Most efficient way to fetch and output Content with 2-Level Comments?

I have some content with up to 2-levels of replies. I am wondering what the most efficient way to fetch and output the replies. I should note that I am planning on开发者_开发知识库 storing the comments with fields content_id and reply_to, where reply_to refers to which comment it is in reply to (if any). Any criticism on this design is welcome.

In pseudo-code (ish), my first attempt would be:

# in outputting content CONTENT_ID
all_comments = fetch all comments where content_id == CONTENT_ID
root_comments =  filter all_comments with reply_to == None
children_comments = filter all_comments with reply_to != None

output_comments = list()
for each root_comment
    children = filter children_comments, reply_to == root_comment.id
    output_coments.append( (root_comment, children) )

send output_comments to template

Is this the best way to do this? Thanks in advance.

Edit: On second thought, I'll want to preserve date-order on the comments, so I'll have to do this a bit differently, or at least just sort the comments afterward.

If you're going to display all the comments anyway, just fetch them all in one operation, then build the tree in Python.