Object Mapping

LiteDB supports POCO class as strong types documents. When you get LiteCollection instance from LiteDatabase.GetCollection<T> you can set <T> type. This type will be your document type. If this <T> was diferent than BsonDocument, LiteDB internal maps your class to BsonDocument. To do this, LiteDB uses BsonMapper class:

// Simple strong type document
public class Customer
{
    public ObjectId CustomerId { get; set; }
    public string Name { get; set; }
    public DateTime CreateDate { get; set; }
    public List<Phone> Phones { get; set; }
    public bool IsActive { get; set; }
}

var typedCustomerCollection = db.GetCollection<Customer>("customer");

var schemelessCollection = db.GetCollection("customer"); // <T> is BsonDocument

BsonMapper convert each property class in a document field folling this rules:

• Classes must be pulic with a public one non-paramter constructor
• Properties must be public
• Properties can be read-only or read/write
• Class must have an Id property, <ClassName>Id property or any property with [BsonId] attribute
• A property can be decorated with [BsonIgnore] to not be mapped to document
• A property can be decorared with [BsonField] to have a custom name field
• No circular references allowed
• Max of 20 deep inner classes
• Class fields are not converted to document
• BsonMapper use a global instance that cache mapping information for a better performance.

Basic .NET data type mapper:

.NET Data Type	BSON Data Type
Int16, UInt16, Byte	Int32
UInt32 , UInt64	Int64
Single, Decimal	Double
Char, Enum	String
IList<T>	Array
IDictionary<K,T>	Document

• Nullable<T> values are accepted. If value is null, BSON data type is Null, otherwise, BSON value will be .NET then underlying type.
• For IDictionary<K, T>, K key must be String or simple data type (convertable using Convert.ToString(..)).

BsonMapper support some options

Name	Default	Description
SerializeNullValues	false	If class property is null do not map in document result
TrimWhitespace	true	Trim string properties before map to document
EmptyStringToNull	true	Empty string conver to null
ResolvePropertyName	(s) => s	A function to map property name to document field name

BsonMapper offers 2 predefined functions to resolve property name: UseCamelCase() and UseLowerCaseDelimiter("_").

BsonMapper.Global.UseLowerCaseDelimiter("_");

public class Customer
{
    public string FirstName { get; set; }
}

var doc = BsonMapper.Global.ToDocument(new Customer { FirstName = "John" });

var john = doc["first_name"];

You can register your own map function, using RegisterType<T> instance method. To register, you need create serialize/deserialize functions.

BsonMapper.Global.RegisterType<Uri>
(
    serialize: (uri) => uri.AbsoluteUri,
    deserialize: (bson) => new Uri(bson.AsString)
);

• Serialize functions pass a <T> object instance as input paramter and expect return a BsonValue
• Deserialize function pass a BsonValue object as input paramter and except return a <T> value
• RegisterType supports complex objects by using BsonDocument or BsonArray

AutoId

By default, typed documents will recive an auto Id value on insert. LiteDB support auto-id for this data types:

.NET data type	New Value
ObjectId	ObjectId.NewObjectId()
Guid	Guid.NewGuid()
Int32	Sequence value starting in 1

But if you don't use or want another data type for Id, you can set previously:

var customer = new Customer { Id = "john-doe", Name = "John Doe" };

You also can create your own auto-id function, using RegisterAutoId<T>:

BsonMapper.Global.RegisterAutoId<long>
(
    isEmpty: (value) => value == 0,
    newId: (collection) => DateTime.UtcNow.Ticks
);

public class Calendar
{
    public long CalendarId { get; set; }
}

• isEmpty returns true to indicate when this type is considerate empty. In this example, zero will be empty value.
• newId returns new id value. Gets LiteCollection<BsonDocument> instance as input parameter if you need use.

Index definition

BsonMapper support index definition direct on property using [BsonIndex] attribute. You can define index options like, Unique or IgnoreCase. This is useful to avoid always run col.EnsureIndex("field") before run a query.

public class Customer
{
    public int Id { get; set; }
    
    [BsonIndex]
    public string Name { get; set; }
    
    [BsonIndex(new IndexOptions { Unique = true, IgnoreCase = false, RemoveAccents = false })]
    public string Email { get; set; }
}

IndexOptions class has this settings:

Name	Default	Description
Unique	false	Do not allow duplicates values in index
IgnoreCase	true	Store string on index in lowercase
RemoveAccents	true	Store string on index removing accents
TrimWhitespace	true	Store string on index removing trimming spaces
EmptyStringToNull	true	If string is empty, convert to Null

To get better performance, [BsonIndex] checks only if index exists but do not check if you are changing options. To change an index option in a existing index you must run EnsureIndex with new index options. This method drop current index and create a new one with new options.