Insert data to column which has two relationships - sql

Question

So, what I want to do in here is I want to insert the data to table1 who has 'username' field, and this 'username' field has two relationships, which is go to student table and teacher table. But, when I insert the data to user table, I have a problem, and the problem caused by this 'username' field. It caused because the data of 'username' field is not the same as a unique key in student table, and when I change it and made the data same, I get an error too, but this time the data is not the same as a unique in teacher table. So, is it possible to make this 'username' field just get one of the table, like if one of student table and teacher table's data is in 'username', it still can be used. Or maybe this is wrong because of bad ERD? Here's my ERD, if you're asking it:
And well, I know this is really, really bad idea but I made the name in teacher table and student table become a unique key, because I can't create foreign key if I didn't do that. Please, I'm really thankful to your answer.
Here's the ddl for that 3 tables :
Student table :
CREATE TABLE [dbo].[student](
[studentid] [int] IDENTITY(2016000001,1) NOT NULL,
[name] [varchar](50) NOT NULL,
[address] [text] NOT NULL,
[gender] [varchar](7) NOT NULL,
[dateofbirth] [date] NOT NULL,
[nohp] [varchar](13) NOT NULL,
CONSTRAINT [PK_student] PRIMARY KEY CLUSTERED
(
[studentid] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY],
CONSTRAINT [IX_student] UNIQUE NONCLUSTERED
(
[name] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
) ON [PRIMARY] TEXTIMAGE_ON [PRIMARY]
GO
Teacher Table :
CREATE TABLE [dbo].[teacher](
[teacherid] [int] IDENTITY(1,1) NOT NULL,
[name] [varchar](50) NOT NULL,
[gender] [varchar](7) NOT NULL,
CONSTRAINT [PK_teacher] PRIMARY KEY CLUSTERED
(
[teacherid] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY],
CONSTRAINT [IX_teacher] UNIQUE NONCLUSTERED
(
[name] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
) ON [PRIMARY]
GO
User table:
CREATE TABLE [dbo].[user](
[userid] [int] IDENTITY(1,1) NOT NULL,
[username] [varchar](50) NOT NULL,
[password] [varchar](20) NOT NULL,
[role] [varchar](10) NOT NULL,
CONSTRAINT [PK_user] PRIMARY KEY CLUSTERED
(
[userid] ASC
)WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [PRIMARY]
) ON [PRIMARY]
GO
ALTER TABLE [dbo].[user] WITH CHECK ADD CONSTRAINT [FK_user_student] FOREIGN KEY([username])
REFERENCES [dbo].[student] ([name])
ON DELETE CASCADE
GO
ALTER TABLE [dbo].[user] CHECK CONSTRAINT [FK_user_student]
GO
ALTER TABLE [dbo].[user] WITH CHECK ADD CONSTRAINT [FK_user_teacher] FOREIGN KEY([username])
REFERENCES [dbo].[teacher] ([name])
ON DELETE CASCADE
GO
ALTER TABLE [dbo].[user] CHECK CONSTRAINT [FK_user_teacher]
GO

Answer 1

The main issue you faced is because your data structure was less than ideal. It did things like prevent you from changing somebody's name. I tossed together a quick example of a cleaner design. There are many assumptions here. I assumed that people have both a first and last name. I would be remiss if I didn't at least point out that assumption is not something you can always make. https://www.kalzumeus.com/2010/06/17/falsehoods-programmers-believe-about-names/ But for a school project it is more than sufficient. I also made the assumption that the addresses are US addresses. Again, this would not work in many real world scenarios. And the last assumption is that everybody can be either Male or Female. In the world today this is not always the case but demonstrates the technique well enough.
Here is how I would probably do this type of design. I would suggest you not just blindly copy this but use it as an idea to get your design more properly normalized.
create table Users
(
UserID int identity not null
, FirstName varchar(50) not null
, LastName varchar(50) not null
, AddressLine1 varchar(50)
, AddressLine2 varchar(50)
, City varchar(50)
, ST char(2)
, ZipCode varchar(9)
, Gender char(1)
, constraint PK_Users primary key clustered
(
UserID
)
, constraint CHK_Users_Gender
CHECK (Gender in ('M', 'F'))
, constraint CHK_Users_ZipCode
CHECK (LEN(ZipCode) in (5,9)) --This ensures you have either the 5 or 9 digiti zip code
)
create table Student
(
StudentID int identity not null
, UserID int not null
, BirthDate date
, constraint PK_Student primary key clustered
(
StudentID
)
, constraint FK_Student_Users foreign key (UserID) references Users(UserID)
)
create table Teacher
(
TeacherID int identity not null
, constraint PK_Teacher primary key clustered
(
TeacherID
)
, constraint FK_Teacher_Users foreign key (TeacherID) references Users(UserID)
)