On Aug 14, 11:07 am, Michael Bayer <mike...@zzzcomputing.com> wrote: > In addition, isolation_level=None with pysqlite disables the DBAPIs > entire transactional system. SQLAlchemy relies upon this system to > handle proper transactional behavior. > > So I consider this a pysqlite bug, and they should offer a mode > by which there is no implicit commit for CREATE TABLE.
Michael, I don't doubt that you're right, but the puzzling thing (for me and, I think, Torsen) is that (if you set echo='debug') you see a BEGIN and ROLLBACK statement apparently being issued through the DBAPI layer, and if there's an implicit COMMIT going on we don't actually see it and don't understand why isolation_level=None isn't preventing it as it apparently does in the other case (test02). I suspect this is because with the logging on, we aren't actually seeing the operations performed, but merely sqlalchemy's report of what it is about to ask for. If the DBAPI is doing something under the covers, we don't know what and therefore can't find a workaround. To be clear, in test02 I believe we're effectively telling sqlite3 "BEGIN; CREATE TABLE ...; ROLLBACK", and in test03 (with the same isolation_level setting) that's exactly what the logging reports is happening, yet the behaviour is different. (I was hoping to put some debugging in the DBAPI layer but unfortunately that's a C extension so harder to deal with. Maybe there's also some feature in sqlite3 itself (not the Python module) which can be configured for debug purposes to show what's really happening here.) -- Peter Hansen Engenuity Corporation -- You received this message because you are subscribed to the Google Groups "sqlalchemy" group. To post to this group, send email to sqlalch...@googlegroups.com. To unsubscribe from this group, send email to sqlalchemy+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/sqlalchemy?hl=en.
