MongoDB allows you to define rich data types, multi-dimensional data structures that are very similar to object-oriented programming language objects. That means that we can define complex objects like arrays with subdocuments embedded or even fields with just subdocuments. Which in turn, can have complex data structures with them. In this particular lesson, we are going to be looking to how to querying on a field of our movies collection, called tomatoes, which corresponds to Rotten Tomatoes' viewer reviews. And how to use the magic of the dot notation to explore and express really complex queries within MongoDB. All right, so we have these documents with subdocuments in them, but what's so special and what's so great about them? Most relational systems can also define columns of data type XML or even BLOB, where we can just throw into it, very complex data structures. The reasoning behind MongoDB is the ability to actually query within those data structures without just having a dump of that data in a particular column in a relational world. Therefore, you can model your data in such a way that it doesn't restrict the queryability of your application. Now let's see how to use this in action, using our mflix application. In this Jupyter notebook, that you have accessed through the handouts of this lesson, you will have a way of expressing those same queries against our mflix database. Now, first step would be to just import the necessary modules. In this case, I'm going to use pymongo and pprint. Second step would be to express our connection string and the way to connect using pymongo.MongoClient. Now, don't forget to get this particular URI. You can just go to your Atlas cluster, go to Connect. Say that you want to connect from your application, and basically copy the URI in place and replace the placeholder for your password by the password of our user that in our case is analytics-password. I've done that due diligence myself, and I can just pure and simply connect. No errors, everything seems to be good. Now to get started in analyzing exactly how the substructures of our documents work, let's have a look to the Titanic movie. Now I know there's some duplicates here, and we'll take care of this later. So therefore, I'm going to just find one where the title is Titanic. As we can see here, we have very, very complex data structures and lots of embed fields in the arrays and all sorts of different Data types here that we might want to treat in a particular fashion. Or even just query on those fields as we would query our normal, let's say, single field root level fields. But what's interesting about this is that we can isolate it a little bit in python. So let's look into tomatoes. Now again tomatoes is basically a field that condenses the information of rotten tomatoes for this particular movie and as we can see here we have a very very rich internal data structure. So tomatoes itself is an field, it has other fields in it which we can use to express queries. Those in that field themselves. Some of them are from to daytimes, to strings and so forth. So can use this data structure to do some very interesting queries. Going a little bit deeper, I can see that tomatoes.viewer has a couple of other embedded fields themselves, meter, numReviews, and rating. Now if I would like to get all the movies that have the same rating as this particular Titanic movie, I can just use the Titanic tomatoes.viewer.rating field, put its value into variable that I'm going to call rating, which in this case is 3.6. And use that value here to express a different query. Now, I'm going to look into all the different movies that have the field tomatoes viewer rating. Again, a subfield inside of another subfield with the same value from previous movie. And I'm going to print only the title, so it doesn't mess up with my view screen here. Let's pause just for a second and look into this dot notation. What is this anyway? The moment we use the dot notation to access elements of a complex object, arrays or sub documents or embedded documents. They are embedded within the documents of a collection. For example here, movies, we have that for accessing a field within tomatoes. And within viewer we would need to use this dot connecting those different fields. If we look back into the structure of this particular document, we see that we have several different data types or fields in this case that could use the same dot notation for us to access the embed information within it. Tomatoes is one of the most complex ones. You could have writers, which has an array and we can access the elements or indexes of those arrays to express queries as well. You always using the dot notation. So every time we want or need to access fields, within sub documents or within arrays, we will be using the dot notation. Now the dot notation is not only used for the find query or for the filter that we express in our queries. Can also be used for our sorts, meaning that, we can sort on any embed field that we wish. In this particular case, we are going to be using tomatoes.lastupdated. We are going to be sorting the results by the ASCENDING last update. Once we do that, we can see that we have our title, rampage that has been last updated on 2012. Then the following is The Suspect, in 2014, and so forth. These capability shows us that no matter how deep our field is, if we need to express a query or even a sort condition for our Queriability of our application, Mongo DB allows us to do that using the dot notation explicitly to get the type of queryability that our application demands, and this is how you can express queries using embedded fields.
