Finding documents in MongoDB using C#

2020, Feb 28    

MongoDB is a very flexible database and is renowned for its easy of use and scalability. It gets rid of all the nasty database tasks that you’d normally end up doing with a more traditional database and just allows you to get on with developing your application, thus making creating your application cheaper and iterations faster.

A lot of people are using MongoDB with Node.js, however there are loads of support for other languages and platforms.

One of these platforms is .NET, and MongoDB still has the same benefits too.

One of the main problems with starting off with MongoDB and C# is most of the examples are based around the Node.js driver or using the MongoDB shell. So we’re going to look at how we map queries written in the MongoDB shell to C#.

Finding a document

When you want to query a document in MongoDB, the driver actually sends a document which represents the query, this document looks something like the below:

{ _id: ObjectId("507f1f77bcf86cd799439011")}

The above query will return documents where the _id property of the object exactly matches the object id of ObjectId("507f1f77bcf86cd799439011").

MongoDB has a large range of query selectors that can be combined to target documents. These are operators such as $eq, $ne, $lte and $gte, we can also use logical operators like $and and $or.

So say we are looking for all documents outside a date ranges, it would look like this:

{
    "$or": [
        {"date" : { "$lt": ISODate("2021-01-01") } },
        {"date" : { "$gte": ISODate("2020-01-01") } }
    ]
}

Finding a document using C#

We’ve not really talked about how this is achieved in C#, but like most things in software development there is lots of ways to achieve the same outcome but they all have their own caveats. So we’ll go through a few examples.

Before we even start finding documents we need to create a MongoClient which we will then use to fetch the database and collection instances.

var client = new MongoClient();

var database = client.GetDatabase("test");

var events = database.GetCollection<BsonDocument>("events");

As you might have noticed we have passed in BsonDocument as a generic argument to the GetCollection method, alternatively we could create our own typed class to represent our event data in the document.

public class Event
{
    public ObjectId Id { get; set; }

    public string Name { get; set; }

    public DateTime At { get; set; }
}

var events = database.GetCollection<Event>("events");

This means we can model our documents as typed objects in C# and they’ll automatically get serialized and deserialized when required.

Basic

The most basic way to find a document using the C# driver is just to pass a string json object in to the Find method on the collection, this means we can take our above matching and pass it as a string.

var @event = await _collection.Find($"{{ _id: ObjectId('507f1f77bcf86cd799439011') }}")
    .SingleAsync();

In theory we can parameterize the string by using string concatenation, which will allow us to pass in any object Id.

var id = new ObjectId("507f1f77bcf86cd799439011");

var @event = await _collection.Find($"{{ _id: ObjectId('{id}') }}")
    .SingleAsync();

Be careful though, if you are accepting arbitrary input as string from the user they might be able to execute something that you wasn’t expecting.

Expressions

Another way to find the document is to pass in an expression, if you’re familiar with .Where in LINQ it’s very similar.

var id = new ObjectId("507f1f77bcf86cd799439011");

var @event = await _collection.Find(x => x.Id == id)
        .SingleAsync();

This stops the problem before with our arbitrary input, but it’s harder to compose parts of queries together.

Builders

Within the MongoDB C# driver we have a Builders<T> object that allows us to build up filter expressions.

var filter = Builders<Event>.Filter.Eq(x => x.Id, id);

var @event = await _collection.Find(filter)
        .SingleAsync();

Using Builders<T>.Filter has the benefit that it allows us to compose filters together.

var filter = Builders<Event>.Filter.Gt(x => x.At, date);
if (filterName)
{
    filter = filter & Builders<Event>.Filter.Eq(x => x.Name, name);
}

var @event = await _collection.Find(filter)
    .SingleAsync();

Our above example allows us to compose filters together, even with conditions of what filters we want to apply and when.

Language Integrated Query (LINQ)

The MongoDB driver supports LINQ, this is very beneficial for C# developers wanting to utilize their current skills from using libraries such as Entity Framework.

Using LINQ is achieved by calling AsQueryable method on the IMongoCollection<T> object, then you can chain the normal LINQ methods together before executing them.

var @event = _collection.Find(filter)
                .AsQueryable()
                .Where(x => x.Id == id)
                .Single();

Even though using LINQ makes the entry level for C# developer to MongoDB easier there are a few things to note.

  • Currently no async methods for the the LINQ syntax which means these are IO blocking call.
  • You might end up with runtime exceptions as not all LINQ expressions map directly to MongoDB queries.

BsonDocument

The last approach is to use a BsonDocument, this is an object that represents the dynamic data of the Bson document but in a typed way, because of this it allows us to do anything we really want.

var filter = new BsonDocument { { "_id", id } };

var @event = await _collection.Find(filter)
    .SingleAsync();

BsonDocument is useful when the Builders<T> object has not been updated with the latest version of the query operators, compared to what is available on the database engine.

Overall

As you can see there are many way to find documents using the C# MongoDB Driver, however the most type safe and flexible way is to use the Builders<T> object.

Twitter Facebook Google+
# MongoDB # .NET # C#