Is there an elegant way to store a dual relationship (i.e. user 1 and user 2 are friends)

Question

I\'ve run into the same problem in two different pieces of work this month:

Version 1: User 1 & User 2 are friends
Version 2: Axis 1 & Axis 2 when gr         


        

Answer 1

You seem to limit the number of friends to 1. If this is the case then I would use something like u1,u2 u2,u1 u3,null u4,u5 u5,u4

u3 does not have a friend.

Answer 2

There is also a way to use the 2nd approach by adding an extra constraint. Check that u1 < u2:

CREATE TABLE User
( Name VARCHAR(10) NOT NULL
, PRIMARY KEY (Name)
) ;

CREATE TABLE MutualFriendship
( u1 VARCHAR(10) NOT NULL
, u2 VARCHAR(10) NOT NULL
, PRIMARY KEY (u1, u2)
, FOREIGN KEY (u1) 
    REFERENCES User(Name)
, FOREIGN KEY (u2) 
    REFERENCES User(Name)
, CHECK (u1 < u2) 
) ;

The rules to read, create, insert or update will have to use the (LEAST(u1,u2), GREATEST(u1,u2)).

Answer 3

"x is a friend of y".

Define a table of (x,y) pairs and enforce a canonical form, e.g. x<y. This will ensure that you cannot have both (p,q) and (q,p) in your database, thus it will ensure "store once".

Create a view as SELECT x,y FROM FRIENDS UNION SELECT x as y, y as x FROM FRIENDS.

Do your updates against the base table (downside : updaters must be aware of the enforced canonical form), do your queries against the view.

Answer 4

For anyone that's interested, I played around with a few bitwise operations and found that the following seems to fulfill the criteria for f(x,y):

#Python, returns 3 tuple
def get_hash(x, y):
  return (x & y, x | y, x * y)

I can't prove it, though.

Answer 5

In SQL it's easy to implement the constraints to support your first approach:

CREATE TABLE MutualFriendship
(u1 VARCHAR(10) NOT NULL,
 u2 VARCHAR(10) NOT NULL,
 PRIMARY KEY (u1,u2),
 FOREIGN KEY (u2,u1) REFERENCES MutualFriendship (u1,u2));

INSERT INTO MutualFriendship VALUES
('Alice','Bob'),
('Bob','Alice');