Django is to making a query much more complicated than it needs to be.
A Sentiment may have a User and a Card, and I am getting the Cards which are not in the passed User's Sentiments
This is the query:
Card.objects.all().exclude(sentiments__in=user.sentiments.all())
this is what Django runs:
SELECT * FROM "cards_card"
WHERE NOT ("cards_card"."id" IN (
SELECT V1."card_id" AS "card_id"
FROM "sentiments_sentiment" V1
WHERE V1."id" IN (
SELECT U0."id"
FROM "sentiments_sentiment" U0
WHERE U0."user_id" = 1
)
)
)
This is a version I came up with which didn't do an N-Times full table scan:
Card.objects.raw('
SELECT DISTINCT "id"
FROM "cards_card"
WHERE NOT "id" IN (
SELECT "card_id"
FROM "sentiments_sentiment"
WHERE "user_id" = ' + user_id + '
)
)')
I don't know why Django has to do it with the N-Times scan. I've been scouring the web for answers, but nothing so far. Any suggestions on how to keep the performance but not have to fall back to raw SQL?
A better way of writing this query without the subqueries would be:
Card.objects.all().exclude(sentiments__user__id=user.id)