Mongoose/mongo many to many with computed field - mongodb

Question

This is my schema. Each user has many posts and each post has list of users who favorited it.
var User = mongoose.model('User', {
name: String,
posts: [{ type: mongoose.Schema.Types.ObjectId, ref: 'Post' }]
});
var Post = mongoose.model('Post', {
name: String,
favorited: [{ type: mongoose.Schema.Types.ObjectId, ref: 'User' }]
});
I need to select all user posts with a boolean flag is it favorited or not.
I've started with:
User.findById('56e14680476f47200f1f598e')
.populate('posts')
.then(u => {
console.log(u);
})
.catch(console.error.bind(console));
Output:
[ { favorited: [Object],
__v: 0,
name: 'post 1',
_id: 56e14680476f47200f1f5991 },
{ favorited: [],
__v: 0,
name: 'post 2',
_id: 56e14680476f47200f1f5992 },
{ favorited: [],
__v: 0,
name: 'post 3',
_id: 56e14680476f47200f1f5993 } ],
__v: 0,
name: 'user 1',
_id: 56e14680476f47200f1f598e }
Aggregation, map-reduce, or maybe re-design schema? Maybe there are good examples of complex structures? Where to dig?

Answer 1

Please try this one
User.findById('56e14680476f47200f1f598e')
.populate('posts')
.exec(function(err, user) {
if (err)
console.log(err);
else{
// populate `posts` again
Post.populate(user.posts, {path:'favorited', model: 'User'}, function(err, ret) {
if (err)
console.log(err);
else
console.log(require('util').inspect(ret, { showHidden: true, depth: null }));
})
}
});
Second thought, if there are litter fields in User collection, you could just nest user information into Post collection, rather than reference it.
var Post = mongoose.model('Post', {
name: String,
favorited: [{ name: String,
// other fields...
}]
});
One concern is that, if there are many favorited users could make the Post collection too large. Make sure no more the collection limit, 16 megabytes.
Whatever, the data schema should meet your data, and facilitate to query it.