使用Python在Google Appengine上创建线程注释的模型

Question

I am trying to write a comments model which can be threaded (no limits to the number of child threads). How do I do this in Appengine? and what is the fastest way to read all comments?

I am trying to do this in a scalable way so that app engine's new pricing does not kill my startup :)

Answer 1

It depends on how nested threads you expect to get and if you want to optimize reading or writing. If we guess that the threads normally are quite shallow and that you want to optimize reading all subcomments on all levels of a comment, I think you should store each comment in a separate entry and then put a reference to it in a list on the parent and the parent's parent and so forth.

That way getting all subcomments are always just one call, but writing a new comment is a bit slower, since you have to modify all parents.

Answer 2

If you need fast way to read all comments, keep in comment model not only a reference to the parent comment but add a reference to the whole topic also. Then you will be able to get all comments for the topic with Comment.query(Comment.topic=12334).fetch() .

If you are worrying about datastore usage cost use http://code.google.com/p/appengine-ndb-experiment/ . It allows to use reduce saved data amount and number of queries. Shortly it will replace current db module.