Compare rows and columns of same table

Question

+--------------+--------------+------------+
| company_name | address_type |  address   |
+--------------+--------------+------------+
| Company A    | Billing      | 111 Street |
| Company A    | Shipping     | 111 Street |
| Company B    | Billing      | 222 Street |
| Company B    | Shipping     | 333 street |
| Company B    | Shipping     | 444 street |
+--------------+--------------+------------+

I have a table similar to this.
What i need is All the companies whose billing address and shipping address are different .

NOTE - Each company has only ONE billing address. But it can have multiple Shipping addresses

This seems like a fairly simple query but I'm not just able to get it.

My attempt - I tried 'subtracting' all the shipping address from Billing but there's just no output. Distinct doesn't help as well

Query:

select company_name 
from tableA 
where address_type='Billing' 
and company_name not in (select to_char(company_name) from tableA where address_type='Shipping');

Output should be Company B (since it's billing and shipping address is different)

EDIT 1 : Tried Indra's query but it runs forever. No response

select A.* from company A inner join company B on A.company_Name = B.company_Name
and (A.address_type = 'Billing' and B.address_type = 'Shipping') 
AND A.address <> B.address 

Answer 1

How about using a join ? The following shows all pairs that are different:

select tb.*, ts.*
from company tb join
     company ts
     on tb.company_name = ts.company_name and
        ts.address_type = 'shipping' and
        tb.address_type = 'billing' and
        ts.address <> tb.address;

If you just want the companies that are different:

select company_name
from company t
group by company_name
having count(distinct case when t.address_type = 'billing' then address end) = 1 and
       count(distinct case when t.address_type = 'shipping' then address end) = 1 and
       (max(case when t.address_type = 'billing' then address end) <>
        max(case when t.address_type = 'shipping' then address end)
       );

Note: this also checks that there is only one distinct billing and shipping address.

Answer 2

Use Below Query

select A.* from company A inner join company B on A.company_Name = B.company_Name
and (A.address_type = 'Billing' and B.address_type = 'Shipping') 
AND A.address <> B.address   

Answer 3

In this SQL Fiddle example , you can find other approach with one more company:

+--------------+--------------+------------+
| company_name | address_type |  address   |
+--------------+--------------+------------+
| Company A    | Billing      | 111 Street |
| Company A    | Shipping     | 111 Street |
| Company B    | Billing      | 222 Street |
| Company B    | Shipping     | 333 street |
| Company B    | Shipping     | 444 street |
| Company C    | Shipping     | 555 street |
| Company C    | Shipping     | 666 street |
| Company C    | Billing      | 555 street |
| Company C    | Billing      | 666 street |
+--------------+--------------+------------+

POSTDATA : SQL Fiddle works fine for me. This is all the code:

Schema:

CREATE TABLE companies
( 
  company_name  VARCHAR2(40),
  address_type  VARCHAR2(40),
  address    VARCHAR2(40)
);

insert into companies values ('Company A','Billing','111 Street' );
insert into companies values ('Company A','Shipping','111 Street' );
insert into companies values ('Company B','Billing','222 Street' );
insert into companies values ('Company B','Shipping','333 street' );
insert into companies values ('Company B','Shipping','444 street');
insert into companies values ('Company C','Billing','555 Street' );
insert into companies values ('Company C','Billing','666 Street' );
insert into companies values ('Company C','Shipping','555 Street' );
insert into companies values ('Company C','Shipping','666 Street' );

Sentece:

SELECT DISTINCT(comp.company_name) FROM
    (select company_name, address, count(*) regs
    from companies
    group by company_name, address) comp
WHERE MOD(comp.regs,2) = 1;

Result:

Company B