Good people of sqlite-users,
I've been porting an application originally designed for SQLite to
PostgreSQL, and the latter failed on a test case where the former does
not. While I've encountered many more such failures during the process
than I'd like, in this case I think PostgreSQL's behaviour is more
desirable. Here's a minimal query:
select (select 'foo' union select 'bar') || 'bar';
SQLite 3.25.3 returns 'barbar' (regardless of the value of the
reverse_unordered_select pragma) while PostgreSQL 11 refuses to process
the query unless the subquery is reduced to a single row. In my
application the actual query was erroneous and would potentially return
data belonging to a user other than the one making the request, which
concerns me quite a bit. Is it possible to make SQLite fail like
PostgreSQL does?
--
J. King
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
