Creating A Complex Tree in MongoDB And Querying For The Data

Question

I'm trying to do something like this in MongoDB:

wiki_db
   |
   +-- Programming --+
                     |
                     |
                     +-- Perl -+
                               |
                               +-- tutorials --+
                                               |
                                               +-- 1 --+
                                                       |
                                                       |      +---- id      = 1
                                                       |      |
                                                       +------+---- title   = Introdaction To Perl
                                                              |
                                                              +---- Date    = 12/11/2100
                                                              |
                                                              +---- Content = "Perl is .... "

It's a small database for a wiki application I'm planning to write it in perl, which :

• wiki_db : is the database.
• programming: is the main section.
• perl: is the sub section.
• tutorials: is the thread.
• 1: is the page number.

Like a tree. So, I've written this code to describe it :

programming = {
    'perl': {
        'tutorials': {
            '1': {
                'id':'1',
                'title':'Hello World!',
                'lastmod':'Sat Jul 23 14:56:22 AST 2011',
                'Content': 'Hello!, This is the first page in the tutorial.'
            }
        }
    }
}

The problem is that when I do some query, it returns nothing:

$ mongo
MongoDB shell version: 1.8.2
connecting to: test
> use wiki
switched to db wiki
> show collections
programming
system.indexes
> db.programming.findOne()
{
    "_id" : ObjectId("4e2f2fadce7012941395b103"),
    "perl" : {
        "tutorials" : {
            "1" : {
                "id" : "1",
                "title" : "Hello World!",
                "lastmod" : "Sat Jul 23 14:56:22 AST 2011",
                "Content" : "Hello!, This is the first page in the tutorial."
            }
        }
    }
}
> db.programming.find({"perl":{"tutorials":{"1":{"id":"1"}}}})
> 

What is the correct way to do write the query? and is the tree design good? I mean will that design slow down the database when the database grow up?

Answer 1

Tutorials could be an array instead:

programming = {
    'perl': {
        'tutorials': [
            {
                'id':'1',
                'title':'Hello World!',
                'lastmod':'Sat Jul 23 14:56:22 AST 2011',
                'Content': 'Hello!, This is the first page in the tutorial.'
            }
        ]
    }
}

You could then do the find like this:

db.programming.find({'perl.tutorials.id': '1'})

However, depending on how many tutorials you have per topic, and how long they are, this model may very well hit the 16MB per document limit. I would suggest splitting tutorials out into their own collection, and use tags to fine them like this:

db.tutorials.save({
                "title" : "Hello World!",
                "lastmod" : "Sat Jul 23 14:56:22 AST 2011",
                "Content" : "Hello!, This is the first page in the tutorial.",
                "Tags":['perl']
            });

Then you could find like this:

db.tutorials.find({Tags:'perl'});

This would prevent hitting the 16MB doc limit and make your model a bit more flexible. See Trees in MongoDB

Answer 2

Do this:

> db.programming.find({'perl.tutorials.1.id': '1'})
{ "_id" : ObjectId("4e2f45ef55bf6c17a7f511e0"), "perl" : { "tutorials" : { "1" : { "id" : "1", "title" : "Hello World" } } } }