Having it happen at the lookup creation time ensures entry points called before the compilation phase (e.g. get_group_by_cols) don't have to duplicate the logic in charge of altering Query instances used as rhs. It also has the nice effect of reducing the amount of time the alteration logic to once as opposed to multiple times if the queryset is compiled more than once. |
||
---|---|---|
.. | ||
backends | ||
migrations | ||
models | ||
__init__.py | ||
transaction.py | ||
utils.py |