On 24-09-10 21:06, Rich Shepard wrote: > On Fri, 24 Sep 2010, John Reed wrote: > >> I compare an application every few days which has sqlite as it's client >> database. I look at the content and check whether documents have made it >> into the application after it has been built. I also check the metadata in >> the sqlite client database for changes. So, I am constantly comparing the >> last database with the newer database. Both databases have exactly the >> same tables, with only the data being changed in most of the 51 tables.The >> largest table has about 3,700,000 rows. Most other tables have much less >> rows in them. Could someone suggest an sql query to find the difference in >> the same table (ta) for both the last database (db1) and the newer >> database (db2)? I can use SQLiteSpy to connect and attach to the >> databases. > You'll want to tune the syntax, but try something like: > > SELECT colA FROM db1 WHERE (NOT EXIST colA IN db2); > > The idea is to match rows in each table and where the equivalent row in db1 > is not in db2, add that to the results table. > > Rich you mean something like: select id1, id2 from callprog a where id1 not in (select b.id1 FROM callprog b where b.id1=a.id1 );
But what is there is more than 1 column? it will grow in complexity when you have a lot of columns..... _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users