Few doubts on writing SQL queries - sql

Question

Few questions i have here
I always see some SQL written like below (not sure if i get it right)
SELECT a.column_1, a.column_2 FROM table_name WHERE b.column_a = 'some value'
i don't quite understand the SQL written in such way. Is it similar to using object in programming, where you can define an object and variables within the object? If it is, where is the definition of a and b for the SQL above (assuming i got the query right)?
I want to make comparisons between 3 columns (say C1 C2 C3) in 3 different tables, say T1 T2 and T3. The condition is to get the values from the C1 in T1, that exists in C2 in T2, but not exists in C3 in T3. Both columns are practically the same, just that some might different or lesser records than the other columns in the other 2 tables, and i want to know what the differences are. Is the query below the right way to do it?
select distinct C1 from T1
and (C1) not in (select C2 from T2)
and (C1) in (select C3 from T3)
order by C1;
And is it possible to extend the condition if i want to include more tables into comparison using the query above?
If i were to customize the query above into something similar to the first question, is the query below the right way to do it?
select a.C1 from T1 a
and (a.C1) not in (select b.C2 from T2 b)
and (a.C1) in (select c.C3 from T3 c)
order by a.C1;
What are the advantanges of writing query in object way (like above), compared to writing it in traditional way? I feel like even if you define a table name as a variable, the variable only can be used within the query where it is defined, and cannot be extended to the other queries.
Thanks

Answer 1

the first point is a and b are "table aliases" (shortcut reference to the table(s) involved in THAT query) e.g.
SELECT a.column_1, a.column_2
FROM table_name_a a ------------------------------- table alias a defined here
INNER JOIN table_name_b b -------------------------- table alias b defined here
ON a.id = b.id
WHERE b.column_a = 'some value'
Your second query has a syntax issue: You need WHERE as shown in uppercase. It also has and performance implications. Distinct adds effort to a query, using IN() is really a syntax shortcut for a series of ORs (it might not scale well). But with the syntax it is valid.
select distinct C1
from T1
WHERE (C1) not in (select C2 from T2)
and (C1) in (select C3 from T3)
order by C1;
Yes (with performance reservations) you could add more tables into that comparison.
You introduce table aliases, done correctly, into your third query - but there is no real advantage in that query structure. Aside from just making code more convenient, aliases serve to distinguish between items that would be ambiguous. In my first query above ON a.id = b.id shows possible ambiguity in that 2 tables both have a field of the same name. Prefixing the field name by a table or table alias solves that ambiguity.

Answer 2

For your first point.
I always see some SQL written like below (not sure if i get it right)
SELECT a.column_1, a.column_2 FROM table_name WHERE b.column_a = 'some value'
This query is wrong. It should be like this -
SELECT a.column_1, a.column_2
FROM table_name a INNER JOIN --(There might be another join also like left join etc..)
table_name b
ON a.id = b.id WHERE b.column_a = 'some value'
so you noted in the above query that a and b are just table alias. Well, there are some cases you must use them, like when you need to join to the same table twice in one query.
For the second point. you can also do it like this
SELECT DISTINCT C1 FROM T1 t1
WHERE NOT EXISTS (
SELECT C2 FROM T2 t2 where t2.C2 = t1.C1)
AND WHERE EXISTS (
SELECT C3 FROM T3 t3 where t3.C3 = t1.C1)
ORDER BY C1;
Personally I prefer aliases, and unless I have a lot of tables they tend to be single letter ones.

Answer 3

I'm not 100% sure on this one so don't quote me, but ill give it my best shot.
I think that when alias' are used it is because if you don't use them your statement can he huge and hard to understand, here is a comparison with two SQL query's, one using alias' and the other without:
SELECT o.OrderID, o.OrderDate, c.CustomerName
FROM Customers AS c, Orders AS o
WHERE c.CustomerName="Around the Horn" AND c.CustomerID=o.CustomerID;
Without:
SELECT Orders.OrderID, Orders.OrderDate, Customers.CustomerName
FROM Customers, Orders
WHERE Customers.CustomerName="Around the Horn" AND Customers.CustomerID=Orders.CustomerID;
As you can see one looks much easier to understand and read than the other and makes your query's much smaller.
Aliases can be useful when:
There are more than one table involved in a query
Functions are used in the query
Column names are big or not very readable
Two or more columns are combined together
http://www.w3schools.com/sql/sql_alias.asp

Answer 4

You have to define the table aliases in the FROM clause, so:
SELECT a.column_1, b.column_1
FROM table1 a, table2 b
WHERE a.id = b.id
It should be possible to make your comparisons in the way you have written your query above, however it is also possible to make aliasing easier by using subqueries in the FROM clause ie:
SELECT tab1.id, tab2.id, tab3.id
FROM table1 as tab1,
(select * from table2)as tab2,
(select * from table3)as tab3
This way you can choose any columns from any of the tables using tab1.xxx etc. and then use the WHERE clause to say NOT IN tab2.column_1 etc...