE @string_2_trim varchar(100);
SET @string_2_trim = 'At the end this sentence,six spaces are there.      ';

SELECT @string_2_trim + ' Starting with next string.';
--At the end this sentence,six spaces are there.       Starting with next string.


SELECT RTRIM(@string_2_trim) + ' Starting with next string.';
--At the end this sentence,six spaces are there. Starting with next string.





















12. LTRIM LTRIM function removes the leading spaces or blanks from the  string or character expression.

Syntax:
LTRIM( string expression )

Examples:
SELECT LTRIM('     Sunil Kumar')
--Sunil Kumar

DECLARE @string_2_trim varchar(100);
SET @string_2_trim = 'First string without any space.';

SELECT @string_2_trim +'      Second string with six space in beginning.';
--First string without any space.      Second string with six space in beginning.

SELECT @string_2_trim +LTRIM('      Second string with six space in beginning.');

--First string without any space.Second string with six space in beginning.




13. SPACE: SPACE function returns string of repeated spaces. It takes integer expression  as the argument which indicates number of spaces .

Syntax:
SPACE( integer expression )


Examples:
SELECT 'Sunil'+'Kumar'
--SunilKumar
SELECT 'Sunil'+SPACE(1)+'Kumar'
--Sunil Kumar

SELECT 'Hello'+SPACE(2)+'How'+SPACE(1)+'are'+SPACE(1)+'you'+SPACE(1)+'?'
--Hello  How are you ?






14.UPPER: UPPER converts all lower case characters into upper case characters in a string.

Syntax:
UPPER( String )


Examples:

SELECT UPPER('sunil kumar')
--SUNIL KUMAR


SELECT UPPER('Sunil Kumar')
--SUNIL KUMAR





15.LOWER: LOWER converts all upper case characters into lower case characters  in a string.

Syntax:
LOWER( String  )


Examples:
SELECT LOWER('SUNIL KUMAR')
--sunil kumar


SELECT LOWER('Sunil Kumar')
--sunil kumar




16.REPLICATE: REPLICATE repeats string or character expression specified number of times.

Syntax:
REPLICATE(String,integer )


Examples:
SELECT REPLICATE('Sunil',4)
--SunilSunilSunilSunil




17.PATINDEX: PATINDEX returns the starting position of first occurrence of specified pattern in specified string. If the pattern is not found, zero is returned.
     If you want to search particular pattern use '%' before and after the pattern. If you want pattern to be your first characters in the string or in a column, do not give '%'character in the beginning of the pattern. If you want pattern to be your last characters in the string or in a column, do not give '%'character in the end of the pattern.
    Both CHARINDEX and PATINDEX is used to find the starting position of the first occurrence of pattern in a string. Both takes two arguments . The difference which lies between them is that we can use wildcard characters in pattern in PATINDEX but we can not use wildcard characters in CHARINDEX.

Syntax:
PATINDEX('%patteren_to_search%',string)


Examples:
SELECT PATINDEX('%NCT%','SUNIL IT JUNCTION 4 ALL SUNIL')
--12
SELECT PATINDEX('%SUNIL%','SUNIL IT JUNCTION 4 ALL SUNIL')
--1
SELECT PATINDEX('SUNIL%','SUNIL IT JUNCTION 4 ALL SUNIL')
--1
SELECT PATINDEX('%SUNIL','SUNIL IT JUNCTION 4 ALL SUNIL')
--25




1 September 2013

SQL-NO LOCK

Q Why and when should one use NO LOCK in SQL query ?

 "WITH (NO LOCK) " is used in SQL Server query to speed up the execution of query. This is particularly used when we are running a long query which fetches excessively a large number of records from a table and we are not much bothered about data integrity.  What exactly happens is that when we fire a SQL query ,database puts a lock on a table. So one process who is in need of  the same table will have to wait till the other process completes reading or writing of a table thus making the query slower. In such scenario where data integrity is not a problem , we can use NO LOCK to avoid lock on a table. Thus this process will read the data without applying lock on a table and also helping other processes as they do not have to wait.
       Use "WITH(NO LOCK) in the from clause just after the table name as shown below.

 SELECT Name, ProductName,Quantity, MRP
 FROM CustomerInfo C WITH(NOLOCK)
 LEFT OUTER JOIN ProductInfo P WITH(NOLOCK)
ON C.ProductID=P.ProductID



WITH(NOLOCK) Illustration

26 August 2013

SQL- Shortcut way to delete duplicate records from a table

I have discussed earlier in one of my post how to delete duplicate records from a table . But recently i came up with even shorter way to delete duplicate records.

Consider a StudentInfo Table as shown  below for demonstration.

SELECT * FROM StudentInfo


StudentInfo Table
Step1:
 First we will create a temporary table '#temp' with same structure  as that of original table 
'StudentInfo' and insert all distinct rows from the 'StudentInfo' table into temporary table as follows: