SQL in relationale Algebra übersetzen

Question

Ich lerne gerade, SQL-Abfragen in relationale Algebra zu übersetzen und umgekehrt.

Customer(customer_name, customer_street, customer_city) 
Loan(loan_number, branch_name, amount) 
Borrower(customer_name, loan_number) 
Account(account_number, branch_name, balance) 
Depositor(customer_name, account_number) 

Ich möchte Kunden finden, die in Fairfax leben, einen Kredit haben, aber kein Bankkonto haben.

CustomersInFFX = PROJECT customer_name (SELECT customer_city = 'Fairfax'(Customer)) 
LoanOwners = (CustomersInFFX NATURAL JOIN Borrower NATURAL JOIN Loan) 
AccountHolders = (Depositor NATURAL JOIN ACCOUNT) 
PROJECT customer_name (LoanOwners - AccountHolders) 

Macht das Sinn?

* SELECT: σ, PROJEKT: π, NATURAL JOIN: ⋈

Answer 1

Hier mein Versuch ist mit Standard-SQL:

WITH CustomersInFFX 
    AS 
    (
     SELECT * 
     FROM Customer 
     WHERE customer_city = 'Fairfax' 
    ), 
    LoanOwners 
    AS 
    (
     SELECT * 
     FROM CustomersInFFX 
      NATURAL JOIN Borrower 
      NATURAL JOIN Loan 
    ), 
    AccountHolders 
    AS 
    (
     SELECT * 
     FROM Depositor 
      NATURAL JOIN ACCOUNT 
    ) 
SELECT * 
    FROM LoanOwners 
EXCEPT CORRESPONDING DISTINCT 
SELECT * 
    FROM AccountHolders; 

Dies wird jedoch nur funktionieren, wenn die LoanOwners und AccountHolders Tabellen die gleiche haben Anzahl der Attribute mit dem gleichen Namen. Dies ist wahrscheinlich nicht der Fall, daher müssen Sie (SELECT) nur die erforderlichen Attribute projizieren, z.

SELECT customer_name 
    FROM LoanOwners 
EXCEPT DISTINCT 
SELECT customer_name 
    FROM AccountHolders;