Types of KEY in MySQL

 

KEY NAMES KEY DESCRIPTION
PRIMARY KEY Name of the column which uniquely identifies each row of the table with no null value in any row of the table.
UNIQUE KEY Name of the column which uniquely identifies each row of the table, but it can accept a null value for any row of the table.
COMPOSITE KEY The combination of two or more column representing as a unique for each row of the table after merging the columns. It acts like a single column in contest of the key.
SURROGATE KEY A key which has no buisness logic or which have no existance in the real world. for more clarity see NATURAL KEY
FOREIGN KEY Name of the column which have the reference of the other table by having the same value of any column other table.
CANDIDATE KEY Any key which can represent as unique for each row is called CANDIDATE KEY. Hence PRIMARY KEY, UNIQUE KEY and COMPOSITE KEY are a member of CANDIDATE KEY
NATURAL KEY A key which have a real existinence in business logic or have the real value in the world like AADHAR NO in India and SSN NO in USA 
   

Store all data in an array to remove duplicate record when using LEFT JOIN in MySql

Sometime you’d stuck like that you’ll need all other rows coming with left join as a duplicate row for all record, where you would need to have all these inside an array variable. For example if you have an tbl_emp and another tbl_contact, in tbl_contact you would have one foreign key emp_id.

You made a query to fetch all employee record with their contact nos, when you left join the query such like LEFT JOIN tbl_contact ON tbl_contact.emp_id=tbl_emp.emp_id, you will get record in associative array something like this.

array (size=3)
  0 => 
    array (size=3)
      'name' => string 'test1' (length=5)
      'age' => int 23
      'email' => string 'Good' (length=4)
  1 => 
    array (size=3)
      'name' => string 'test2' (length=5)
      'age' => int 24
      'email' => string 'Good' (length=4)
  2 => 
    array (size=3)
      'name' => string 'test2' (length=5)
      'age' => int 24
      'email' => string 'Best' (length=4)

 Here you are getting two record for test2 employee, because this employee has two email as a contact.

So, to simplify this you will need to have all contact nos inside a single variable key and the duplicate record to be removed, then here is the logic to do that.

 

 

  1. <?php
  2. $total_size=sizeof($arr);
  3. for($i=0;$i<$total_size;$i++){
  4. $arr[$i]['emails']=array();
  5. $arr[$i]['emails'][]=$arr[$i]['email'];
  6. $unset=0;
  7. foreach($arr[$i] as $key=>$value){
  8. if($key=='name' && $arr[$i]['name']==$arr[$i+1]['name']){
  9. $arr[$i]['emails'][]=$arr[$i+1]['email'];
  10. $unset=1;
  11. }
  12. }
  13. unset($arr[$i]['email']);
  14. if($unset==1){
  15. unset($arr[$i+1]);
  16. $i++;
  17. }
  18. }
  19. var_dump($arr);
  20. ?>

 

 

Now you’ll get the result something like that

array (size=2)
  0 => 
    array (size=3)
      'name' => string 'test1' (length=5)
      'age' => int 23
      'emails' => 
        array (size=1)
          0 => string 'Good' (length=4)
  1 => 
    array (size=3)
      'name' => string 'test2' (length=5)
      'age' => int 24
      'emails' => 
        array (size=2)
          0 => string 'Good' (length=4)
          1 => string 'Best' (length=4)

 

 

Difference between “WHERE” and ”HAVING” clause

where clause can not accept aggregate function like sum, count, etc. whereas having clause can accept the aggregate functions like sum, count, average, etc.

Let’s see it by example.

emp_salary_paid
paid_id emp_id  salary_given salary_month
1 1 1000 jan
2 2 1500 jan
3 3 2000 jan
4 1 1200 feb
5 2 1500 feb
6 3 1800 feb

 

emp_detail
emp_id emp_name
1 Sajid
2 John
3 Ahmad

 

Query:

  1. SELECT emp_detail.emp_name, emp_salary_paid.salary_month, emp_salary_paid.salary_given
  2. FROM emp_detail, emp_salary_paid
  3. WHERE emp_detail.emp_id=emp_salary_paid.emp_id

 

will produce result like this one. because no aggregate function is being used here.

emp_name salary_month salary_given
Sajid jan 1000
Sajid feb 1200
John jan 1500
John feb 1500
Ahmad jan 2000
Ahmad feb 1800
  1. SELECT emp_detail.emp_name, SUM(emp_salary_paid.salary_given) AS total_paid
  2. FROM emp_detail, emp_salary_paid
  3. GROUP BY emp_detail.emp_id
  4. WHERE SUM(emp_salary_paid.salary_given) >2000

This will produce an error, because aggregate function is being used with where clause. To overcome this, we use having clause like this one.

  1. SELECT emp_detail.emp_name, SUM(emp_salary_paid.salary_given)
  2. FROM emp_detail, emp_salary_paid
  3. GROUP BY emp_detail.emp_id
  4. HAVING SUM(emp_salary_paid.salary_given) > 2000

This will produce the correct result as follows:

emp_name total_paid
Sajid 9000
John 9000
Ahmad 9000