AutoMapper¶
A convention-based object-object mapper.
AutoMapper uses a fluent configuration API to define an object-object mapping strategy. AutoMapper uses a convention-based matching algorithm to match up source to destination values. AutoMapper is geared towards model projection scenarios to flatten complex object models to DTOs and other simple objects, whose design is better suited for serialization, communication, messaging, or simply an anti-corruption layer between the domain and application layer.
AutoMapper supports the following platforms:
- .NET 4.6.1+
- .NET Standard 2.0+ https://docs.microsoft.com/en-us/dotnet/standard/net-standard
New to AutoMapper? Check out the Getting Started Guide page first.
Getting Started Guide¶
What is AutoMapper?¶
AutoMapper is an object-object mapper. Object-object mapping works by transforming an input object of one type into an output object of a different type. What makes AutoMapper interesting is that it provides some interesting conventions to take the dirty work out of figuring out how to map type A to type B. As long as type B follows AutoMapper’s established convention, almost zero configuration is needed to map two types.
Why use AutoMapper?¶
Mapping code is boring. Testing mapping code is even more boring. AutoMapper provides simple configuration of types, as well as simple testing of mappings. The real question may be “why use object-object mapping?” Mapping can occur in many places in an application, but mostly in the boundaries between layers, such as between the UI/Domain layers, or Service/Domain layers. Concerns of one layer often conflict with concerns in another, so object-object mapping leads to segregated models, where concerns for each layer can affect only types in that layer.
How do I use AutoMapper?¶
First, you need both a source and destination type to work with. The destination type’s design can be influenced by the layer in which it lives, but AutoMapper works best as long as the names of the members match up to the source type’s members. If you have a source member called “FirstName”, this will automatically be mapped to a destination member with the name “FirstName”. AutoMapper also supports Flattening.
AutoMapper will ignore null reference exceptions when mapping your source to your target. This is by design. If you don’t like this approach, you can combine AutoMapper’s approach with custom value resolvers if needed.
Once you have your types you can create a map for the two types using a MapperConfiguration
or the static Mapper
instance and CreateMap. You only need one MapperConfiguration
instance typically per AppDomain and should be instantiated during startup. Alternatively, you can just use Mapper.Initialize
(more examples of initial setup see in Static-and-Instance-API.
Mapper.Initialize(cfg => cfg.CreateMap<Order, OrderDto>());
//or
var config = new MapperConfiguration(cfg => cfg.CreateMap<Order, OrderDto>());
The type on the left is the source type, and the type on the right is the destination type. To perform a mapping, use the static or instance Mapper methods, depending on static or instance initialization:
var mapper = config.CreateMapper();
// or
var mapper = new Mapper(config);
OrderDto dto = mapper.Map<OrderDto>(order);
// or
OrderDto dto = Mapper.Map<OrderDto>(order);
Most applications can use dependency injection to inject the created IMapper
instance.
AutoMapper also has non-generic versions of these methods, for those cases where you might not know the type at compile time.
Where do I configure AutoMapper?¶
If you’re using the static Mapper method, configuration should only happen once per AppDomain. That means the best place to put the configuration code is in application startup, such as the Global.asax file for ASP.NET applications. Typically, the configuration bootstrapper class is in its own class, and this bootstrapper class is called from the startup method. The bootstrapper class should call Mapper.Initialize to configure the type maps.
For ASP.NET Core the Dependency Injection article shows how to configure AutoMapper in your application.
How do I test my mappings?¶
To test your mappings, you need to create a test that does two things:
- Call your bootstrapper class to create all the mappings
- Call MapperConfiguration.AssertConfigurationIsValid
Here’s an example:
var config = AutoMapperConfiguration.Configure();
config.AssertConfigurationIsValid();
Can I see a demo?¶
Check out the dnrTV episode on AutoMapper.
5.0 Upgrade Guide¶
Initialization¶
You now must use either Mapper.Initialize
or new MapperConfiguration()
to initialize AutoMapper. If you prefer to keep the static usage, use Mapper.Initialize
.
If you have a lot of Mapper.CreateMap
calls everywhere, move those to a Profile, or into Mapper.Initialize
, called once at startup.
For examples see here.
Profiles¶
Instead of overriding a Configure method, you configure directly via the constructor:
public class MappingProfile : Profile {
public MappingProfile() {
CreateMap<Foo, Bar>();
RecognizePrefix("m_");
}
}
IgnoreAllNonExisting extension¶
A popular Stack Overflow post introduced the idea of ignoring all non-existing members on the destination type. It used things that don’t exist anymore in the configuration API. This functionality is really only intended for configuration validation.
In 5.0, you can use ReverseMap or CreateMap passing in the MemberList enum to validate against the source members (or no members). Any place you have this IgnoreAllNonExisting extension, use the CreateMap overload that validates against the source or no members:
cfg.CreateMap<ProductDto, Product>(MemberList.None);
Resolution Context things¶
ResolutionContext used to capture a lot of information, source and destination values, along with a hierarchical parent model. For source/destination values, all of the interfaces (value resolvers and type converters) along with config options now include the source/destination values, and if applicable, source/destination members.
If you’re trying to access some parent object in your model, you will need to add those relationships to your models and access them through those relationships, and not through AutoMapper’s hierarchy. The ResolutionContext was pared down for both performance and sanity reasons.
Value resolvers¶
The signature of a value resolver has changed to allow access to the source/destination models. Additionally, the base class is gone in favor of interfaces. For value resolvers that do not have a member redirection, the interface is now:
public interface IValueResolver<in TSource, in TDestination, TDestMember>
{
TDestMember Resolve(TSource source, TDestination destination, TDestMember destMember, ResolutionContext context);
}
You have access now to the source model, destination model, and destination member this resolver is configured against.
If you are using a ResolveUsing and passing in the FromMember
configuration, this is now a new resolver interface:
public interface IMemberValueResolver<in TSource, in TDestination, in TSourceMember, TDestMember>
{
TDestMember Resolve(TSource source, TDestination destination, TSourceMember sourceMember, TDestMember destMember, ResolutionContext context);
}
This is now configured directly as ForMember(dest => dest.Foo, opt => opt.ResolveUsing<MyCustomResolver, string>(src => src.Bar)
Type converters¶
The base class for a type converter is now gone in favor of a single interface that accepts the source and destination objects and returns the destination object:
public interface ITypeConverter<in TSource, TDestination>
{
TDestination Convert(TSource source, TDestination destination, ResolutionContext context);
}
Circular references¶
Previously, AutoMapper could handle circular references by keeping track of what was mapped, and on every mapping, check a local hashtable of source/destination objects to see if the item was already mapped. It turns out this tracking is very expensive, and you need to opt-in using PreserveReferences for circular maps to work. Alternatively, you can configure MaxDepth:
// Self-referential mapping
cfg.CreateMap<Category, CategoryDto>().MaxDepth(3);
// Circular references between users and groups
cfg.CreateMap<User, UserDto>().PreserveReferences();
Starting from 6.1.0 PreserveReferences is set automatically at config time whenever the recursion can be detected statically. If that doesn’t happen in your case, open an issue with a full repro and we’ll look into it.
UseDestinationValue¶
UseDestinationValue tells AutoMapper not to create a new object for some member, but to use the existing property of the destination object. It used to be true by default. Consider whether this applies to your case. Check recent issues.
cfg.CreateMap<Source, Destination>()
.ForMember(d => d.Child, opt => opt.UseDestinationValue());
8.0 Upgrade Guide¶
ProjectUsing¶
The ProjectUsing
method consolidated with ConvertUsing
:
// IMappingExpression
// Old
void ConvertUsing(Func<TSource, TDestination> mappingFunction);
void ProjectUsing(Expression<Func<TSource, TDestination>> mappingExpression);
// New
void ConvertUsing(Expression<Func<TSource, TDestination>> mappingExpression);
To migrate, replace all usages of ProjectUsing
with ConvertUsing
.
The ConvertUsing
expression-based method will be used for both in-memory mapping and LINQ projections. You cannot have separate configuration for in-memory vs. LINQ projections.
Existing ConvertUsing
usages¶
The change from Func
to Expression
may break some existing usages. Namely:
ConvertUsing
using lambda statements, method groups, or delegates- Dual configuration of
ProjectUsing
andConvertUsing
For the first case, you may either:
- Convert to a lambda expression
- Move to the
Func
-based overloads
The Func
-based overloads accept more parameters, so you may have to add the parameters to your delegates.
Motivation¶
Simplify overloads, and to make it clear that you cannot have separate configuration for LINQ projections vs. in-memory mapping.
ConstructProjectionUsing¶
The ConstructProjectionUsing
method consolidated with ConstructUsing
:
// IMappingExpression<TSource, TDestination>
// Old
IMappingExpression<TSource, TDestination> ConstructUsing(Func<TSource, TDestination> ctor);
IMappingExpression<TSource, TDestination> ConstructUsing(Func<TSource, ResolutionContext, TDestination> ctor);
IMappingExpression<TSource, TDestination> ConstructProjectionUsing(Expression<Func<TSource, TDestination>> ctorExpression);
// New
IMappingExpression<TSource, TDestination> ConstructUsing(Expression<Func<TSource, TDestination>> ctor);
IMappingExpression<TSource, TDestination> ConstructUsing(Func<TSource, ResolutionContext, TDestination> ctor);
// IMappingExpression
// Old
IMappingExpression ConstructUsing(Func<object, object> ctor);
IMappingExpression ConstructUsing(Func<object, ResolutionContext, object> ctor);
IMappingExpression ConstructProjectionUsing(LambdaExpression ctorExpression);
// New
IMappingExpression ConstructUsing(Expression<Func<object, object>> ctor);
IMappingExpression ConstructUsing(Func<object, ResolutionContext, object> ctor);
To migrate, replace all usages of ConstructProjectionUsing
with ConstructUsing
.
The ConstructUsing
expression-based method will be used for both in-memory mapping and LINQ projections. You cannot have separate configuration for in-memory vs. LINQ projections.
Existing ConstructUsing
usages¶
The change from Func
to Expression
may break some existing usages. Namely:
ConstructUsing
using lambda statements, method groups, or delegates- Dual configuration of
ConstructProjectionUsing
andConstructUsing
For the first case, you may either:
- Convert to a lambda expression
- Move to the
Func
-based overload
The Func
-based overload accepts more parameters, so you may have to add the parameters to your delegates.
Motivation¶
Simplify overloads, and to make it clear that you cannot have separate configuration for LINQ projections vs. in-memory mapping.
ResolveUsing¶
The ResolveUsing
method consolidated with MapFrom
:
// IMappingExpression
// Old
void ResolveUsing(Func<TSource, TDestination> mappingFunction);
void ResolveUsing(Func<TSource, TDestination, TDestination> mappingFunction);
void ResolveUsing<TResult>(Func<TSource, TDestination, TMember, TResult> mappingFunction);
// Many, many overloads
void MapFrom(Expression<Func<TSource, TDestination>> mapExpression);
// New
void MapFrom(Expression<Func<TSource, TDestination>> mappingExpression);
void MapFrom<TResult>(Func<TSource, TDestination, TResult> mappingFunction);
void MapFrom<TResult>(Func<TSource, TDestination, TMember, TResult> mappingFunction);
To migrate, replace all usages of ResolveUsing
with MapFrom
.
The MapFrom
expression-based method will be used for both in-memory mapping and LINQ projections. You cannot have separate configuration for in-memory vs. LINQ projections.
Existing ResolveUsing
usages¶
The change from Func
to Expression
may break some existing usages. Namely:
ResolveUsing
using lambda statements, method groups, or delegates- Dual configuration of
ResolveUsing
andMapFrom
For the first case, you may either:
- Convert to a lambda expression
- Move to the
Func
-based overloads
The Func
-based overloads accept more parameters, so you may have to add the parameters to your delegates.
Motivation¶
Simplify overloads, and to make it clear that you cannot have separate configuration for LINQ projections vs. in-memory mapping.
UseValue¶
Underneath the covers, UseValue
called MapFrom
. UseValue
consolidated with MapFrom
.
To migrate, replace all usages of UseValue
with MapFrom
:
// Old
cfg.CreateMap<Source, Dest>()
.ForMember(dest => dest.Date, opt => opt.UseValue(DateTime.Now));
// New
cfg.CreateMap<Source, Dest>()
.ForMember(dest => dest.Date, opt => opt.MapFrom(src => DateTime.Now));
This can be simplified to a global find and replace of UseValue(
with MapFrom(src =>
.
Motivation¶
To make the underlying configuration more explicit. Historically, MapFrom
only allowed mapping from an individual source member. This restriction went away with 5.0, so there is no longer a need for additional redundant configuration options originally meant to work around this restriction.
ForSourceMember Ignore¶
ISourceMemberConfigurationExpression.Ignore
was renamed to DoNotValidate
to avoid confusion. It only applies when validating source members, with MemberList.Source
. It does not affect the mapping itself or validation for the default case, MemberList.Destination
.
To migrate, replace all usages of Ignore
with DoNotValidate
:
// Old
cfg.CreateMap<Source, Dest>()
.ForSourceMember(source => source.Date, opt => opt.Ignore());
// New
cfg.CreateMap<Source, Dest>()
.ForSourceMember(source => source.Date, opt => opt.DoNotValidate());
Generic maps validation¶
Generic maps are now validated. The generic map itself is validated at configuration time for the non generic members, so AssertConfigurationIsValid should catch errors for those. And the closed generic map will be validated when it’s used, possibly at runtime. If you don’t care about those errors, you need to override them.
Static and Instance API¶
In 4.2.1 version of AutoMapper and later, AutoMapper provides two APIs: a static and an instance API. The static API:
Mapper.Initialize(cfg => {
cfg.AddProfile<AppProfile>();
cfg.CreateMap<Source, Dest>();
});
var dest = Mapper.Map<Source, Dest>(new Source());
And the instance API:
var config = new MapperConfiguration(cfg => {
cfg.AddProfile<AppProfile>();
cfg.CreateMap<Source, Dest>();
});
var mapper = config.CreateMapper();
// or
IMapper mapper = new Mapper(config);
var dest = mapper.Map<Source, Dest>(new Source());
Gathering configuration before initialization¶
AutoMapper also lets you gather configuration before initialization:
var cfg = new MapperConfigurationExpression();
cfg.CreateMap<Source, Dest>();
cfg.AddProfile<MyProfile>();
MyBootstrapper.InitAutoMapper(cfg);
Mapper.Initialize(cfg);
// or
var mapperConfig = new MapperConfiguration(cfg);
IMapper mapper = new Mapper(mapperConfig);
LINQ projections¶
For the instance API, you can use IMapper.ProjectTo. If you prefer to keep using the IQueryable extension methods, you have to pass in the MapperConfiguration instance:
public class ProductsController : Controller {
public ProductsController(MapperConfiguration config) {
this.config = config;
}
private MapperConfiguration config;
public ActionResult Index(int id) {
var dto = dbContext.Products
.Where(p => p.Id == id)
.ProjectTo<ProductDto>(config)
.SingleOrDefault();
return View(dto);
}
}
Unsupported operations¶
One “feature” of AutoMapper allowed you to modify configuration at runtime. That caused many problems, so the new static API does not allow you to do this. You’ll need to move all your Mapper.CreateMap
calls into a profile, and into a Mapper.Initialize
.
For dynamic mapping, such as Mapper.DynamicMap
, you can configure AutoMapper to create missing maps as needed:
Mapper.Initialize(cfg => cfg.CreateMissingTypeMaps = true);
Internally this uses conventions to create maps as necessary.
Conventions¶
Conditional Object Mapper¶
Conditional Object Mappers make new type maps based on conditional between the source and the destination type.
var config = new MapperConfiguration(cfg => {
cfg.AddConditionalObjectMapper().Where((s, d) => s.Name == d.Name + "Dto");
});
Member Configuration¶
Member Configuration is like Configuration but you can have complete control on what is and isn’t used.
var config = new MapperConfiguration(cfg => { cfg.AddMemberConfiguration(); });
AddMemberConfiguration() starts off with a blank slate. Everything that applies in Configuration by default is gone to start with.
Naming Conventions¶
AddMemberConfiguration().AddMember<NameSplitMember>()
gets you default naming convention functionality.
Can overwrite the source and destination member naming conventions by passing a lambda through the parameter. SourceExtentionMethods can also be set here.
If you don’t set anything AutoMapper will use DefaultMember, which will only check using the Name of the property.
PS: If you don’t set this Flattening of objects will be disabled.
Replacing characters¶
AddMemberConfiguration().AddName<ReplaceName>(_ => _.AddReplace("Ä", "A").AddReplace("í", "i"));
Recognizing pre/postfixes¶
AddMemberConfiguration()
.AddName<PrePostfixName>(_ => _.AddStrings(p => p.Prefixes, "Get", "get")
.AddStrings(p => p.DestinationPostfixes, "Set"));
Attribute Support¶
AddMemberConfiguration().AddName<SourceToDestinationNameMapperAttributesMember>();
* Currently is always on
Looks for instances of SourceToDestinationMapperAttribute for Properties/Fields and calls user defined isMatch function to find member matches.
MapToAttribute is one of them which will match the property based on name provided.
public class Foo
{
[MapTo("SourceOfBar")]
public int Bar { get; set; }
}
Getting AutoMapper Defaults¶
AddMemberConfiguration()
.AddMember<NameSplitMember>()
.AddName<PrePostfixName>(_ => _.AddStrings(p => p.Prefixes, "Get"))
Is the default values set by Configuration if you don’t use AddMemberConfiguration().
Expand-ability¶
Each of the AddName and AddMember types are based on an interface ISourceToDestinationNameMapper and IChildMemberConfiguration. You can make your own classes off the interface and configure their properties through the lambda statement argument, so you can have fine tune control over how AutoMapper resolves property maps.
Multiple Configurations¶
Each configuration is its own set of rules that all must pass in order to say a property is mapped. If you make multiple configurations they are completely separate from one another.
Profiles¶
These can be added to Profile as well as the ConfigurationStore.
Each Profiles rules are separate from one another, and won’t share any conditions. If a map is generated from AddConditionalObjectMapper of one profile, the AddMemberConfigurations of only that profile can be used to resolve the property maps.
Example¶
Shown below is two profiles for making conventions for transferring to and from a Data Transfer Object. Each one is isolated to one way of the mappings, and the rules are explicitly stated for each side.
// Flattens with NameSplitMember
// Only applies to types that have same name with destination ending with Dto
// Only applies Dto post fixes to the source properties
public class ToDTO : Profile
{
protected override void Configure()
{
AddMemberConfiguration().AddMember<NameSplitMember>().AddName<PrePostfixName>(
_ => _.AddStrings(p => p.Postfixes, "Dto"));
AddConditionalObjectMapper().Where((s, d) => s.Name == d.Name + "Dto");
}
}
// Doesn't Flatten Objects
// Only applies to types that have same name with source ending with Dto
// Only applies Dto post fixes to the destination properties
public class FromDTO : Profile
{
protected override void Configure()
{
AddMemberConfiguration().AddName<PrePostfixName>(
_ => _.AddStrings(p => p.DestinationPostfixes, "Dto"));
AddConditionalObjectMapper().Where((s, d) => d.Name == s.Name + "Dto");
}
}
Understanding Your Mappings¶
AutoMapper creates an execution plan for your mapping. That execution plan can be viewed as an expression tree during debugging. You can get a better view of the resulting code by installing the ReadableExpressions VS extension. If you need to see the code outside VS, you can use the ReadableExpressions package directly.
var configuration = new MapperConfiguration(cfg => cfg.CreateMap<Foo, Bar>());
var executionPlan = configuration.BuildExecutionPlan(typeof(Foo), typeof(Bar));
Be sure to remove all such code before release.
For ProjectTo, you need to inspect IQueryable.Expression.
var expression = context.Entities.ProjectTo<Dto>().Expression;
The MyGet Build¶
AutoMapper uses MyGet to publish development builds based on the master branch. This means that the MyGet build sometimes contains fixes that are not available in the current NuGet package. Please try the latest MyGet build before reporting issues, in case your issue has already been fixed but not released.
The AutoMapper MyGet gallery is available here. Be sure to include prereleases.
Installing the Package¶
If you want to install the latest MyGet package into a project, you can use the following command:
Install-Package AutoMapper -Source https://www.myget.org/F/automapperdev/api/v3/index.json -IncludePrerelease
Flattening¶
One of the common usages of object-object mapping is to take a complex object model and flatten it to a simpler model. You can take a complex model such as:
public class Order
{
private readonly IList<OrderLineItem> _orderLineItems = new List<OrderLineItem>();
public Customer Customer { get; set; }
public OrderLineItem[] GetOrderLineItems()
{
return _orderLineItems.ToArray();
}
public void AddOrderLineItem(Product product, int quantity)
{
_orderLineItems.Add(new OrderLineItem(product, quantity));
}
public decimal GetTotal()
{
return _orderLineItems.Sum(li => li.GetTotal());
}
}
public class Product
{
public decimal Price { get; set; }
public string Name { get; set; }
}
public class OrderLineItem
{
public OrderLineItem(Product product, int quantity)
{
Product = product;
Quantity = quantity;
}
public Product Product { get; private set; }
public int Quantity { get; private set;}
public decimal GetTotal()
{
return Quantity*Product.Price;
}
}
public class Customer
{
public string Name { get; set; }
}
We want to flatten this complex Order object into a simpler OrderDto that contains only the data needed for a certain scenario:
public class OrderDto
{
public string CustomerName { get; set; }
public decimal Total { get; set; }
}
When you configure a source/destination type pair in AutoMapper, the configurator attempts to match properties and methods on the source type to properties on the destination type. If for any property on the destination type a property, method, or a method prefixed with “Get” does not exist on the source type, AutoMapper splits the destination member name into individual words (by PascalCase conventions).
// Complex model
var customer = new Customer
{
Name = "George Costanza"
};
var order = new Order
{
Customer = customer
};
var bosco = new Product
{
Name = "Bosco",
Price = 4.99m
};
order.AddOrderLineItem(bosco, 15);
// Configure AutoMapper
Mapper.Initialize(cfg => cfg.CreateMap<Order, OrderDto>());
// Perform mapping
OrderDto dto = Mapper.Map<Order, OrderDto>(order);
dto.CustomerName.ShouldEqual("George Costanza");
dto.Total.ShouldEqual(74.85m);
We configured the type map in AutoMapper with the CreateMap method. AutoMapper can only map type pairs it knows about, so we have explicitly register the source/destination type pair with CreateMap. To perform the mapping, we use the Map method.
On the OrderDto type, the Total property matched to the GetTotal() method on Order. The CustomerName property matched to the Customer.Name property on Order. As long as we name our destination properties appropriately, we do not need to configure individual property matching.
IncludeMembers¶
If you need more control when flattening, you can use IncludeMembers. You can map members of a child object to the destination object when you already have a map from the child type to the destination type (unlike the classic flattening that doesn’t require a map for the child type).
class Source
{
public string Name { get; set; }
public InnerSource InnerSource { get; set; }
public OtherInnerSource OtherInnerSource { get; set; }
}
class InnerSource
{
public string Name { get; set; }
public string Description { get; set; }
}
class OtherInnerSource
{
public string Name { get; set; }
public string Description { get; set; }
public string Title { get; set; }
}
class Destination
{
public string Name { get; set; }
public string Description { get; set; }
public string Title { get; set; }
}
cfg.CreateMap<Source, Destination>().IncludeMembers(s=>s.InnerSource, s=>s.OtherInnerSource);
cfg.CreateMap<InnerSource, Destination>(MemberList.None);
cfg.CreateMap<OtherInnerSource, Destination>(MemberList.None);
var source = new Source { Name = "name", InnerSource = new InnerSource{ Description = "description" },
OtherInnerSource = new OtherInnerSource{ Title = "title" } };
var destination = mapper.Map<Destination>(source);
destination.Name.ShouldBe("name");
destination.Description.ShouldBe("description");
destination.Title.ShouldBe("title");
So this allows you to reuse the configuration in the existing maps for the child types InnerSource
and OtherInnerSource
when mapping the parent types Source
and Destination
. It works in a similar way to mapping inheritance, but it uses composition, not inheritance.
The order of the parameters in the IncludeMembers
call is relevant. When mapping a destination member, the first match wins, starting with the source object itself and then with the included child objects in the order you specified. So in the example above, Name
is mapped from the source object itself and Description
from InnerSource
because it’s the first match.
IncludeMembers integrates with ReverseMap
. An included member will be reversed to
ForMember(destination => destination.IncludedMember, member => member.MapFrom(source => source))
and the other way around. If that’s not what you want, you can avoid ReverseMap
(explicitly create the reverse map) or you can override the default settings (using Ignore
or IncludeMembers
without parameters respectively).
For details, check the tests.
Reverse Mapping and Unflattening¶
Starting with 6.1.0, AutoMapper now supports richer reverse mapping support. Given our entities:
public class Order {
public decimal Total { get; set; }
public Customer Customer { get; set; }
}
public class Customer {
public string Name { get; set; }
}
We can flatten this into a DTO:
public class OrderDto {
public decimal Total { get; set; }
public string CustomerName { get; set; }
}
We can map both directions, including unflattening:
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.ReverseMap();
});
By calling ReverseMap
, AutoMapper creates a reverse mapping configuration that includes unflattening:
var customer = new Customer {
Name = "Bob"
};
var order = new Order {
Customer = customer,
Total = 15.8m
};
var orderDto = Mapper.Map<Order, OrderDto>(order);
orderDto.CustomerName = "Joe";
Mapper.Map(orderDto, order);
order.Customer.Name.ShouldEqual("Joe");
Unflattening is only configured for ReverseMap
. If you want unflattening, you must configure Entity
-> Dto
then call ReverseMap
to create an unflattening type map configuration from the Dto
-> Entity
.
Customizing reverse mapping¶
AutoMapper will automatically reverse map “Customer.Name” from “CustomerName” based on the original flattening. If you use MapFrom, AutoMapper will attempt to reverse the map:
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.CustomerName, opt => opt.MapFrom(src => src.Customer.Name))
.ReverseMap();
As long as the MapFrom
path are member accessors, AutoMapper will unflatten from the same path (CustomerName
=> Customer.Name
).
If you need to customize this, for a reverse map you can use ForPath
:
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.CustomerName, opt => opt.MapFrom(src => src.Customer.Name))
.ReverseMap()
.ForPath(s => s.Customer.Name, opt => opt.MapFrom(src => src.CustomerName));
For most cases you shouldn’t need this, as the original MapFrom will be reversed for you. Use ForPath when the path to get and set the values are different.
If you do not want unflattening behavior, you can remove the call to ReverseMap
and create two separate maps. Or, you can use Ignore:
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.CustomerName, opt => opt.MapFrom(src => src.Customer.Name))
.ReverseMap()
.ForPath(s => s.Customer.Name, opt => opt.Ignore());
IncludeMembers¶
ReverseMap
also integrates with IncludeMembers
and configuration like
ForMember(destination => destination.IncludedMember, member => member.MapFrom(source => source))
Projection¶
Projection transforms a source to a destination beyond flattening the object model. Without extra configuration, AutoMapper requires a flattened destination to match the source type’s naming structure. When you want to project source values into a destination that does not exactly match the source structure, you must specify custom member mapping definitions. For example, we might want to turn this source structure:
public class CalendarEvent
{
public DateTime Date { get; set; }
public string Title { get; set; }
}
Into something that works better for an input form on a web page:
public class CalendarEventForm
{
public DateTime EventDate { get; set; }
public int EventHour { get; set; }
public int EventMinute { get; set; }
public string Title { get; set; }
}
Because the names of the destination properties do not exactly match the source property (CalendarEvent.Date
would need to be CalendarEventForm.EventDate
), we need to specify custom member mappings in our type map configuration:
// Model
var calendarEvent = new CalendarEvent
{
Date = new DateTime(2008, 12, 15, 20, 30, 0),
Title = "Company Holiday Party"
};
// Configure AutoMapper
Mapper.Initialize(cfg =>
cfg.CreateMap<CalendarEvent, CalendarEventForm>()
.ForMember(dest => dest.EventDate, opt => opt.MapFrom(src => src.Date.Date))
.ForMember(dest => dest.EventHour, opt => opt.MapFrom(src => src.Date.Hour))
.ForMember(dest => dest.EventMinute, opt => opt.MapFrom(src => src.Date.Minute)));
// Perform mapping
CalendarEventForm form = Mapper.Map<CalendarEvent, CalendarEventForm>(calendarEvent);
form.EventDate.ShouldEqual(new DateTime(2008, 12, 15));
form.EventHour.ShouldEqual(20);
form.EventMinute.ShouldEqual(30);
form.Title.ShouldEqual("Company Holiday Party");
Each custom member configuration uses an action delegate to configure each individual member. In the above example, we used the MapFrom
option to perform custom source-to-destination member mappings. The MapFrom
method takes a lambda expression as a parameter, which is then evaluated later during mapping. The MapFrom
expression can be any Func<TSource, object>
lambda expression.
Configuration Validation¶
Hand-rolled mapping code, though tedious, has the advantage of being testable. One of the inspirations behind AutoMapper was to eliminate not just the custom mapping code, but eliminate the need for manual testing. Because the mapping from source to destination is convention-based, you will still need to test your configuration.
AutoMapper provides configuration testing in the form of the AssertConfigurationIsValid method. Suppose we have slightly misconfigured our source and destination types:
public class Source
{
public int SomeValue { get; set; }
}
public class Destination
{
public int SomeValuefff { get; set; }
}
In the Destination type, we probably fat-fingered the destination property. Other typical issues are source member renames. To test our configuration, we simply create a unit test that sets up the configuration and executes the AssertConfigurationIsValid method:
Mapper.Initialize(cfg =>
cfg.CreateMap<Source, Destination>());
Mapper.Configuration.AssertConfigurationIsValid();
Executing this code produces an AutoMapperConfigurationException, with a descriptive message. AutoMapper checks to make sure that every single Destination type member has a corresponding type member on the source type.
Overriding configuration errors¶
To fix a configuration error (besides renaming the source/destination members), you have three choices for providing an alternate configuration:
- Custom Value Resolvers
- Projection
- Use the Ignore() option
With the third option, we have a member on the destination type that we will fill with alternative means, and not through the Map operation.
Mapper.Initialize(cfg =>
cfg.CreateMap<Source, Destination>()
.ForMember(dest => dest.SomeValuefff, opt => opt.Ignore())
);
Selecting members to validate¶
By default, AutoMapper uses the destination type to validate members. It assumes that all destination members need to be mapped. To modify this behavior, use the CreateMap
overload to specify which member list to validate against:
Mapper.Initialize(cfg =>
cfg.CreateMap<Source, Destination>(MemberList.Source);
cfg.CreateMap<Source2, Destination2>(MemberList.None);
);
To skip validation altogether for this map, use MemberList.None
.
Inline Mapping¶
AutoMapper creates type maps on the fly (new in 6.2.0). When you call Mapper.Map
for the first time, AutoMapper will create the type map configuration and compile the mapping plan. Subsequent mapping calls will use the compiled map.
Inline configuration¶
To configure an inline map, use the mapping options:
var source = new Source();
var dest = Mapper.Map<Source, Dest>(source, opt => opt.ConfigureMap().ForMember(dest => dest.Value, m => m.MapFrom(src => src.Value + 10)));
You can use local functions to make the configuration a little easier to read:
var source = new Source();
void ConfigureMap(IMappingOperationOptions<Source, Dest> opt) {
opt.ConfigureMap()
.ForMember(dest => dest.Value, m => m.MapFrom(src => src.Value + 10))
};
var dest = Mapper.Map<Source, Dest>(source, ConfigureMap);
You can use closures in this inline map as well to capture and use runtime values in your configuration:
int valueToAdd = 10;
var source = new Source();
void ConfigureMap(IMappingOperationOptions<Source, Dest> opt) {
opt.ConfigureMap()
.ForMember(dest => dest.Value, m => m.MapFrom(src => src.Value + valueToAdd))
};
var dest = Mapper.Map<Source, Dest>(source, ConfigureMap);
Inline validation¶
The first time the map is used, AutoMapper validates the map using the default validation configuration (destination members must all be mapped). Subsequent map calls skip mapping validation. This ensures you can safely map your objects.
You can configure the member list used to validate, to validate the source, destination, or no members to validate per map:
var source = new Source();
var dest = Mapper.Map<Source, Dest>(source, opt => opt.ConfigureMap(MemberList.None);
You can also turn off inline map validation altogether (not recommended unless you’re explicitly testing all of your maps):
Mapper.Initialize(cfg => cfg.ValidateInlineMaps = false);
Disabling inline maps¶
To turn off inline mapping:
Mapper.Initialize(cfg => cfg.CreateMissingTypeMaps = false);
Lists and Arrays¶
AutoMapper only requires configuration of element types, not of any array or list type that might be used. For example, we might have a simple source and destination type:
public class Source
{
public int Value { get; set; }
}
public class Destination
{
public int Value { get; set; }
}
All the basic generic collection types are supported:
Mapper.Initialize(cfg => cfg.CreateMap<Source, Destination>());
var sources = new[]
{
new Source { Value = 5 },
new Source { Value = 6 },
new Source { Value = 7 }
};
IEnumerable<Destination> ienumerableDest = Mapper.Map<Source[], IEnumerable<Destination>>(sources);
ICollection<Destination> icollectionDest = Mapper.Map<Source[], ICollection<Destination>>(sources);
IList<Destination> ilistDest = Mapper.Map<Source[], IList<Destination>>(sources);
List<Destination> listDest = Mapper.Map<Source[], List<Destination>>(sources);
Destination[] arrayDest = Mapper.Map<Source[], Destination[]>(sources);
To be specific, the source collection types supported include:
- IEnumerable
- IEnumerable<T>
- ICollection
- ICollection<T>
- IList
- IList<T>
- List<T>
- Arrays
For the non-generic enumerable types, only unmapped, assignable types are supported, as AutoMapper will be unable to “guess” what types you’re trying to map. As shown in the example above, it’s not necessary to explicitly configure list types, only their member types.
When mapping to an existing collection, the destination collection is cleared first. If this is not what you want, take a look at AutoMapper.Collection.
Handling null collections¶
When mapping a collection property, if the source value is null AutoMapper will map the destination field to an empty collection rather than setting the destination value to null. This aligns with the behavior of Entity Framework and Framework Design Guidelines that believe C# references, arrays, lists, collections, dictionaries and IEnumerables should NEVER be null, ever.
This behavior can be changed by setting the AllowNullCollections
property to true when configuring the mapper.
Mapper.Initialize(cfg => {
cfg.AllowNullCollections = true;
cfg.CreateMap<Source, Destination>();
});
Polymorphic element types in collections¶
Many times, we might have a hierarchy of types in both our source and destination types. AutoMapper supports polymorphic arrays and collections, such that derived source/destination types are used if found.
public class ParentSource
{
public int Value1 { get; set; }
}
public class ChildSource : ParentSource
{
public int Value2 { get; set; }
}
public class ParentDestination
{
public int Value1 { get; set; }
}
public class ChildDestination : ParentDestination
{
public int Value2 { get; set; }
}
AutoMapper still requires explicit configuration for child mappings, as AutoMapper cannot “guess” which specific child destination mapping to use. Here is an example of the above types:
Mapper.Initialize(c=> {
c.CreateMap<ParentSource, ParentDestination>()
.Include<ChildSource, ChildDestination>();
c.CreateMap<ChildSource, ChildDestination>();
});
var sources = new[]
{
new ParentSource(),
new ChildSource(),
new ParentSource()
};
var destinations = Mapper.Map<ParentSource[], ParentDestination[]>(sources);
destinations[0].ShouldBeInstanceOf<ParentDestination>();
destinations[1].ShouldBeInstanceOf<ChildDestination>();
destinations[2].ShouldBeInstanceOf<ParentDestination>();
Nested Mappings¶
As the mapping engine executes the mapping, it can use one of a variety of methods to resolve a destination member value. One of these methods is to use another type map, where the source member type and destination member type are also configured in the mapping configuration. This allows us to not only flatten our source types, but create complex destination types as well. For example, our source type might contain another complex type:
public class OuterSource
{
public int Value { get; set; }
public InnerSource Inner { get; set; }
}
public class InnerSource
{
public int OtherValue { get; set; }
}
We could simply flatten the OuterSource.Inner.OtherValue to one InnerOtherValue property, but we might also want to create a corresponding complex type for the Inner property:
public class OuterDest
{
public int Value { get; set; }
public InnerDest Inner { get; set; }
}
public class InnerDest
{
public int OtherValue { get; set; }
}
In that case, we would need to configure the additional source/destination type mappings:
var config = new MapperConfiguration(cfg => {
cfg.CreateMap<OuterSource, OuterDest>();
cfg.CreateMap<InnerSource, InnerDest>();
});
config.AssertConfigurationIsValid();
var source = new OuterSource
{
Value = 5,
Inner = new InnerSource {OtherValue = 15}
};
var mapper = config.CreateMapper();
var dest = mapper.Map<OuterSource, OuterDest>(source);
dest.Value.ShouldEqual(5);
dest.Inner.ShouldNotBeNull();
dest.Inner.OtherValue.ShouldEqual(15);
A few things to note here:
- Order of configuring types does not matter
- Call to Map does not need to specify any inner type mappings, only the type map to use for the source value passed in
With both flattening and nested mappings, we can create a variety of destination shapes to suit whatever our needs may be.
Custom Type Converters¶
Sometimes, you need to take complete control over the conversion of one type to another. This is typically when one type looks nothing like the other, a conversion function already exists, and you would like to go from a “looser” type to a stronger type, such as a source type of string to a destination type of Int32.
For example, suppose we have a source type of:
public class Source
{
public string Value1 { get; set; }
public string Value2 { get; set; }
public string Value3 { get; set; }
}
But you would like to map it to:
public class Destination
{
public int Value1 { get; set; }
public DateTime Value2 { get; set; }
public Type Value3 { get; set; }
}
If we were to try and map these two types as-is, AutoMapper would throw an exception (at map time and configuration-checking time), as AutoMapper does not know about any mapping from string to int, DateTime or Type. To create maps for these types, we must supply a custom type converter, and we have three ways of doing so:
void ConvertUsing(Func<TSource, TDestination> mappingFunction);
void ConvertUsing(ITypeConverter<TSource, TDestination> converter);
void ConvertUsing<TTypeConverter>() where TTypeConverter : ITypeConverter<TSource, TDestination>;
The first option is simply any function that takes a source and returns a destination (there are several overloads too). This works for simple cases, but becomes unwieldy for larger ones. In more difficult cases, we can create a custom ITypeConverter<TSource, TDestination>:
public interface ITypeConverter<in TSource, TDestination>
{
TDestination Convert(TSource source, TDestination destination, ResolutionContext context);
}
And supply AutoMapper with either an instance of a custom type converter, or simply the type, which AutoMapper will instantiate at run time. The mapping configuration for our above source/destination types then becomes:
[Test]
public void Example()
{
Mapper.Initialize(cfg => {
cfg.CreateMap<string, int>().ConvertUsing(s => Convert.ToInt32(s));
cfg.CreateMap<string, DateTime>().ConvertUsing(new DateTimeTypeConverter());
cfg.CreateMap<string, Type>().ConvertUsing<TypeTypeConverter>();
cfg.CreateMap<Source, Destination>();
});
Mapper.AssertConfigurationIsValid();
var source = new Source
{
Value1 = "5",
Value2 = "01/01/2000",
Value3 = "AutoMapperSamples.GlobalTypeConverters.GlobalTypeConverters+Destination"
};
Destination result = Mapper.Map<Source, Destination>(source);
result.Value3.ShouldEqual(typeof(Destination));
}
public class DateTimeTypeConverter : ITypeConverter<string, DateTime>
{
public DateTime Convert(string source, DateTime destination, ResolutionContext context)
{
return System.Convert.ToDateTime(source);
}
}
public class TypeTypeConverter : ITypeConverter<string, Type>
{
public Type Convert(string source, Type destination, ResolutionContext context)
{
return Assembly.GetExecutingAssembly().GetType(source);
}
}
In the first mapping, from string to Int32, we simply use the built-in Convert.ToInt32 function (supplied as a method group). The next two use custom ITypeConverter implementations.
The real power of custom type converters is that they are used any time AutoMapper finds the source/destination pairs on any mapped types. We can build a set of custom type converters, on top of which other mapping configurations use, without needing any extra configuration. In the above example, we never have to specify the string/int conversion again. Where as Custom Value Resolvers have to be configured at a type member level, custom type converters are global in scope.
System Type Converters¶
The .NET Framework also supports the concepts of type converters, through the TypeConverter class. AutoMapper supports these types of type converters, in configuration checking and mapping, without the need for any manual configuration. AutoMapper uses the TypeDescriptor.GetConverter method for determining if the source/destination type pair can be mapped.
Custom Value Resolvers¶
Although AutoMapper covers quite a few destination member mapping scenarios, there are the 1 to 5% of destination values that need a little help in resolving. Many times, this custom value resolution logic is domain logic that can go straight on our domain. However, if this logic pertains only to the mapping operation, it would clutter our source types with unnecessary behavior. In these cases, AutoMapper allows for configuring custom value resolvers for destination members. For example, we might want to have a calculated value just during mapping:
public class Source
{
public int Value1 { get; set; }
public int Value2 { get; set; }
}
public class Destination
{
public int Total { get; set; }
}
For whatever reason, we want Total to be the sum of the source Value properties. For some other reason, we can’t or shouldn’t put this logic on our Source type. To supply a custom value resolver, we’ll need to first create a type that implements IValueResolver:
public interface IValueResolver<in TSource, in TDestination, TDestMember>
{
TDestMember Resolve(TSource source, TDestination destination, TDestMember destMember, ResolutionContext context);
}
The ResolutionContext contains all of the contextual information for the current resolution operation, such as source type, destination type, source value and so on. An example implementation:
public class CustomResolver : IValueResolver<Source, Destination, int>
{
public int Resolve(Source source, Destination destination, int member, ResolutionContext context)
{
return source.Value1 + source.Value2;
}
}
Once we have our IValueResolver implementation, we’ll need to tell AutoMapper to use this custom value resolver when resolving a specific destination member. We have several options in telling AutoMapper a custom value resolver to use, including:
- MapFrom<TValueResolver>
- MapFrom(typeof(CustomValueResolver))
- MapFrom(aValueResolverInstance)
In the below example, we’ll use the first option, telling AutoMapper the custom resolver type through generics:
Mapper.Initialize(cfg =>
cfg.CreateMap<Source, Destination>()
.ForMember(dest => dest.Total, opt => opt.MapFrom<CustomResolver>()));
Mapper.AssertConfigurationIsValid();
var source = new Source
{
Value1 = 5,
Value2 = 7
};
var result = Mapper.Map<Source, Destination>(source);
result.Total.ShouldEqual(12);
Although the destination member (Total) did not have any matching source member, specifying a custom resolver made the configuration valid, as the resolver is now responsible for supplying a value for the destination member.
If we don’t care about the source/destination types in our value resolver, or want to reuse them across maps, we can just use “object” as the source/destination types:
public class MultBy2Resolver : IValueResolver<object, object, int> {
public int Resolve(object source, object dest, int destMember, ResolutionContext context) {
return destMember * 2;
}
}
Custom constructor methods¶
Because we only supplied the type of the custom resolver to AutoMapper, the mapping engine will use reflection to create an instance of the value resolver.
If we don’t want AutoMapper to use reflection to create the instance, we can supply it directly:
Mapper.Initialize(cfg => cfg.CreateMap<Source, Destination>()
.ForMember(dest => dest.Total,
opt => opt.MapFrom(new CustomResolver())
));
AutoMapper will use that specific object, helpful in scenarios where the resolver might have constructor arguments or need to be constructed by an IoC container.
Customizing the source value supplied to the resolver¶
By default, AutoMapper passes the source object to the resolver. This limits the reusability of resolvers, since the resolver is coupled to the source type. If, however, we supply a common resolver across multiple types, we configure AutoMapper to redirect the source value supplied to the resolver, and also use a different resolver interface so that our resolver can get use of the source/destination members:
Mapper.Initialize(cfg => {
cfg.CreateMap<Source, Destination>()
.ForMember(dest => dest.Total,
opt => opt.MapFrom<CustomResolver, decimal>(src => src.SubTotal));
cfg.CreateMap<OtherSource, OtherDest>()
.ForMember(dest => dest.OtherTotal,
opt => opt.MapFrom<CustomResolver, decimal>(src => src.OtherSubTotal));
});
public class CustomResolver : IMemberValueResolver<object, object, decimal, decimal> {
public decimal Resolve(object source, object destination, decimal sourceMember, decimal destinationMember, ResolutionContext context) {
// logic here
}
}
Passing in key-value to Mapper¶
When calling map you can pass in extra objects by using key-value and using a custom resolver to get the object from context.
Mapper.Map<Source, Dest>(src, opt => opt.Items["Foo"] = "Bar");
This is how to setup the mapping for this custom resolver
cfg.CreateMap<Source, Dest>()
.ForMember(dest => dest.Foo, opt => opt.MapFrom((src, dest, destMember, context) => context.Items["Foo"]));
Resolvers and conditions¶
For each property mapping, AutoMapper attempts to resolve the destination value before evaluating the condition. So it needs to be able to do that without throwing an exception even if the condition will prevent the resulting value from being used.
As an example, here’s sample output from BuildExecutionPlan (displayed using ReadableExpressions) for a single property:
try
{
var resolvedValue =
{
try
{
return // ... tries to resolve the destination value here
}
catch (NullReferenceException)
{
return null;
}
catch (ArgumentNullException)
{
return null;
}
};
if (condition.Invoke(src, typeMapDestination, resolvedValue))
{
typeMapDestination.WorkStatus = resolvedValue;
}
}
catch (Exception ex)
{
throw new AutoMapperMappingException(
"Error mapping types.",
ex,
AutoMapper.TypePair,
AutoMapper.TypeMap,
AutoMapper.PropertyMap);
};
The default generated code for resolving a property, if you haven’t customized the mapping for that member, generally doesn’t have any problems. But if you’re using custom code to map the property that will crash if the condition isn’t met, the mapping will fail despite the condition.
This example code would fail:
public class SourceClass
{
public string Value { get; set; }
}
public class TargetClass
{
public int ValueLength { get; set; }
}
// ...
var source = new SourceClass { Value = null };
var target = new TargetClass;
CreateMap<SourceClass, TargetClass>()
.ForMember(d => d.ValueLength, o => o.MapFrom(s => s.Value.Length))
.ForAllMembers(o => o.Condition((src, dest, value) => value != null));
The condition prevents the Value property from being mapped onto the target, but the custom member mapping would fail before that point because it calls Value.Length, and Value is null.
Prevent this by using a PreCondition instead or by ensuring the custom member mapping code can complete safely regardless of conditions:
.ForMember(d => d.ValueLength, o => o.MapFrom(s => s != null ? s.Value.Length : 0))
Value Converters¶
Value converters are a cross between Type Converters and Value Resolvers. Type converters are globally scoped, so that any time you map from type Foo
to type Bar
in any mapping, the type converter will be used. Value converters are scoped to a single map, and receive the source and destination objects to resolve to a value to map to the destination member. Optionally value converters can receive the source member as well.
In simplified syntax:
- Type converter =
Func<TSource, TDestination, TDestination>
- Value resolver =
Func<TSource, TDestination, TDestinationMember>
- Member value resolver =
Func<TSource, TDestination, TSourceMember, TDestinationMember>
- Value converter =
Func<TSourceMember, TDestinationMember>
To configure a value converter, use at the member level:
public class CurrencyFormatter : IValueConverter<decimal, string> {
public string Convert(decimal source)
=> source.ToString("c");
}
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.Amount, opt => opt.ConvertUsing(new CurrencyFormatter()));
cfg.CreateMap<OrderLineItem, OrderLineItemDto>()
.ForMember(d => d.Total, opt => opt.ConvertUsing(new CurrencyFormatter()));
});
You can customize the source member when the source member name does not match:
public class CurrencyFormatter : IValueConverter<decimal, string> {
public string Convert(decimal source)
=> source.ToString("c");
}
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.Amount, opt => opt.ConvertUsing(new CurrencyFormatter(), src => src.OrderAmount));
cfg.CreateMap<OrderLineItem, OrderLineItemDto>()
.ForMember(d => d.Total, opt => opt.ConvertUsing(new CurrencyFormatter(), src => src.LITotal));
});
If you need the value converters instantiated by the service locator, you can specify the type instead:
public class CurrencyFormatter : IValueConverter<decimal, string> {
public string Convert(decimal source)
=> source.ToString("c");
}
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.ForMember(d => d.Amount, opt => opt.ConvertUsing<CurrencyFormatter, decimal>());
cfg.CreateMap<OrderLineItem, OrderLineItemDto>()
.ForMember(d => d.Total, opt => opt.ConvertUsing<CurrencyFormatter, decimal>());
});
If you do not know the types or member names at runtime, use the various overloads that accept System.Type
and string
-based members:
public class CurrencyFormatter : IValueConverter<decimal, string> {
public string Convert(decimal source)
=> source.ToString("c");
}
Mapper.Initialize(cfg => {
cfg.CreateMap(typeof(Order), typeof(OrderDto))
.ForMember("Amount", opt => opt.ConvertUsing(new CurrencyFormatter(), "OrderAmount"));
cfg.CreateMap(typeof(OrderLineItem), typeof(OrderLineItemDto))
.ForMember("Total", opt => opt.ConvertUsing(new CurrencyFormatter(), "LITotal"));
});
Value converters are only used for in-memory mapping execution. They will not work for ProjectTo
.
Value Transformers¶
Value transformers apply an additional transformation to a single type. Before assigning the value, AutoMapper will check to see if the value to be set has any value transformations associated, and will apply them before setting.
You can create value transformers at several different levels:
- Globally
- Profile
- Map
- Member
Mapper.Initialize(cfg => {
cfg.ValueTransformers.Add<string>(val => val + "!!!");
});
var source = new Source { Value = "Hello" };
var dest = Mapper.Map<Dest>(source);
dest.Value.ShouldBe("Hello!!!");
Null Substitution¶
Null substitution allows you to supply an alternate value for a destination member if the source value is null anywhere along the member chain. This means that instead of mapping from null, it will map from the value you supply.
var config = new MapperConfiguration(cfg => cfg.CreateMap<Source, Dest>()
.ForMember(destination => destination.Value, opt => opt.NullSubstitute("Other Value")));
var source = new Source { Value = null };
var mapper = config.CreateMapper();
var dest = mapper.Map<Source, Dest>(source);
dest.Value.ShouldEqual("Other Value");
source.Value = "Not null";
dest = mapper.Map<Source, Dest>(source);
dest.Value.ShouldEqual("Not null");
The substitute is assumed to be of the source member type, and will go through any mapping/conversion after to the destination type.
Before and After Map Action¶
Occasionally, you might need to perform custom logic before or after a map occurs. These should be a rarity, as it’s more obvious to do this work outside of AutoMapper. You can create global before/after map actions:
Mapper.Initialize(cfg => {
cfg.CreateMap<Source, Dest>()
.BeforeMap((src, dest) => src.Value = src.Value + 10)
.AfterMap((src, dest) => dest.Name = "John");
});
Or you can create before/after map callbacks during mapping:
int i = 10;
Mapper.Map<Source, Dest>(src, opt => {
opt.BeforeMap((src, dest) => src.Value = src.Value + i);
opt.AfterMap((src, dest) => dest.Name = HttpContext.Current.Identity.Name);
});
The latter configuration is helpful when you need contextual information fed into before/after map actions.
Using IMappingAction
¶
You can encapsulate Before and After Map Actions into small reusable classes. Those classes need to implement the IMappingAction<in TSource, in TDestination>
interface.
Using the previous example, here is an encapsulation of naming some objects “John”:
public class NameMeJohnAction : IMappingAction<SomePersonObject, SomeOtherPersonObject>
{
public void Process(SomePersonObject source, SomeOtherPersonObject destination)
{
destination.Name = "John";
}
}
Mapper.Initialize(cfg => {
cfg.CreateMap<SomePersonObject, SomeOtherPersonObject>()
.AfterMap<NameMeJohnAction>();
});
Asp.Net Core and AutoMapper.Extensions.Microsoft.DependencyInjection
¶
If you are using Asp.Net Core and the AutoMapper.Extensions.Microsoft.DependencyInjection
package, this is also a good way of using Dependency Injection. You can’t inject dependencies into Profile
classes, but you can do it in IMappingAction
implementations.
The following example shows how to connect an IMappingAction
accessing the current HttpContext
to a Profile
after map action, leveraging Dependency Injection:
public class SetTraceIdentifierAction : IMappingAction<SomeModel, SomeOtherModel>
{
private readonly IHttpContextAccessor _httpContextAccessor;
public SetTraceIdentifierAction(IHttpContextAccessor httpContextAccessor)
{
_httpContextAccessor = httpContextAccessor ?? throw new ArgumentNullException(nameof(httpContextAccessor));
}
public void Process(SomeModel source, SomeOtherModel destination)
{
destination.TraceIdentifier = _httpContextAccessor.HttpContext.TraceIdentifier;
}
}
public class SomeProfile : Profile
{
public SomeProfile()
{
CreateMap<SomeModel, SomeOtherModel>()
.AfterMap<SetTraceIdentifierAction>();
}
}
Everything is connected together by:
public class Startup
{
public void ConfigureServices(IServiceCollection services)
{
services.AddAutoMapper(typeof(Startup).Assembly);
}
//..
}
See AutoMapper.Extensions.Microsoft.DependencyInjection
for more info.
Dependency Injection¶
AutoMapper supports the ability to construct Custom Value Resolvers, Custom Type Converters, and Value Converters using static service location:
Mapper.Initialize(cfg =>
{
cfg.ConstructServicesUsing(ObjectFactory.GetInstance);
cfg.CreateMap<Source, Destination>();
});
Or dynamic service location, to be used in the case of instance-based containers (including child/nested containers):
var mapper = new Mapper(Mapper.Configuration, childContainer.GetInstance);
var dest = mapper.Map<Source, Destination>(new Source { Value = 15 });
Queryable Extensions¶
Starting with 8.0 you can use IMapper.ProjectTo. For older versions you need to pass the configuration to the extension method IQueryable.ProjectTo<T>(IConfigurationProvider)
.
Note that IQueryable.ProjectTo is more limited than IMapper.Map, as only what is allowed by the underlying LINQ provider is supported. That means you cannot use DI with value resolvers and converters as you can with Map.
Examples¶
ASP.NET Core¶
There is a NuGet package to be used with the default injection mechanism described here and used in this project.
Once the nuget package is downloaded, simply add AutoMapper to your IServiceCollection in your startup.cs class:
services.AddAutoMapper(assembly1, assembly2 /*, ...*/);
or marker types:
services.AddAutoMapper(type1, type2 /*, ...*/);
Now you can inject AutoMapper at runtime into your services/controllers:
public class EmployeesController {
private readonly IMapper _mapper;
public EmployeesController(IMapper mapper)
=> _mapper = mapper;
// use _mapper.Map to map
}
Ninject¶
For those using Ninject here is an example of a Ninject module for AutoMapper
public class AutoMapperModule : NinjectModule
{
public override void Load()
{
Bind<IValueResolver<SourceEntity, DestModel, bool>>().To<MyResolver>();
var mapperConfiguration = CreateConfiguration();
Bind<MapperConfiguration>().ToConstant(mapperConfiguration).InSingletonScope();
// This teaches Ninject how to create automapper instances say if for instance
// MyResolver has a constructor with a parameter that needs to be injected
Bind<IMapper>().ToMethod(ctx =>
new Mapper(mapperConfiguration, type => ctx.Kernel.Get(type)));
}
private MapperConfiguration CreateConfiguration()
{
var config = new MapperConfiguration(cfg =>
{
// Add all profiles in current assembly
cfg.AddProfiles(GetType().Assembly);
});
return config;
}
}
Simple Injector¶
The workflow is as follows:
- Register your types via MyRegistrar.Register
- The MapperProvider allows you to directly inject an instance of IMapper into your other classes
- SomeProfile resolves a value using PropertyThatDependsOnIocValueResolver
- PropertyThatDependsOnIocValueResolver has IService injected into it, which is then able to be used
The ValueResolver has access to IService because we register our container via MapperConfigurationExpression.ConstructServicesUsing
public class MyRegistrar
{
public void Register(Container container)
{
// Injectable service
container.RegisterSingleton<IService, SomeService>();
// Automapper
container.RegisterSingleton(() => GetMapper(container));
}
private AutoMapper.IMapper GetMapper(Container container)
{
var mp = container.GetInstance<MapperProvider>();
return mp.GetMapper();
}
}
public class MapperProvider
{
private readonly Container _container;
public MapperProvider(Container container)
{
_container = container;
}
public IMapper GetMapper()
{
var mce = new MapperConfigurationExpression();
mce.ConstructServicesUsing(_container.GetInstance);
mce.AddProfiles(typeof(SomeProfile).Assembly);
var mc = new MapperConfiguration(mce);
mc.AssertConfigurationIsValid();
IMapper m = new Mapper(mc, t => _container.GetInstance(t));
return m;
}
}
public class SomeProfile : Profile
{
public SomeProfile()
{
var map = CreateMap<MySourceType, MyDestinationType>();
map.ForMember(d => d.PropertyThatDependsOnIoc, opt => opt.MapFrom<PropertyThatDependsOnIocValueResolver>());
}
}
public class PropertyThatDependsOnIocValueResolver : IValueResolver<MySourceType, object, int>
{
private readonly IService _service;
public PropertyThatDependsOnIocValueResolver(IService service)
{
_service = service;
}
int IValueResolver<MySourceType, object, int>.Resolve(MySourceType source, object destination, int destMember, ResolutionContext context)
{
return _service.MyMethod(source);
}
}
Castle Windsor¶
For those using Castle Windsor here is an example of an installer for AutoMapper
public class AutoMapperInstaller : IWindsorInstaller
{
public void Install(IWindsorContainer container, IConfigurationStore store)
{
// Register all mapper profiles
container.Register(
Classes.FromAssemblyInThisApplication(GetType().Assembly)
.BasedOn<Profile>().WithServiceBase());
// Register IConfigurationProvider with all registered profiles
container.Register(Component.For<IConfigurationProvider>().UsingFactoryMethod(kernel =>
{
return new MapperConfiguration(configuration =>
{
kernel.ResolveAll<Profile>().ToList().ForEach(configuration.AddProfile);
});
}).LifestyleSingleton());
// Register IMapper with registered IConfigurationProvider
container.Register(
Component.For<IMapper>().UsingFactoryMethod(kernel =>
new Mapper(kernel.Resolve<IConfigurationProvider>(), kernel.Resolve)));
}
}
Mapping Inheritance¶
Mapping inheritance serves two functions:
- Inheriting mapping configuration from a base class or interface configuration
- Runtime polymorphic mapping
Inheriting base class configuration is opt-in, and you can either explicitly specify the mapping to inherit from the base type configuration with Include
or in the derived type configuration with IncludeBase
:
CreateMap<BaseEntity, BaseDto>()
.Include<DerivedEntity, DerivedDto>()
.ForMember(dest => dest.SomeMember, opt => opt.MapFrom(src => src.OtherMember));
CreateMap<DerivedEntity, DerivedDto>();
or
CreateMap<BaseEntity, BaseDto>()
.ForMember(dest => dest.SomeMember, opt => opt.MapFrom(src => src.OtherMember));
CreateMap<DerivedEntity, DerivedDto>()
.IncludeBase<BaseEntity, BaseDto>();
In each case above, the derived mapping inherits the custom mapping configuration from the base mapping configuration.
To include all derived maps, from the base type map configuration:
CreateMap<BaseEntity, BaseDto>()
.IncludeAllDerived();
CreateMap<DerivedEntity, DerivedDto>();
Runtime polymorphism¶
Take:
public class Order { }
public class OnlineOrder : Order { }
public class MailOrder : Order { }
public class OrderDto { }
public class OnlineOrderDto : OrderDto { }
public class MailOrderDto : OrderDto { }
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.Include<OnlineOrder, OnlineOrderDto>()
.Include<MailOrder, MailOrderDto>();
cfg.CreateMap<OnlineOrder, OnlineOrderDto>();
cfg.CreateMap<MailOrder, MailOrderDto>();
});
// Perform Mapping
var order = new OnlineOrder();
var mapped = Mapper.Map(order, order.GetType(), typeof(OrderDto));
Assert.IsType<OnlineOrderDto>(mapped);
You will notice that because the mapped object is a OnlineOrder, AutoMapper has seen you have a more specific mapping for OnlineOrder than OrderDto, and automatically chosen that.
Specifying inheritance in derived classes¶
Instead of configuring inheritance from the base class, you can specify inheritance from the derived classes:
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.ForMember(o => o.Id, m => m.MapFrom(s => s.OrderId));
cfg.CreateMap<OnlineOrder, OnlineOrderDto>()
.IncludeBase<Order, OrderDto>();
cfg.CreateMap<MailOrder, MailOrderDto>()
.IncludeBase<Order, OrderDto>();
});
As¶
For simple cases, you can use As
to redirect a base map to an existing derived map:
cfg.CreateMap<Order, OnlineOrderDto>();
cfg.CreateMap<Order, OrderDto>().As<OnlineOrderDto>();
mapper.Map<OrderDto>(new Order()).ShouldBeOfType<OnlineOrderDto>();
Inheritance Mapping Priorities¶
This introduces additional complexity because there are multiple ways a property can be mapped. The priority of these sources are as follows
- Explicit Mapping (using .MapFrom())
- Inherited Explicit Mapping
- Ignore Property Mapping
- Convention Mapping (Properties that are matched via convention)
To demonstrate this, lets modify our classes shown above
//Domain Objects
public class Order { }
public class OnlineOrder : Order
{
public string Referrer { get; set; }
}
public class MailOrder : Order { }
//Dtos
public class OrderDto
{
public string Referrer { get; set; }
}
//Mappings
Mapper.Initialize(cfg => {
cfg.CreateMap<Order, OrderDto>()
.Include<OnlineOrder, OrderDto>()
.Include<MailOrder, OrderDto>()
.ForMember(o=>o.Referrer, m=>m.Ignore());
cfg.CreateMap<OnlineOrder, OrderDto>();
cfg.CreateMap<MailOrder, OrderDto>();
});
// Perform Mapping
var order = new OnlineOrder { Referrer = "google" };
var mapped = Mapper.Map(order, order.GetType(), typeof(OrderDto));
Assert.IsNull(mapped.Referrer);
Notice that in our mapping configuration, we have ignored Referrer
(because it doesn’t exist in the order base class) and that has a higher priority than convention mapping, so the property doesn’t get mapped.
If you do want the Referrer
property to be mapped in the mapping from OnlineOrder
to OrderDto
you should include an explicit mapping in the mapping like this:
cfg.CreateMap<OnlineOrder, OrderDto>()
.ForMember(o=>o.Referrer, m=>m.MapFrom(x=>x.Referrer));
Overall this feature should make using AutoMapper with classes that leverage inheritance feel more natural.
Queryable Extensions¶
When using an ORM such as NHibernate or Entity Framework with AutoMapper’s standard Mapper.Map
functions, you may notice that the ORM will query all the fields of all the objects within a graph when AutoMapper is attempting to map the results to a destination type.
If your ORM exposes IQueryable
s, you can use AutoMapper’s QueryableExtensions helper methods to address this key pain.
Using Entity Framework for an example, say that you have an entity OrderLine
with a relationship with an entity Item
. If you want to map this to an OrderLineDTO
with the Item
’s Name
property, the standard Mapper.Map
call will result in Entity Framework querying the entire OrderLine
and Item
table.
Use this approach instead.
Given the following entities:
public class OrderLine
{
public int Id { get; set; }
public int OrderId { get; set; }
public Item Item { get; set; }
public decimal Quantity { get; set; }
}
public class Item
{
public int Id { get; set; }
public string Name { get; set; }
}
And the following DTO:
public class OrderLineDTO
{
public int Id { get; set; }
public int OrderId { get; set; }
public string Item { get; set; }
public decimal Quantity { get; set; }
}
You can use the Queryable Extensions like so:
Mapper.Initialize(cfg =>
cfg.CreateMap<OrderLine, OrderLineDTO>()
.ForMember(dto => dto.Item, conf => conf.MapFrom(ol => ol.Item.Name)));
public List<OrderLineDTO> GetLinesForOrder(int orderId)
{
using (var context = new orderEntities())
{
return context.OrderLines.Where(ol => ol.OrderId == orderId)
.ProjectTo<OrderLineDTO>().ToList();
}
}
The .ProjectTo<OrderLineDTO>()
will tell AutoMapper’s mapping engine to emit a select
clause to the IQueryable that will inform entity framework that it only needs to query the Name column of the Item table, same as if you manually projected your IQueryable
to an OrderLineDTO
with a Select
clause.
Note that for this feature to work, all type conversions must be explicitly handled in your Mapping. For example, you can not rely on the ToString()
override of the Item
class to inform entity framework to only select from the Name
column, and any data type changes, such as Double
to Decimal
must be explicitly handled as well.
The instance API¶
Starting with 8.0 there are similar ProjectTo methods on IMapper that feel more natural when you use IMapper with DI.
Preventing lazy loading/SELECT N+1 problems¶
Because the LINQ projection built by AutoMapper is translated directly to a SQL query by the query provider, the mapping occurs at the SQL/ADO.NET level, and not touching your entities. All data is eagerly fetched and loaded into your DTOs.
Nested collections use a Select to project child DTOs:
from i in db.Instructors
orderby i.LastName
select new InstructorIndexData.InstructorModel
{
ID = i.ID,
FirstMidName = i.FirstMidName,
LastName = i.LastName,
HireDate = i.HireDate,
OfficeAssignmentLocation = i.OfficeAssignment.Location,
Courses = i.Courses.Select(c => new InstructorIndexData.InstructorCourseModel
{
CourseID = c.CourseID,
CourseTitle = c.Title
}).ToList()
};
This map through AutoMapper will result in a SELECT N+1 problem, as each child Course
will be queried one at a time, unless specified through your ORM to eagerly fetch. With LINQ projection, no special configuration or specification is needed with your ORM. The ORM uses the LINQ projection to build the exact SQL query needed.
Custom projection¶
In the case where members names don’t line up, or you want to create calculated property, you can use MapFrom (the expression-based overload) to supply a custom expression for a destination member:
Mapper.Initialize(cfg => cfg.CreateMap<Customer, CustomerDto>()
.ForMember(d => d.FullName, opt => opt.MapFrom(c => c.FirstName + " " + c.LastName))
.ForMember(d => d.TotalContacts, opt => opt.MapFrom(c => c.Contacts.Count()));
AutoMapper passes the supplied expression with the built projection. As long as your query provider can interpret the supplied expression, everything will be passed down all the way to the database.
If the expression is rejected from your query provider (Entity Framework, NHibernate, etc.), you might need to tweak your expression until you find one that is accepted.
Custom Type Conversion¶
Occasionally, you need to completely replace a type conversion from a source to a destination type. In normal runtime mapping, this is accomplished via the ConvertUsing method. To perform the analog in LINQ projection, use the ConvertUsing method:
cfg.CreateMap<Source, Dest>().ConvertUsing(src => new Dest { Value = 10 });
The expression-based ConvertUsing
is slightly more limited than Func-based ConvertUsing
overloads as only what is allowed in an Expression and the underlying LINQ provider will work.
Custom destination type constructors¶
If your destination type has a custom constructor but you don’t want to override the entire mapping, use the ConstructUsing expression-based method overload:
cfg.CreateMap<Source, Dest>()
.ConstructUsing(src => new Dest(src.Value + 10));
AutoMapper will automatically match up destination constructor parameters to source members based on matching names, so only use this method if AutoMapper can’t match up the destination constructor properly, or if you need extra customization during construction.
String conversion¶
AutoMapper will automatically add ToString()
when the destination member type is a string and the source member type is not.
public class Order {
public OrderTypeEnum OrderType { get; set; }
}
public class OrderDto {
public string OrderType { get; set; }
}
var orders = dbContext.Orders.ProjectTo<OrderDto>().ToList();
orders[0].OrderType.ShouldEqual("Online");
Explicit expansion¶
In some scenarios, such as OData, a generic DTO is returned through an IQueryable controller action. Without explicit instructions, AutoMapper will expand all members in the result. To control which members are expanded during projection, set ExplicitExpansion in the configuration and then pass in the members you want to explicitly expand:
dbContext.Orders.ProjectTo<OrderDto>(
dest => dest.Customer,
dest => dest.LineItems);
// or string-based
dbContext.Orders.ProjectTo<OrderDto>(
null,
"Customer",
"LineItems");
// for collections
dbContext.Orders.ProjectTo<OrderDto>(
null,
dest => dest.LineItems.Select(item => item.Product));
For more information, see the tests.
Aggregations¶
LINQ can support aggregate queries, and AutoMapper supports LINQ extension methods. In the custom projection example, if we renamed the TotalContacts
property to ContactsCount
, AutoMapper would match to the Count()
extension method and the LINQ provider would translate the count into a correlated subquery to aggregate child records.
AutoMapper can also support complex aggregations and nested restrictions, if the LINQ provider supports it:
cfg.CreateMap<Course, CourseModel>()
.ForMember(m => m.EnrollmentsStartingWithA,
opt => opt.MapFrom(c => c.Enrollments.Where(e => e.Student.LastName.StartsWith("A")).Count()));
This query returns the total number of students, for each course, whose last name starts with the letter ‘A’.
Parameterization¶
Occasionally, projections need runtime parameters for their values. Consider a projection that needs to pull in the current username as part of its data. Instead of using post-mapping code, we can parameterize our MapFrom configuration:
string currentUserName = null;
cfg.CreateMap<Course, CourseModel>()
.ForMember(m => m.CurrentUserName, opt => opt.MapFrom(src => currentUserName));
When we project, we’ll substitute our parameter at runtime:
dbContext.Courses.ProjectTo<CourseModel>(Config, new { currentUserName = Request.User.Name });
This works by capturing the name of the closure’s field name in the original expression, then using an anonymous object/dictionary to apply the value to the parameter value before the query is sent to the query provider.
You may also use a dictionary to build the projection values:
dbContext.Courses.ProjectTo<CourseModel>(Config, new Dictionary<string, object> { {"currentUserName", Request.User.Name} });
However, using a dictionary will result in hard-coded values in the query instead of a parameterized query, so use with caution.
Supported mapping options¶
Not all mapping options can be supported, as the expression generated must be interpreted by a LINQ provider. Only what is supported by LINQ providers is supported by AutoMapper:
- MapFrom (Expression-based)
- ConvertUsing (Expression-based)
- Ignore
- NullSubstitute
Not supported:
- Condition
- SetMappingOrder
- UseDestinationValue
- MapFrom (Func-based)
- Before/AfterMap
- Custom resolvers
- Custom type converters
- ForPath
- Any calculated property on your domain object
Additionally, recursive or self-referencing destination types are not supported as LINQ providers do not support this. Typically hierarchical relational data models require common table expressions (CTEs) to correctly resolve a recursive join.
Configuration¶
Create a MapperConfiguration
instance and initialize configuration via the constructor:
var config = new MapperConfiguration(cfg => {
cfg.CreateMap<Foo, Bar>();
cfg.AddProfile<FooProfile>();
});
The MapperConfiguration
instance can be stored statically, in a static field or in a dependency injection container. Once created it cannot be changed/modified.
Alternatively, you can use the static Mapper instance to initialize AutoMapper:
Mapper.Initialize(cfg => {
cfg.CreateMap<Foo, Bar>();
cfg.AddProfile<FooProfile>();
});
Profile Instances¶
A good way to organize your mapping configurations is with profiles.
Create classes that inherit from Profile
and put the configuration in the constructor:
// This is the approach starting with version 5
public class OrganizationProfile : Profile
{
public OrganizationProfile()
{
CreateMap<Foo, FooDto>();
// Use CreateMap... Etc.. here (Profile methods are the same as configuration methods)
}
}
// How it was done in 4.x - as of 5.0 this is obsolete:
// public class OrganizationProfile : Profile
// {
// protected override void Configure()
// {
// CreateMap<Foo, FooDto>();
// }
// }
In earlier versions the Configure
method was used instead of a constructor.
As of version 5, Configure()
is obsolete. It will be removed in 6.0.
Configuration inside a profile only applies to maps inside the profile. Configuration applied to the root configuration applies to all maps created.
Assembly Scanning for auto configuration¶
Profiles can be added to the main mapper configuration in a number of ways, either directly:
cfg.AddProfile<OrganizationProfile>();
cfg.AddProfile(new OrganizationProfile());
or by automatically scanning for profiles:
// Scan for all profiles in an assembly
// ... using instance approach:
var config = new MapperConfiguration(cfg => {
cfg.AddMaps(myAssembly);
});
// ... or static approach:
Mapper.Initialize(cfg => cfg.AddProfiles(myAssembly));
// Can also use assembly names:
Mapper.Initialize(cfg =>
cfg.AddMaps(new [] {
"Foo.UI",
"Foo.Core"
});
);
// Or marker types for assemblies:
Mapper.Initialize(cfg =>
cfg.AddMaps(new [] {
typeof(HomeController),
typeof(Entity)
});
);
AutoMapper will scan the designated assemblies for classes inheriting from Profile and add them to the configuration.
Naming Conventions¶
You can set the source and destination naming conventions
Mapper.Initialize(cfg => {
cfg.SourceMemberNamingConvention = new LowerUnderscoreNamingConvention();
cfg.DestinationMemberNamingConvention = new PascalCaseNamingConvention();
});
This will map the following properties to each other:
property_name -> PropertyName
You can also set this at a per profile level
public class OrganizationProfile : Profile
{
public OrganizationProfile()
{
SourceMemberNamingConvention = new LowerUnderscoreNamingConvention();
DestinationMemberNamingConvention = new PascalCaseNamingConvention();
//Put your CreateMap... Etc.. here
}
}
Replacing characters¶
You can also replace individual characters or entire words in source members during member name matching:
public class Source
{
public int Value { get; set; }
public int Ävíator { get; set; }
public int SubAirlinaFlight { get; set; }
}
public class Destination
{
public int Value { get; set; }
public int Aviator { get; set; }
public int SubAirlineFlight { get; set; }
}
We want to replace the individual characters, and perhaps translate a word:
Mapper.Initialize(c =>
{
c.ReplaceMemberName("Ä", "A");
c.ReplaceMemberName("í", "i");
c.ReplaceMemberName("Airlina", "Airline");
});
Recognizing pre/postfixes¶
Sometimes your source/destination properties will have common pre/postfixes that cause you to have to do a bunch of custom member mappings because the names don’t match up. To address this, you can recognize pre/postfixes:
public class Source {
public int frmValue { get; set; }
public int frmValue2 { get; set; }
}
public class Dest {
public int Value { get; set; }
public int Value2 { get; set; }
}
Mapper.Initialize(cfg => {
cfg.RecognizePrefixes("frm");
cfg.CreateMap<Source, Dest>();
});
Mapper.AssertConfigurationIsValid();
By default AutoMapper recognizes the prefix “Get”, if you need to clear the prefix:
Mapper.Initialize(cfg => {
cfg.ClearPrefixes();
cfg.RecognizePrefixes("tmp");
});
Global property/field filtering¶
By default, AutoMapper tries to map every public property/field. You can filter out properties/fields with the property/field filters:
Mapper.Initialize(cfg =>
{
// don't map any fields
cfg.ShouldMapField = fi => false;
// map properties with a public or private getter
cfg.ShouldMapProperty = pi =>
pi.GetMethod != null && (pi.GetMethod.IsPublic || pi.GetMethod.IsPrivate);
});
Configuring visibility¶
By default, AutoMapper only recognizes public members. It can map to private setters, but will skip internal/private methods and properties if the entire property is private/internal. To instruct AutoMapper to recognize members with other visibilities, override the default filters ShouldMapField and/or ShouldMapProperty :
Mapper.Initialize(cfg =>
{
// map properties with public or internal getters
cfg.ShouldMapProperty = p => p.GetMethod.IsPublic || p.GetMethod.IsAssembly;
cfg.CreateMap<Source, Destination>();
});
Map configurations will now recognize internal/private members.
Configuration compilation¶
Because expression compilation can be a bit resource intensive, AutoMapper lazily compiles the type map plans on first map. However, this behavior is not always desirable, so you can tell AutoMapper to compile its mappings directly:
Mapper.Initialize(cfg => {});
Mapper.Configuration.CompileMappings();
For a few hundred mappings, this may take a couple of seconds.
Resetting static mapping configuration¶
The static Mapper.Initialize
is intended to be called only once. To reset the static mapping configuration (for example, at the start of tests):
Mapper.Reset();
Mapper.Initialize(cfg => { /* configure again */ });
Reset
should not be used in production code. It is intended to support testing scenarios only.
Attribute Mapping¶
In addition to fluent configuration is the ability to declare and configure maps via attributes. Attribute maps can supplement or replace fluent mapping configuration.
Type Map configuration¶
In order to search for maps to configure, use the AddMaps
method:
var configuration = new MapperConfiguration(cfg => cfg.AddMaps("MyAssembly"));
var mapper = new Mapper(configuration);
AddMaps
looks for fluent map configuration (Profile
classes) and attribute-based mappings.
To declare an attribute map, decorate your destination type with the AutoMapAttribute
:
[AutoMap(typeof(Order))]
public class OrderDto {
// destination members
This is equivalent to a CreateMap<Order, OrderDto>()
configuration.
Customizing type map configuration¶
To customize the overall type map configuration, you can set the following properties on the AutoMapAttribute
:
- ReverseMap (bool)
- ConstructUsingServiceLocator (bool)
- MaxDepth (int)
- PreserveReferences (bool)
- DisableCtorValidation (bool)
- IncludeAllDerived (bool)
- TypeConverter (Type)
These all correspond to the similar fluent mapping configuration options. Only the sourceType
value is required to map.
Member configuration¶
For attribute-based maps, you can decorate individual members with additional configuration. Because attributes have limitations in C# (no expressions, for example), the configuration options available are a bit limited.
Member-based attributes are declared in the AutoMapper.Configuration.Annotations
namespace.
If the attribute-based configuration is not available or will not work, you can combine both attribute and profile-based maps (though this may be confusing).
Ignoring members¶
Use the IgnoreAttribute
to ignore an individual destination member from mapping and/or validation:
using AutoMapper.Configuration.Annotations;
[AutoMap(typeof(Order))]
public class OrderDto {
[Ignore]
public decimal Total { get; set; }
Redirecting to a different source member¶
It is not possible to use MapFrom
with an expression in an attribute, but SourceMemberAttribute
can redirect to a separate named member:
using AutoMapper.Configuration.Annotations;
[AutoMap(typeof(Order))]
public class OrderDto {
[SourceMember("OrderTotal")]
public decimal Total { get; set; }
Or use the nameof
operator:
using AutoMapper.Configuration.Annotations;
[AutoMap(typeof(Order))]
public class OrderDto {
[SourceMember(nameof(Order.OrderTotal))]
public decimal Total { get; set; }
You cannot flatten with this attribute, only redirect source type members (i.e. no “Order.Customer.Office.Name” in the name). Configuring flattening is only available with the fluent configuration.
Additional configuration options¶
Additional attribute-based configuration options include:
MapAtRuntimeAttribute
MappingOrderAttribute
NullSubstituteAttribute
UseExistingValueAttribute
ValueConverterAttribute
ValueResolverAttribute
Each corresponds to the same fluent configuration mapping option.
Construction¶
AutoMapper can map to destination constructors based on source members:
public class Source {
public int Value { get; set; }
}
public class SourceDto {
public SourceDto(int value) {
_value = value;
}
private int _value;
public int Value {
get { return _value; }
}
}
Mapper.Initialize(cfg => cfg.CreateMap<Source, SourceDto>());
If the destination constructor parameter names don’t match, you can modify them at config time:
public class Source {
public int Value { get; set; }
}
public class SourceDto {
public SourceDto(int valueParamSomeOtherName) {
_value = valueParamSomeOtherName;
}
private int _value;
public int Value {
get { return _value; }
}
}
Mapper.Initialize(cfg =>
cfg.CreateMap<Source, SourceDto>()
.ForCtorParam("valueParamSomeOtherName", opt => opt.MapFrom(src => src.Value))
);
This works for both LINQ projections and in-memory mapping.
You can also disable constructor mapping:
Mapper.Initialize(cfg => cfg.DisableConstructorMapping());
You can configure which constructors are considered for the destination object:
// don't map private constructors
Mapper.Initialize(cfg => cfg.ShouldUseConstructor = ci => !ci.IsPrivate);
Conditional Mapping¶
AutoMapper allows you to add conditions to properties that must be met before that property will be mapped.
This can be used in situations like the following where we are trying to map from an int to an unsigned int.
class Foo{
public int baz;
}
class Bar {
public uint baz;
}
In the following mapping the property baz will only be mapped if it is greater than or equal to 0 in the source object.
Mapper.Initialize(cfg => {
cfg.CreateMap<Foo,Bar>()
.ForMember(dest => dest.baz, opt => opt.Condition(src => (src.baz >= 0)));
});
Preconditions¶
Similarly, there is a PreCondition method. The difference is that it runs sooner in the mapping process, before the source value is resolved (think MapFrom). So the precondition is called, then we decide which will be the source of the mapping (resolving), then the condition is called and finally the destination value is assigned.
You can see the steps yourself.
See here for a concrete example.
Open Generics¶
AutoMapper can support an open generic type map. Create a map for the open generic types:
public class Source<T> {
public T Value { get; set; }
}
public class Destination<T> {
public T Value { get; set; }
}
// Create the mapping
Mapper.Initialize(cfg => cfg.CreateMap(typeof(Source<>), typeof(Destination<>)));
You don’t need to create maps for closed generic types. AutoMapper will apply any configuration from the open generic mapping to the closed mapping at runtime:
var source = new Source<int> { Value = 10 };
var dest = mapper.Map<Source<int>, Destination<int>>(source);
dest.Value.ShouldEqual(10);
Because C# only allows closed generic type parameters, you have to use the System.Type version of CreateMap to create your open generic type maps. From there, you can use all of the mapping configuration available and the open generic configuration will be applied to the closed type map at runtime.
AutoMapper will skip open generic type maps during configuration validation, since you can still create closed types that don’t convert, such as Source<Foo> -> Destination<Bar>
where there is no conversion from Foo to Bar.
You can also create an open generic type converter:
Mapper.Initialize(cfg =>
cfg.CreateMap(typeof(Source<>), typeof(Destination<>)).ConvertUsing(typeof(Converter<>)));
AutoMapper also supports open generic type converters with any number of generic arguments:
Mapper.Initialize(cfg =>
cfg.CreateMap(typeof(Source<>), typeof(Destination<>)).ConvertUsing(typeof(Converter<,>)));
The closed type from Source
will be the first generic argument, and the closed type of Destination
will be the second argument to close Converter<,>
.
Dynamic and ExpandoObject Mapping¶
AutoMapper can map to/from dynamic objects without any explicit configuration:
public class Foo {
public int Bar { get; set; }
public int Baz { get; set; }
}
dynamic foo = new MyDynamicObject();
foo.Bar = 5;
foo.Baz = 6;
Mapper.Initialize(cfg => {});
var result = Mapper.Map<Foo>(foo);
result.Bar.ShouldEqual(5);
result.Baz.ShouldEqual(6);
dynamic foo2 = Mapper.Map<MyDynamicObject>(result);
foo2.Bar.ShouldEqual(5);
foo2.Baz.ShouldEqual(6);
Similarly you can map straight from Dictionary<string, object> to objects, AutoMapper will line up the keys with property names.
Expression Translation (UseAsDataSource)¶
Automapper supports translating Expressions from one object to another in a separate package. This is done by substituting the properties from the source class to what they map to in the destination class.
Given the example classes:
public class OrderLine
{
public int Id { get; set; }
public int OrderId { get; set; }
public Item Item { get; set; }
public decimal Quantity { get; set; }
}
public class Item
{
public int Id { get; set; }
public string Name { get; set; }
}
public class OrderLineDTO
{
public int Id { get; set; }
public int OrderId { get; set; }
public string Item { get; set; }
public decimal Quantity { get; set; }
}
Mapper.Initialize(cfg =>
{
cfg.CreateMap<OrderLine, OrderLineDTO>()
.ForMember(dto => dto.Item, conf => conf.MapFrom(ol => ol.Item.Name));
cfg.CreateMap<OrderLineDTO, OrderLine>()
.ForMember(ol => ol.Item, conf => conf.MapFrom(dto => dto));
cfg.CreateMap<OrderLineDTO, Item>()
.ForMember(i => i.Name, conf => conf.MapFrom(dto => dto.Item));
});
When mapping from DTO Expression
Expression<Func<OrderLineDTO, bool>> dtoExpression = dto=> dto.Item.StartsWith("A");
var expression = Mapper.Map<Expression<Func<OrderLine, bool>>>(dtoExpression);
Expression will be translated to ol => ol.Item.Name.StartsWith("A")
Automapper knows dto.Item
is mapped to ol.Item.Name
so it substituted it for the expression.
Expression translation can work on expressions of collections as well.
Expression<Func<IQueryable<OrderLineDTO>,IQueryable<OrderLineDTO>>> dtoExpression = dtos => dtos.Where(dto => dto.Quantity > 5).OrderBy(dto => dto.Quantity);
var expression = Mapper.Map<Expression<Func<IQueryable<OrderLine>,IQueryable<OrderLine>>>(dtoExpression);
Resulting in ols => ols.Where(ol => ol.Quantity > 5).OrderBy(ol => ol.Quantity)
Supported Mapping options¶
Much like how Queryable Extensions can only support certain things that the LINQ providers support, expression translation follows the same rules as what it can and can’t support.
UseAsDataSource¶
Mapping expressions to one another is a tedious and produces long ugly code.
UseAsDataSource().For<DTO>()
makes this translation clean by not having to explicitly map expressions.
It also calls ProjectTo<TDO>()
for you as well, where applicable.
Using EntityFramework as an example
dataContext.OrderLines.UseAsDataSource().For<OrderLineDTO>().Where(dto => dto.Name.StartsWith("A"))
Does the equivalent of
dataContext.OrderLines.Where(ol => ol.Item.Name.StartsWith("A")).ProjectTo<OrderLineDTO>()
When ProjectTo() is not called¶
Expression Translation works for all kinds of functions, including Select
calls. If Select
is used after UseAsDataSource()
and changes return type, then ProjectTo<>()
won’t be called and value with be returned instead using Mapper.Map
.
Example:
dataContext.OrderLines.UseAsDataSource().For<OrderLineDTO>().Select(dto => dto.Name)
Does the equivalent of
dataContext.OrderLines.Select(ol => ol.Item.Name)
Register a callback, for when an UseAsDataSource() query is enumerated¶
Sometimes, you may want to edit the collection, that is returned from a mapped query before forwarding it to the next application layer.
With .ProjectTo<TDto>
this is quite simple, as there is no sense in directly returning the resulting IQueryable<TDto>
because you cannot edit it anymore anyways. So you will most likely do this:
Mapper.Initialize(cfg =>
cfg.CreateMap<OrderLine, OrderLineDTO>()
.ForMember(dto => dto.Item, conf => conf.MapFrom(ol => ol.Item.Name)));
public List<OrderLineDTO> GetLinesForOrder(int orderId)
{
using (var context = new orderEntities())
{
var dtos = context.OrderLines.Where(ol => ol.OrderId == orderId)
.ProjectTo<OrderLineDTO>().ToList();
foreach(var dto in dtos)
{
// edit some property, or load additional data from the database and augment the dtos
}
return dtos;
}
}
However, if you did this with the .UseAsDataSource()
approach, you would loose all its power - namely its ability to modify the internal expression until it is enumerated.
To solve that problem, we introduced the .OnEnumerated
callback.
Using it, you can do the following:
Mapper.Initialize(cfg =>
cfg.CreateMap<OrderLine, OrderLineDTO>()
.ForMember(dto => dto.Item, conf => conf.MapFrom(ol => ol.Item.Name)));
public IQueryable<OrderLineDTO> GetLinesForOrder(int orderId)
{
using (var context = new orderEntities())
{
return context.OrderLines.Where(ol => ol.OrderId == orderId)
.UseAsDataSource()
.For<OrderLineDTO>()
.OnEnumerated((dtos) =>
{
foreach(var dto in dtosCast<OrderLineDTO>())
{
// edit some property, or load additional data from the database and augment the dtos
}
}
}
}
this OnEnumerated(IEnumerable)
callback is executed, when the IQueryable<OrderLineDTO>
itself is enumerated.
So this also works with the OData samples mentioned above: The OData $filter and $orderby expressions are still converted into SQL, and the OnEnumerated()
callback is provided with the filtered, ordered resultset from the database.
Examples¶
The source code contains unit tests for all of the features listed above. Use the GitHub search to find relevant examples.
Housekeeping¶
The latest builds can be found at NuGet
The dev builds can be found at MyGet
The discussion group is hosted on Google Groups