Docs Menu

Docs HomeDevelop ApplicationsMongoDB DriversC#/.NET

Operations with Builders

On this page

  • Overview
  • Sample Class
  • Construct a Filter
  • Array Operators
  • Create a Projection
  • Lambda Expressions
  • Define a Sort
  • Define an Update
  • Define Index Keys
  • Build an Aggregation Pipeline
  • Build an Atlas Search Query
  • Additional Information
  • API Documentation

In this guide, you can learn about the helper classes, or builders, that the .NET/C# Driver provides to create types used in your operations. Using builders helps you identify errors at compile time and avoid them at runtime. This guide provides information on builder classes that you can use for the following tasks:

  • Creating a filter definition

  • Creating a projection

  • Defining a sort order

  • Defining an update operation

  • Selecting index keys

Tip

MongoDB Analyzer

The MongoDB Analyzer is a tool that helps you analyze your builders definitions and understand how your .NET/C# code translates into the MongoDB Query API. For more information and installation instructions, see the MongoDB Analyzer reference page.

You should read this guide if you want to learn more about how to construct definitions and build up syntax using builders.

The code examples in this guide demonstrate how you can use builders to create types to interact with documents in the sample collection plants.flowers. Documents in this collection are modeled by the following Flower class:

public class Flower
{
public ObjectId Id { get; set; }
public string Name { get; set; }
public string Category { get; set; }
public double Price { get; set; }
public List<string> Season { get; set; }
public double Stock { get; set; }
public string Description { get; set; }
}

Each builder class takes a generic type parameter TDocument which represents the type of document that you are working with. In this guide, the Flower class is the document type used in each builder class example.

The FilterDefinitionBuilder class provides a type-safe interface for building up queries. Suppose you want to query your collection for documents matching the following criteria:

  • Price field value less than 20

  • Category field value is "Perennial"

Use builders to create the filter definition with the typed variant:

var builder = Builders<Flower>.Filter;
var filter = builder.Lt(f => f.Price, 20) & builder.Eq(f => f.Category, "Perennial");

Using the typed variant form provides compile-time safety. Additionally, your IDE can provide refactoring support.

Alternatively, you can use string-based field names to contruct the filter:

var builder = Builders<Flower>.Filter;
var filter = builder.Lt("Price", 20) & builder.Eq("Category", "Perennial");

If your document has properties or fields that serialize to arrays, you can use the methods beginning with Any, such as AnyEq() or AnyLt(), to compare the entire array against a single item.

Use builders to check which documents in the collection have a Season array that includes "winter":

var builder = Builders<Flower>.Filter;
var filter = builder.AnyEq(f => f.Season, "winter");

The ProjectionDefinitionBuilder class provides a type-safe interface for defining a projection. Suppose you want to create a projection on the Name and Price fields, but exclude the Id field.

Use builders to create the projection definition with the typed variant:

var builder = Builders<Flower>.Projection;
var projection = builder.Include(f => f.Name).Include(f => f.Price).Exclude(f => f.Id);

You can also use string-based field names to define the projection:

var builder = Builders<Flower>.Projection;
var projection = builder.Include("Name").Include("Price").Exclude("Id");

Finally, you can use the Expression() method to define the projection:

var builder = Builders<Flower>.Projection;
var projection = builder.Expression(f => new { Name = f.Name, Price = f.Price });

This definition has a return type of ProjectionDefinition<TDocument, TProjection> whereas the others return a ProjectionDefinition<TDocument>.

The driver supports using lambda expressions to render projections. When you define a Find() projection with the Expression() method to create a lambda expression, the driver inspects the expression to determine which fields are referenced and automatically constructs a server-side projection to return only those fields.

You can also use lambda expressions to create new fields by performing operations on values in your documents. The following example shows how you can use a lambda expression to project a new Profit field using the Price and Stock fields:

var builder = Builders<Flower>.Projection;
var projection = builder.Expression(f => new { Profit = f.Price * f.Stock });

Note

Id Field Exclusion

When you create a projection using a lambda expression, the output automatically excludes the Id field unless you explicitly include is as a projection field.

The SortDefinitionBuilder class provides a type-safe interface for building up sort syntax. Suppose you want to define a sort with the following order:

  • Ascending on Price

  • Descending on Category

Use builders to create the sort definition with the typed variant:

var builder = Builders<Flower>.Sort;
var sort = builder.Ascending(f => f.Price).Descending(f => f.Category);

Alternatively, you can use string-based field names to define the sort:

var builder = Builders<Flower>.Sort;
var sort = builder.Ascending("Price").Descending("Category");

The UpdateDefinitionBuilder class provides a type-safe interface for building up an update specification. Suppose you want to create an update specification with the following criteria:

  • Create the new field SunRequirement

  • Multiply the Price field value by 0.9

Use builders to create the update specification with the typed variant:

var builder = Builders<Flower>.Update;
var update = builder.Set(f => f.SunRequirement, "Full sun").Mul(f => f.Price, 0.9);

Alternatively, you can use string-based field names to define the update:

var builder = Builders<Flower>.Update;
var update = builder.Set("SunRequirement", "Full sun").Mul("Price", 0.9);

The IndexKeysDefinitionBuilder class provides a type-safe interface for defining index keys. Suppose you want to select Category as an ascending index key.

Use builders to select the index key with the typed variant:

var builder = Builders<Flower>.IndexKeys;
var keys = builder.Ascending(f => f.Category);

Alternatively, you can use string-based field names to select the index key:

var builder = Builders<BsonDocument>.IndexKeys;
var keys = builder.Ascending("Category");

The IndexKeysDefinitionBuilder class also provides methods to build a wildcard index. You can create a wildcard index using All field paths or A single field path, in this case using Category:

The PipelineDefinitionBuilder class provides a type-safe interface for defining an aggregation pipeline. An aggregation pipeline is a series of stages that are used to transform a document. Suppose you want to create a pipeline that performs the following operations:

  • Matches all documents with "spring" in the Season field

  • Sorts the results by the Category field

  • Groups the documents by category and shows the average price and total available for all documents in that category

Use PipelineDefinitionBuilder classes to build the pipeline:

var sortBuilder = Builders<Flower>.Sort.Ascending(f => f.Category);
var matchFilter = Builders<Flower>.Filter.AnyEq(f => f.Season, "spring");
var pipeline = new EmptyPipelineDefinition<Flower>()
.Match(matchFilter)
.Sort(sortBuilder)
.Group(f => f.Category,
g => new
{
name = g.Key,
avgPrice = g.Average(f => f.Price),
totalAvailable = g.Sum(f => f.Stock)
}
);

The preceding example creates the following pipeline:

[{ "$match" : { "season" : "spring" } }, { "$sort" : { "category" : 1 } }, { "$group" : { "_id" : "$category", "avgPrice" : { "$avg" : "$price" }, "totalAvailable" : { "$sum" : "$stock" } } }]

You can add stages to your pipeline that don't have corresponding type-safe methods in the PipelineDefinitionBuilder interface by providing your query as a BsonDocument to the AppendStage() method.

var pipeline = new EmptyPipelineDefinition<BsonDocument>().AppendStage<BsonDocument, BsonDocument, BsonDocument>("{ $set: { field1: '$field2' } }");

Note

When using a BsonDocument to define your pipeline stage, the driver does not take into account any BsonClassMap, serialization attributes or serialization conventions. The field names used in the BsonDocument must match those stored on the server.

For more information on providing a query as a BsonDocument, see our FAQ page.

To learn more about the Aggregation Pipeline, see the Aggregation Pipeline server manual page.

The Search class provides a type-safe interface for creating a $search pipeline stage.

To learn how to construct search queries with the Search class, see Atlas Search.

Find runnable examples using builders for various operations under Usage Examples.

To learn more about any of the methods or types discussed in this guide, see the following API Documentation:

← Count Documents