[Django]-Django – save() update on duplicate key

41👍

To update an existing rating, you actually have to have the one you want to update. If you know the object may not exist, use get_or_create:

rate, created = VideoRate.objects.get_or_create(user_id=1, video_id=1, crit_id=1)
rate.rate = 2
rate.save()

You can short-cut the process by using update():

VideoRate.objects.filter(user_id=1, video_id=1, crit_id=1).update(rate=2)

But this will silently fail if the rating does not exist – it won’t create one.

10👍

First, you must check if the rating exists. So you may either use what Daniel Roseman said or use exists, but you can not solve this with a simple update since update do not create new records…

rating = 2
rate, created = VideoRate.objects.get_or_create(user_id=1, video_id=1, crit_id=1,
    defaults={'rate':rating})#if create, also save the rate infdormation

if not created:# update
    rate.rate = rating
    rate.save()

You can use defaults to pass exrta arguments, so if it is an insert, database record will be created with all required information and you do not need to update it again…

Documentation

Update: This answer is quite old just like the question. As @peterthomassen mention, Django now have update_or_create() method

👤Mp0int

3👍

Django 4.1 has added the support for such conflicts. In the backend, it runs a INSERT...ON DUPLICATE KEY UPDATE query.

Example of above in a single query:

rate = VideoRate(user_id=1, video_id=1, crit_id=1, rate=2)
rate.save()

# updating rate object if it exists, else create one
rate = VideoRate(user_id=1, video_id=1, crit_id=1, rate=3)
VideoRate.objects.bulk_create(
    [rate],
    update_conflicts=True,
    update_fields=['rate']
)

Leave a comment