Difference between SQL Server codes? - sql

Question

For a school assignment, I have to create a database and run reports. I've created code and a classmate has also created code and it runs the same thing, but his is in a format I've not seen and don't quite understand.
Here is mine:
SELECT
Course.Name AS 'Course Name',
Program.Name AS 'Program Name'
FROM
Course, Program, ProgramCourse
WHERE
ProgramCourse.CourseID = Course.ID
AND
ProgramCourse.ProgramID = Program.ID
GO
And here's his:
CREATE VIEW NumberOfCoursePerProgram AS
SELECT
p.name AS ProgramName,
c.name AS CourseName
FROM
Program p
JOIN
ProgramCourse pc ON pc.ProgramID = p.ID
JOIN
Course c ON c.ID = pc.CourseID
GO
I ran both queries using the data in the tables I've created. They return practically the same results, just in a slightly different order but it fulfills the assignment question. Anyway, if I delete the p from Program p from his code, it returns an error
The multi-part identifier "p.name" could not be bound.
So how is SQL Server able to accept p.name and p.ID, etc. when I haven't ever established these variables? I don't quite understand how the code is working on his. Mine seems simple and straightforward, and I definitely understand what's going on there. So can someone explain his?
Thanks

Answer 1

There's a few differences. First off, he's creating a VIEW rather than just a select statement:
CREATE VIEW NumberOfCoursePerProgram AS
Once the view is created, you can query the view just as you would a table:
SELECT * FROM NumberOfCoursePerProgram;
Second, he's using an ANSI JOIN rather than an implicit JOIN. His method is more modern and most likely considered more correct by today's standards:
JOIN ProgramCourse pc ON pc.ProgramID = p.ID
JOIN Course c ON c.ID= pc.CourseID
Rather than:
FROM Course, Program, ProgramCourse
Also, note he's assigning table aliases when he refers to a table:
FROM Program p
The p at the end allows you to substitute p rather than specify the entire table name of Program elsewhere in the query. For example, you can now say WHERE p.Foo > 5 rather than WHERE Program.Foo > 5. In this case, it's just a shortcut and saves a few characters. However, suppose you were referring to the same table twice (for example, JOINing in two different rows on the same table). In that case, you might have to provide aliases for each table to disambiguate which one is which.

Answer 2

These are called alias in SQL. Alias is basically created to give more readability and for better ease of writing code.
The readability of a SELECT statement can be improved by giving a
table an alias, also known as a correlation name or range variable. A
table alias can be assigned either with or without the AS keyword:
table_name AS table alias
table_name table_alias
So in your query p is an alias to Program so that means now you can refer your table Program by the name of p instead of writing the whole name Program everywhere.
Similarly you can access the names of the columns of your table Program by simply writing p with a dot and then the column name. Something like p.column. This technique is very useful when you using JOINS and some your tables have similar names of the columns.
EDIT:-
Although most of the points are covered in other's answer. I am just adding a point that you should avoid the habit of JOINING table the way you are doing it right now.
You may check Bad habits to kick : using old-style JOINs by Aaron Bertrand for reference.

Answer 3

CREATE VIEW NumberOfCoursePerProgram AS
BEGIN
SELECT
p.name AS ProgramName,
c.name AS CourseName
FROM
Program p
JOIN
ProgramCourse pc ON pc.ProgramID= p.ID
JOIN Course c ON c.ID= pc.CourseID
END
GO
Observe that both tables Program and Course have a table alias defined.
The select part must specify the table from which the column name comes from. Which is exactly what you did. Your partner just added aliases to the tables names. These aliases are shorter, and makes the query look a bit less like a big wall of text.
The other difference is the use of joints. The joins are usually used to link results from two tables that has a corresponding column.
The columns are usually the primary key, and the foreign key in the second table.
Your query is fine, but the join syntax is preferred.
Edit : Once the view is created, it is now compiled and can be used in a select just like a table.
SELECT * FROM NumberOfCoursePerProgram
If you need to modify the view, you can use
ALTER VIEW NumberOfCoursePerProgram AS
......
......

Answer 4

There are a few things going on here:
if I delete the "p" from "Program p" from his code, it returns "The multi-part identifier "p.name" could not be bound."
The P is an alias for program. By doing it that way. If you ever need to use Program you can just call it P.
As far as what he is doing: it's better to do a JOIN in this case as opposed to doing a multi-table SELECT that you are doing. In simple cases it doesn't really matter but what if course and program both had millions of rows. In your way you are basically selecting everything. By doing the join on ProgramID you are only getting those items in ProgramCourse that correspond to an entry in Program as well (tied together by ID and CourseID).
One more important note. You are doing a simple SELECT statement. In SQL there are objects called VIEWS that act as a virtual table. He can now do any time a SELECT * FROM NumberOfCoursePerProgram and he will never have to do any of the joins, and selects again.
Hope that helps...

Answer 5

In SQL Server you can give a table an alias without AS, its optional.
If you were to add the AS it would will work just like you have it.
It may also be worthy to note that he is using a join as opposed to a where clause join, which is my preference because it's easier to read and more update code ethic.
Sometimes you need to join on a WHERE clause because of multiple conditions on join, but that's pretty rare in your case.