Ekspresi F untuk Pembaruan Atomik

Loading a value, changing it in Python, then saving creates a race: two requests can read the same number and overwrite each other.

An F expression refers to a column by name, letting the database do the math itself instead of round-tripping the value to Python.

Use F to add to a column based on its own current value, computed safely inside a single SQL statement.

The whole change runs as one UPDATE, so concurrent requests stack correctly instead of clobbering one another's results.

Because F runs in SQL, you can apply a change across a whole queryset in one query, like giving every product a raise.

F expressions also shine in filters, letting you compare one column against another directly in the query.

An in-memory instance does not change when you run update, so call refresh_from_db to pull the new value back.

You can also assign an F expression to a field and then save; Django turns it into the same column-aware SQL.

F values support full arithmetic, so you can combine columns, for example storing a total computed from price and quantity.

After saving with an F expression, the Python attribute still holds the expression object, not the number, until you refresh.

Pick F whenever an update depends on a column's current value, especially counters, balances, and stock that many requests touch.

Two checkouts hit the same product at once. Which approach avoids losing a decrement?

F expressions push math into the database for atomic, race-free updates across one or many rows. Just refresh your instance afterward. 💾