Lab 8 Solutions

Exercise 1
Q.1 ) Create a trigger on the table employees, which after an update or insert, converts all the
values of first and last names of the updated or inserted rows to upper case. (Hint: Use cursors
to retrieve the values of each row and modify them.)
Solution:
>> create trigger converttoupper on employees

after update , insert
as
declare employee_cursor CURSOR FOR SELECT firstname, lastname
from inserted;
declare @fname varchar(100), @lname varchar(100);
OPEN employee_cursor;
fetch next from employee_cursor into @fname,@lname;
WHILE @@FETCH_STATUS = 0 begin

update employees set firstname = upper(@fname), lastname =
upper(@lname) where CURRENT OF employee_cursor;
FETCH NEXT FROM employee_cursor INTO @fname,@lname;

end;
-- close and deallocate cursors

close employee_cursor;
deallocate employee_cursor;
go
DROP INDEX employees.i_func_employees_lastname;
2) Create a trigger that restores the values before an update operation on the employees table
if the salary exceeds 100000. (Hint: Use the inserted and deleted tables to look at the old and
new values in the table respectively.)
Solution :
>> create trigger salary_checker on employees

instead of UPDATE
as
begin transaction trans;
save transaction check_1;

if exists(select * from inserted where salary > 100000)
begin RAISERROR ('salary > 100000 not allowed', 16, 1);
rollback transaction check_1;
end;
commit transaction trans;

go;
3) Create a trigger to insert into the view alex such that the underlying base tables are populated
correctly. Handle cases where the input is incorrect and rollback if a wrong input is asked to be
inserted. Assume that you cannot create a new department by an insert query.
>> create trigger tr_alex_insert

on alex
instead of insert
as
begin transaction temp;
-- Just in case we get an incorrect tuple to insert

SAVE TRANSACTION p1;
declare @employeeid numeric(9);

declare @firstname varchar(10), @lastname varchar(20);
declare @code char(5);
declare @salary numeric(9,2);
select @employeeid = employeeid from inserted;

select @firstname = firstname from inserted;
select @lastname = lastname from inserted;
select @code = code from inserted;
select @salary = salary from inserted;
-- Ensure that primary key constraint is followed and

Ensure that department code is valid
if ( @employeeid in (select employeeid from employees) ) or
(@code not in (select code from departments) ) begin
rollback transaction p1;
end
else begin
insert into employees
values(@employeeid,@firstname,@lastname,@code,@salary;
end
commit transaction temp;
go
4) Create a trigger to delete from the view alex such that the corresponding rows in the
underlying base tables are removed correctly. Handle cases where the rows requested to be
deleted are not possible and rollback accordingly. Assume that a department cannot be left
without a manager.
>> create trigger tr_alex_delete

on alex
instead of delete
as
-- make sure to mention the name of the transaction
begin transaction temp
save transaction t1;
declare @id numeric(9);

declare @code char(5);
declare @first_name varchar(10), @last_name varchar(20);
declare @salary numeric(9,2);
declare @int_1 numeric(9);

select @int_1 = count(*) from (select distinct managerid from

departments except select employeeid from deleted) d2;
select @int_2 = count(*) from (select distinct managerid from

departments) d1;
-- if any of the employees are managers do nothing

if (@int_2 > @int_1) begin
rollback transaction t1;
end
-- else continue with the deletion
else begin
declare delete_cursor CURSOR for select * from deleted;
open delete_cursor;
fetch next from delete_cursor into
@id,@first_name,@last_name,@salary,@code;
while @@FETCH_STATUS = 0
begin
-- delete all workson entries
delete from workson where employeeid = @id;
-- delete all employee tables

delete from employees where employeeid = @id;
-- fetch next
@id,@first_name,@last_name,@salary,@code;
end
-- close and deallocate the cursors

close delete_cursor;
deallocate delete_cursor;
end

commit transaction temp;
go
5) Create a trigger to modify from the view alex such that the corresponding rows in the
underlying base tables are changed correctly. Handle cases where the modifications asked for are
not possible and rollback accordingly. Assume that a department cannot be left without a
manager.
>> create trigger tr_alex_update

on alex
instead of update
as
begin transaction trans;
save transaction check_1;

select @int_1 = count( distinct employeeid ) from inserted;

select @int_2 = count(*) from inserted;
select @int_3 = count(*) from (select employeeid from inserted
where employeeid in (select employeeid from employees)) d1;
-- check if the primary keys are same for any two rows
-- check if primary key does not already exist
if( (@int_1 != @int_2) or (@int_3 > 0) ) begin
rollback transaction check_1;
end
else begin
declare @insert_id numeric(9);
declare @insert_code char(5);
declare @insert_first_name varchar(10), @insert_last_name
varchar(20);
declare @insert_salary numeric(9,2);
declare @delete_id numeric(9);

declare @delete_code char(5);
declare @delete_first_name varchar(10), @delete_last_name
varchar(20);
declare @delete_salary numeric(9,2);
select * from inserted;

select * from deleted;
declare insert_cursor cursor for select * from inserted;

open insert_cursor;
fetch next from insert_cursor into
@insert_id,@insert_first_name,@insert_last_name,@inser
t_salary,@insert_code;
declare delete_cursor CURSOR for select * from deleted;

open delete_cursor;
@delete_id,@delete_first_name,@delete_last_name,@delete_sal
ary,@delete_code;
while @@FETCH_STATUS = 0 begin
-- insert the new row

insert into employees
values(@insert_id,@insert_first_name,@insert_last_name
,@insert_code,@insert_salary);
-- modify the foreign key relation values

update workson set employeeid=@insert_id where
employeeid=@delete_id;
update departments set managerid=@insert_id where
managerid=@delete_id;
-- delete the old row

delete from employees where employeeid=@delete_id;
-- fetching the new rows

fetch next from insert_cursor into
@insert_id,@insert_first_name,@insert_last_name,@inser
t_salary,@insert_code;
@delete_id,@delete_first_name,@delete_last_name,@delet
e_salary,@delete_code;
end
-- close and deallocate cursor

close delete_cursor;
deallocate delete_cursor;
end
commit transaction trans;

go
Exercise 2
1. Create a unique composite non-clustered index on the first_name and order in the
descending order.
>> create UNIQUE Nonclustered index IX_student_info on

student_info(first_name asc);
2. Create a composite clustered index on the first_name and the last_name and order both
in the increasing order.
>> create clustered index IX_student_info2 on

student_info(first_name asc, last_name asc);
3. Create a unique Clustered index on the age and order in the increasing order
>> drop index student_info.IX_student_info2 create UNIQUE
clustered index IX_student_info3 on student_info(age asc);
4. Create composite non-clustered index on the id,age both in the increasing order
>> create Nonclustered index IX_student_info4 on student_info(id

asc ,age asc);
5. Create a non-clustered unique function based index on the upper(first_name + ' ' +
lastname)
>> alter table student_info add complete_name as upper(first_name

+ ' ' + last_name);
create unique Nonclustered index IX_student_info5 on
student_info(complete_name);

Lab 8 Solutions

Uploaded by

Copyright:

Available Formats

Lab 8 Solutions

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lab 8 Solutions

Uploaded by

Copyright:

Available Formats

Exercise 1

>> create trigger converttoupper on employees

WHILE @@FETCH_STATUS = 0 begin

FETCH NEXT FROM employee_cursor INTO @fname,@lname;

-- close and deallocate cursors

DROP INDEX employees.i_func_employees_lastname;

>> create trigger salary_checker on employees

save transaction check_1;

commit transaction trans;

>> create trigger tr_alex_insert

-- Just in case we get an incorrect tuple to insert

declare @employeeid numeric(9);

select @employeeid = employeeid from inserted;

-- Ensure that primary key constraint is followed and

>> create trigger tr_alex_delete

save transaction t1;

declare @id numeric(9);

declare @int_1 numeric(9);

select @int_1 = count(*) from (select distinct managerid from

select @int_2 = count(*) from (select distinct managerid from

-- if any of the employees are managers do nothing

-- delete all employee tables

-- close and deallocate the cursors

-- make sure to mention the name of the transaction

>> create trigger tr_alex_update

save transaction check_1;

declare @int_1 numeric(9);

select @int_1 = count( distinct employeeid ) from inserted;

declare @delete_id numeric(9);

select * from inserted;

declare insert_cursor cursor for select * from inserted;

declare delete_cursor CURSOR for select * from deleted;

while @@FETCH_STATUS = 0 begin

-- insert the new row

-- modify the foreign key relation values

-- delete the old row

-- fetching the new rows

-- close and deallocate cursor

commit transaction trans;

>> create UNIQUE Nonclustered index IX_student_info on

>> create clustered index IX_student_info2 on

>> create Nonclustered index IX_student_info4 on student_info(id

>> alter table student_info add complete_name as upper(first_name

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.