Andreas Ntaflos <[EMAIL PROTECTED]> wrote: > this question is about the AUTOINCREMENT usage in SQLite when > modelling weak entities. Suppose there are two tables 'Building' and > 'Room': > > CREATE TABLE Building ( > buildingID INTEGER PRIMARY KEY AUTOINCREMENT, > ... > ); > > CREATE TABLE Room ( > roomID INTEGER, > buildingID INTEGER CONSTRAINT fk_room_building_id REFERENCES > Building(buildingID), > PRIMARY KEY(roomID, buildingID) > ); > > CREATE TRIGGER statements follow to enforce the foreign key > constraint. > > Naturally a room cannot be identified without a building so it is a > weak entity (this seems to be the canonical example in all database > books I've seen). > > I'd like to auto-increment the ID numbers in both tables but as far > as I can see the AUTOINCREMENT keyword only works the way it is used > in the Building table. That is, only "INTEGER PRIMARY KEY > AUTOINCREMENT" is syntactically correct.
Wait: if you want roomID to be autoincrement, then it's going to be unique (which, after all, is the point of autoincrement). Run it by my again: why can't you identify a room without a building? Igor Tandetnik _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
