There were a couple of errors in ._dirty flag handling:
* It started as None, but was never reset to None.
* The _dirty flag was sometimes used to indicate if the connection
was inside transaction management, but this was not done
consistently. This also meant the flag had three separate values.
* The None value had a special meaning, causing for example inability
to commit() on new connection unless enter/leave tx management was
done.
* The _dirty was tracking "connection in transaction" state, but only
in managed transactions.
* Some tests never reset the transaction state of the used connection.
* And some additional less important changes.
This commit has some potential for regressions, but as the above list
shows, the current situation isn't perfect either.
Implemented this by adding an internal hook for work that should be performed
before that point.
Also, regarding the `DatabaseCreation.set_autocommit()` method:
* Stop using it for such tasks
* Stop providing an implementation that tries to cover all the possible
idioms a third party database backend DB-API 2 driver could need to activate
autocommit. It is now left for third party backends to implement.
This can be backwards incompatible in the case of user applications that:
* Had started using this method
* Use a third a party database backend
git-svn-id: http://code.djangoproject.com/svn/django/trunk@16520 bcc190cf-cafb-0310-a4f2-bffc1f526a37
Anssi Kääriäinen
Claude Paroz
Aymeric Augustin
Ramiro Morales
Jannis Leidel
Russell Keith-Magee
Adrian Holovaty