Limit number of objects of a self reference model with Prefetch Django - django

Question

I have 2 models: Post and Comment. Like this:
class Post(models.Model):
content = models.TextField()
class Comment(models.Model)
content = models.TextField()
owner = models.ForeignKey(User, null=False, on_delete=models.CASCADE)
post_parent = models.ForeignKey(
Post,
on_delete=models.CASCADE,
null=False,
db_column='post_parent',
related_name='comments',
)
comment_ref = models.ForeignKey(
'self',
on_delete=models.CASCADE,
null=True,
related_name='children_comment'
)
Suppose that comments can have sub-comments in 1 level. With a post's ID, I want to get 5 newest comments (not sub-comment), each of them is prefetched with their 3 newest sub-comments.
My current implementation after searching is:
sub_comments = Comment.objects.filter(
comment_ref=OuterRef('pk'),
post_parent=post_id
).order_by('-timestamp').values_list('id', flat=True)[:3]
queryset = Comment.objects.filter(
comment_ref=None,post_parent=post_id
).select_related('owner').prefetch_related(
Prefetch('children_comment', queryset=Comment.objects.filter(
id__in=Subquery(sub_comments)
), to_attr="cmt"))
That does not work. It doesn't limit the number of prefetched sub-comments. Please help me :D

Answer 1

You will need a date field on the Comment table so you can show the latest comments and sub comments.
I would introduce another model here so its easier for queries and writing tests personally for me.
Models.py
class Post(models.Model):
content = models.TextField()
class Comment(models.Model)
content = models.TextField()
owner = models.ForeignKey(User, null=False, on_delete=models.CASCADE)
post_parent = models.ForeignKey(
Post,
on_delete=models.CASCADE,
null=False,
db_column='post_parent',
related_name='comments',
)
date_posted = models.DateTimeField()
class ChildComment(models.Model):
parent_comment = models.ForeignKey(
Comment,
on_delete=models.CASCADE,
null=True,
related_name="child_comments"
)
Views.py
post = Post.objects.get(id=<some_id>)
comments = Comment.objects.filter(
post_parent_id=post.id
).prefetch_related(
'child_comments'
).order_by(
'-date_posted'
)[:5]
recent_comments = {}
for comment in comments:
recent_comments[comment] = [
comment.child_comments.all().order_by(
'-parent_comment__date_posted'
)[:3]
]
Pass recent_comments to the template context and then in a template file use it like:
{% for recent_comment, child_comments in recent_comments.items %}
{{ recent_comment.content }}
{% for child_comment in child_comments %}
{{ child_comment.content }}
{% endfor %}
{% endfor %}

Answer 2

You will need a DateTimeField to get the latest comments and child comments in an order_by, but this query should work. This may not be ideal if you were returning thousands of comments, but only 5 at a time, this will be fast.
comments = Comment.objects.filter(
post_parent_id=post.id
)[:5]
root_comment_ids = [comment.id for comment in comments]
child_comments = Comment.objects.filter(
comment_ref_id__in=root_comment_ids
)[:3]