I am adding a httpclient as a service (within ConfigureServices(IServiceCollection services)) as follows:
services.AddHttpClient<RequestsClient>(c =>
{
c.DefaultRequestHeaders.Add("User-Agent", "HttpClientFactory");
})
.ConfigurePrimaryHttpMessageHandler(() =>
{
return new HttpClientHandler()
{
UseDefaultCredentials = true
};
});
the RequestsClient class is constructed as:
public RequestsClient (HttpClient client, IHttpContextAccessor hca, ILogger<RequestsClient > log, Configuration config)
to use the RequestsClient service in any class/component needing it i'm injecting it as:
[Inject]
protected RequestsClient requestsClient { get; set; }
all this works great.
I'm now in need of creating a second service, lets call it "TimedService".
How can I use my RequestsClient service from within my second service, TimedService?
injecting it like i do with components won't work as the RequestsClient always is null.
is there a way to give TimedService service access to my RequestsClient service?
I'm sorry if this is a stupid question, I'm fairly new to this
You need to use constructor injection in TimedService.
public class TimedService
{
private RequestsClient _requestClient;
public TimedService(RequestsClient client)
{
_requestClient = client;
}
}
I would recommend to go into typed client version. It is a bit more clearer in separation of code wise. See the official docs how to do this.
I have some class libraries that I use in my ASP.NET Web API app that handle all my backend stuff e.g. CRUD operations to multiple databases like Azure SQL Database, Cosmos DB, etc.
I don't want to re-invent the wheel and able to use them in a new Azure Functions that I'm creating in Visual Studio 2017. All my repository methods use an interface. So, how will I implement dependency injection in my new Azure function?
I'm not seeing any support for DI but I'm a bit confused. It appears Azure Functions are based on the same SDK as WebJobs and I think last year Microsoft had started supporting DI in WebJobs - I know for sure because I implemented it using Ninject.
Is there way around this so that I can use my existing libraries in my new Azure Functions project?
I see these two techniques in addition to the service locator (anti)pattern. I asked the Azure Functions team for their comments as well.
https://blog.wille-zone.de/post/azure-functions-dependency-injection/
https://blog.wille-zone.de/post/azure-functions-proper-dependency-injection/
There is an open feature request on the GitHub pages for Azure Functions concerning this matter.
However, the way I'm approaching this is using some kind of 'wrapper' entry point, resolve this using the service locator and and start the function from there.
This looks a bit like this (simplified)
var builder = new ContainerBuilder();
//register my types
var container = builder.Build();
using(var scope = container.BeginLifetimeScope())
{
var functionLogic = scope.Resolve<IMyFunctionLogic>();
functionLogic.Execute();
}
This is a bit hacky of course, but it's the best there is until there is at the moment (to my knowledge).
I've seen the willie-zone blog mentioned a lot when it comes to this topic, but you don't need to go that route to use DI with Azure functions.
If you are using Version2 you can make your Azure functions non-static. Then you can add a public constructor for injecting your dependencies. The next step is to add an IWebJobsStartup class. In your startup class you will be able to register your services like you would for any other .Net Core project.
I have a public repo that is using this approach here: https://github.com/jedi91/MovieSearch/tree/master/MovieSearch
Here is a direct link to the startup class: https://github.com/jedi91/MovieSearch/blob/master/MovieSearch/Startup.cs
And here is the function: https://github.com/jedi91/MovieSearch/blob/master/MovieSearch/Functions/Search.cs
Hope this approach helps. If you are wanting to keep your Azure Functions static then the willie-zone approach should work, but I really like this approach and it doesn't require any third party libraries.
One thing to note is the Directory.Build.target file. This file will copy your extensions over in the host file so that DI will work once the function is deployed to Azure. Running the function locally does not require this file.
Azure Functions Depdendency Injection was announced at MSBuild 2019. Here's an example on how to do it:
[assembly: FunctionsStartup(typeof(MyNamespace.Startup))]
namespace MyNamespace
{
public class Startup : FunctionsStartup
{
public override void Configure(IFunctionsHostBuilder builder)
{
builder.Services.AddHttpClient();
builder.Services.AddSingleton((s) => {
return new CosmosClient(Environment.GetEnvironmentVariable("COSMOSDB_CONNECTIONSTRING"));
});
builder.Services.AddSingleton<ILoggerProvider, MyLoggerProvider>();
}
}
}
GitHub Example
Documentation
As stated above, it was just announced at Build 2019. It can now be setup almost exactly like you would in an ASP .Net Core app.
Microsoft Documentation
Short Blog I Wrote
Actually there is a much nicer and simpler way provided out of the box by Microsoft. It is a bit hard to find though. You simply create a start up class and add all required services here, and then you can use constructor injection like in regular web apps and web apis.
This is all you need to do.
First I create my start up class, I call mine Startup.cs to be consistent with Razor web apps, although this is for Azure Functions, but still it's the Microsoft way.
using System;
using com.paypal;
using dk.commentor.bl.command;
using dk.commentor.logger;
using dk.commentor.sl;
using Microsoft.Azure.WebJobs;
using Microsoft.Azure.WebJobs.Hosting;
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Logging;
using org.openerp;
[assembly:Microsoft.Azure.WebJobs.Hosting.WebJobsStartup(typeof(dk.commentor.starterproject.api.Startup))]
namespace dk.commentor.starterproject.api
{
public class Startup : IWebJobsStartup
{
public void Configure(IWebJobsBuilder builder)
{
builder.Services.AddSingleton<ILogger, CommentorLogger>();
builder.Services.AddSingleton<IPaymentService, PayPalService>();
builder.Services.AddSingleton<IOrderService, OpenERPService>();
builder.Services.AddSingleton<ProcessOrderCommand>();
Console.WriteLine("Host started!");
}
}
}
Next I change the method call in the function from static to non-static, and I add a constructor to the class (which is now also non-static). In this constructor I simply add the services I require as constructor parameters.
using System;
using System.IO;
using System.Threading.Tasks;
using Microsoft.AspNetCore.Mvc;
using Microsoft.Azure.WebJobs;
using Microsoft.Azure.WebJobs.Extensions.Http;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Logging;
using Newtonsoft.Json;
using dk.commentor.bl.command;
namespace dk.commentor.starterproject.api
{
public class ProcessOrder
{
private ProcessOrderCommand processOrderCommand;
public ProcessOrder(ProcessOrderCommand processOrderCommand) {
this.processOrderCommand = processOrderCommand;
}
[FunctionName("ProcessOrder")]
public async Task<IActionResult> Run([HttpTrigger(AuthorizationLevel.Function, "post", Route = null)] HttpRequest req, ILogger log)
{
log.LogInformation("C# HTTP trigger ProcessOrder called!");
log.LogInformation(System.Environment.StackTrace);
string jsonRequestData = await new StreamReader(req.Body).ReadToEndAsync();
dynamic requestData = JsonConvert.DeserializeObject(jsonRequestData);
if(requestData?.orderId != null)
return (ActionResult)new OkObjectResult($"Processing order with id {requestData.orderId}");
else
return new BadRequestObjectResult("Please pass an orderId in the request body");
}
}
}
Hopes this helps.
I would like to add my 2 cents to it. I used the technique that it's used by Host injecting ILogger. If you look at the Startup project I created GenericBindingProvider that implements IBindingProvider. Then for each type I want to be injected I register it as follow:
builder.Services.AddTransient<IWelcomeService, WelcomeService>();
builder.Services.AddSingleton<IBindingProvider, GenericBindingProvider<IWelcomeService>>();
The downside is that you need to register the type you want to be injected into the function twice.
Sample code:
Azure Functions V2 Dependency Injection sample
I have been using SimpleInjector perfectly fine in Azure Functions. Just create a class (let's call it IoCConfig) that has the registrations and make a static instance of that class in function class so that each instance will use the existing instance.
public interface IIoCConfig
{
T GetInstance<T>() where T : class;
}
public class IoCConfig : IIoCConfig
{
internal Container Container;
public IoCConfig(ExecutionContext executionContext, ILogger logger)
{
var configurationRoot = new ConfigurationBuilder()
.SetBasePath(executionContext.FunctionAppDirectory)
.AddJsonFile("local.settings.json", optional: true, reloadOnChange: true)
.AddEnvironmentVariables()
.Build();
Container = new Container();
Configure(configurationRoot, logger);
}
public IoCConfig(IConfigurationRoot configurationRoot, ILogger logger)
{
Container = new Container();
Configure(configurationRoot, logger);
}
private void Configure(IConfigurationRoot configurationRoot, ILogger logger)
{
Container.RegisterInstance(typeof(IConfigurationRoot), configurationRoot);
Container.Register<ISomeType, SomeType>();
}
public T GetInstance<T>() where T : class
{
return Container.GetInstance<T>();
}
}
Then in root:
public static class SomeFunction
{
public static IIoCConfig IoCConfig;
[FunctionName("SomeFunction")]
public static async Task Run(
[ServiceBusTrigger("some-topic", "%SUBSCRIPTION_NAME%", Connection = "AZURE_SERVICEBUS_CONNECTIONSTRING")]
SomeEvent msg,
ILogger log,
ExecutionContext executionContext)
{
Ensure.That(msg).IsNotNull();
if (IoCConfig == null)
{
IoCConfig = new IoCConfig(executionContext, log);
}
var someType = IoCConfig.GetInstance<ISomeType>();
await someType.Handle(msg);
}
}
AzureFunctions.Autofac is very easy to use.
Just add a config file:
public class DIConfig
{
public DIConfig(string functionName)
{
DependencyInjection.Initialize(builder =>
{
builder.RegisterType<Sample>().As<ISample>();
...
}, functionName);
}
}
Add the DependencyInjectionConfig attribute then inject:
[DependencyInjectionConfig(typeof(DIConfig))]
public class MyFunction
{
[FunctionName("MyFunction")]
public static HttpResponseMessage Run([HttpTrigger(AuthorizationLevel.Function, "get", Route = null)]HttpRequestMessage request,
TraceWriter log,
[Inject]ISample sample)
{
https://github.com/introtocomputerscience/azure-function-autofac-dependency-injection
I think this is a better solution:
https://github.com/junalmeida/autofac-azurefunctions
https://www.nuget.org/packages/Autofac.Extensions.DependencyInjection.AzureFunctions
Install the NuGet in your project and then make a Startup.cs and put this in it:
[assembly: FunctionsStartup(typeof(Startup))]
public class Startup
{
public override void Configure(IFunctionsHostBuilder builder)
{
builder
.UseAppSettings() // this is optional, this will bind IConfiguration in the container.
.UseAutofacServiceProviderFactory(ConfigureContainer);
}
private void ConfigureContainer(ContainerBuilder builder)
{
// do DI registration against Autofac like normal! (builder is just the normal ContainerBuilder from Autofac)
}
...
Then in your function code you can do normal constructor injection via DI:
public class Function1 : Disposable
{
public Function1(IService1 service1, ILogger logger)
{
// logger and service1 injected via autofac like normal
// ...
}
[FunctionName(nameof(Function1))]
public async Task Run([QueueTrigger("myqueue-items", Connection = "AzureWebJobsStorage")]string myQueueItem)
{
//...
Support for Dependency injection begins with Azure Functions 2.x which means Dependency Injection in Azure function can now leverage .NET Core Dependency Injection features.
Before you can use dependency injection, you must install the following NuGet packages:
Microsoft.Azure.Functions.Extensions
Microsoft.NET.Sdk.Functions
Having Dependency Injection eases things like DBContext, Http client usage (Httpclienfactory), Iloggerfactory, cache support etc.
Firstly, update the Startup class as shown below
namespace DemoApp
{
public class Startup: FunctionsStartup
{
public override void Configure(IFunctionsHostBuilder builder)
{
builder.Services.AddScoped<IHelloWorld, HelloWorld>();
// Registering Serilog provider
var logger = new LoggerConfiguration()
.WriteTo.Console()
.CreateLogger();
builder.Services.AddLogging(lb => lb.AddSerilog(logger));
//Reading configuration section can be added here etc.
}
}
}
Secondly, Removal of Static keyword in Function class and method level
public class DemoFunction
{
private readonly IHelloWorld _helloWorld;
public DemoFunction(IHelloWorld helloWorld)
{
_helloWorld = helloWorld;
}
[FunctionName("HttpDemoFunction")]
public async Task<IActionResult> Run(
[HttpTrigger(AuthorizationLevel.Anonymous, "get", "post", Route = null)] HttpRequest req,
ILogger log)
{
log.LogInformation("C# HTTP trigger function processed a request.");
}
If we look into above e.g. IHelloWorld is injected using .NET Core DI
**Note:**In-spite of having latest version of Azure function v3 for Dependency Injection to enable few steps are manual as shown above
Sample code on github can be found here
I've got REST FeignClient defined in my application:
#FeignClient(name = "gateway", configuration = FeignAuthConfig.class)
public interface AccountsClient extends Accounts {
}
I share endpoint interface between server and client:
#RequestMapping(API_PATH)
public interface Accounts {
#PostMapping(path = "/register",
produces = APPLICATION_JSON_VALUE,
consumes = APPLICATION_JSON_VALUE)
ResponseEntity<?> registerAccount(#RequestBody ManagedPassUserVM managedUserDTO)
throws EmailAlreadyInUseException, UsernameAlreadyInUseException, URISyntaxException;
}
Everythng works fine except that my FeignClient definition in my client application also got registered as independent REST endpoint.
At the moment I try to prevent this behavior using filter which returns 404 status code for FeignClinet client mappings in my client application. However this workeraund seems very inelegant.
Is there another way how to prevent feign clients registering as separate REST endpoints?
It is a known limitation of Spring Cloud's feign support. By adding #RequestMapping to the interface, Spring MVC (not Spring Cloud) assumes you want as an endpoint. #RequestMapping on Feign interfaces is not currently supported.
I've used workaround for this faulty Spring Framework behavior:
#Configuration
#ConditionalOnClass({Feign.class})
public class FeignMappingDefaultConfiguration {
#Bean
public WebMvcRegistrations feignWebRegistrations() {
return new WebMvcRegistrationsAdapter() {
#Override
public RequestMappingHandlerMapping getRequestMappingHandlerMapping() {
return new FeignFilterRequestMappingHandlerMapping();
}
};
}
private static class FeignFilterRequestMappingHandlerMapping extends RequestMappingHandlerMapping {
#Override
protected boolean isHandler(Class<?> beanType) {
return super.isHandler(beanType) && (AnnotationUtils.findAnnotation(beanType, FeignClient.class) == null);
}
}
}
I found it in SpringCloud issue
We are using classes inheriting from Registry to configure our StructureMap container in our ASP.NET MVC 4 application startup.
Some excerpt from one of the registry-classes:
For<ISomeInterface>().HybridHttpOrThreadLocalScoped().Use<SomeImplementation>();
We would like use different instances of our interfaces depending on the context. (For example switching from database "online" mode to "maintenance" mode where everything is saved on filesystem; therefore using other interfaces (i.e. repositories) all over the place in our application)
For example by default it should use SomeImplementation but when passing some kind of querystring in the url (to name a simple "context" scenario) it should use SomeOtherImplementation.
How can this be achieved for multiple interfaces/types?
Should we use named instances for this? Like:
For<ISomeInterface>().HybridHttpOrThreadLocalScoped().Use<SomeOtherImplementation>().Named("other");
I read about StructureMap Profiles but i'm not sure if this is the right way to go.
Should we use profiles for this? Like i.e.:
Profile("other", profileExpression =>
{
For<ISomeInterface>().HybridHttpOrThreadLocalScoped().Use<SomeOtherImplementation>();
});
How can we switch different configurations on the fly?
ObjectFactory.Container.SetDefaultsToProfile("other");
This way? (At what stage in mvc "life-cycle" this can happen at the earliest?)
Can this be a temporary switch for just the current request or current users session?
Thanks in advance!
From my experience, runtime configuration like this is best achieved using an abstract factory that is responsible for creating your dependency during runtime.
This dependency can then be registered with StructureMap like so:
Your registry:
public class StorageRegistry : Registry
{
public StorageRegistry()
{
...
this.For<IDataStoreInstance>().Use(ctx => ctx.GetInstance<DataStoreAbstractFactory>().ConfigureStorage());
...
}
}
Now your DataStoreAbstractFactory is responsible for creating and configure the necessary storage instance based on your configuration. As DataStoreAbstractFactory is now registered with StructureMap you're able to inject the necessary dependencies into it for determining which storage method to use.
Implementation example:
public class DataStoreAbstractFactory
{
public DataStoreAbstractFactory()
{
// Dependencies to figure out data storage method can be injected here.
}
public IDataStoreInstance ConfigureStorage()
{
// This method can be used to return type of storage based on your configuration (ie: online or maintenance)
}
}
public interface IDataStoreInstance
{
void Save();
}
public class DatabaseStorage : IDataStoreInstance
{
public void Save()
{
// Implementation details of persisting data in a database
}
}
public class FileStorage : IDataStoreInstance
{
public void Save()
{
// Implementation details of persisting data in a file system
}
}
Usage:
Your controller/services or whatever are now completely unaware of what storage method they're using when accessing and persisting data.
public class UpdateController : Controller
{
public IDataStoreInstance StorageInstance { get; set; }
public UpdateController(IDataStoreInstance storageInstance)
{
StorageInstance = storageInstance;
}
[HttpPost]
public ActionResult Index()
{
...
this.StorageInstance.Save();
...
}
...
}
I am looking to integrate NServiceBus into an existing ServiceStack web host. ServiceStack is currently using the built in Funq IoC container. NServiceBus has been configured (elsewhere in the system) to use Unity for IoC.
ServiceStack has a feature whereby IRequestContext will be automatically injected when it finds the IRequiresRequestContext interface on a class. NServiceBus has a similar feature for Message Mutators, via the IMutateOutgoingTransportMessages interface.
The application is a multi-tenant application. Single application, which via an API Key, passes an account code through to the NServiceBus handler (indirectly via a provider that is called on the construction of the handler's constructor injection using Unity).
My problem arises in ServiceStack. I am using a request filter to drag the API key out of the request headers, which I look-up in a database, and then finally write into the IHttpRequest.Items collection:
appHost.RequestFilters.Add((req, res, requestDto) =>
{
var tenant = tenantRepository.GetByApiKey(
req.Items.Add("AccountCode", tenant.AccountCode);
}
I then have an NServiceBus transport message mutator, that implements that IRequiresRequestContext interface, and this class is located in the same assembly as the ServiceStack services registered in the AppHost:
public class MessageHeaderMutator : IMutateOutgoingTransportMessages, INeedInitialization, IRequiresRequestContext
{
#region IRequiresRequestContext Members
public IRequestContext RequestContext { get; set; }
#endregion
#region IMutateOutgoingTransportMessages Members
public void MutateOutgoing(object[] messages, NServiceBus.TransportMessage transportMessage)
{
transportMessage.Headers.Add("AccountCode", RequestContext.Get<IHttpRequest>().Items["AccountCode"].ToString());
}
#endregion
#region INeedInitialization Members
public void Init()
{
Configure.Instance.Configurer.ConfigureComponent<MessageHeaderMutator>(DependencyLifecycle.InstancePerCall);
}
#endregion
}
However, RequestContext is never injected, and is always null. My theory is that the two interface injections, injected via the two separate frameworks, are somehow clashing.
I have a workaround, which is to use the ServiceStack HostContext.Items instead, as per this discussion, but I am concerned that the HostContext is not a per request collection, so I might end up writing data to the wrong tenant. Workaround is:
// app host
appHost.RequestFilters.Add((req, res, requestDto) =>
{
var accountCode = tenantRepository.GetByApiKey(
HostContext.Instance.Items.Add("AccountCode", client.AccountCode);
}
// message mutator
public class MessageHeaderMutator : IMutateOutgoingTransportMessages, INeedInitialization
{
#region IMutateOutgoingTransportMessages Members
public void MutateOutgoing(object[] messages, NServiceBus.TransportMessage transportMessage)
{
var accountCode = HostContext.Instance.Items["AccountCode"].ToString();
transportMessage.Headers.Add("AccountCode", accountCode);
}
#endregion
#region INeedInitialization Members
public void Init()
{
Configure.Instance.Configurer.ConfigureComponent<MessageHeaderMutator>(DependencyLifecycle.InstancePerCall);
}
#endregion
}
My question is therefore twofold:
The first is, why does IRequiresRequestContext not correctly inject RequestContext into the Message mutator, or is there a way to inject RequestContext manually?
Is the use of the HostContext safe to assume it is per request?
Bonus question: Is the use of two separate IoC containers in the same project (Unity in NServiceBus and Funq in ServiceStack) a really bad idea? Would it be smarter to get ServiceStack to use the same Unity IoC container as NServiceBus?
P.S. This is NServiceBus 4 (beta at time of writing).
It is precisely because you're using 2 different containers that DI doesn't work for you out of the box for objects registered in different containers.
You don't necessarily need to standardize on a single container (though it would save you from dealing with these sorts of issues all the time).
What you can do to keep working with both containers is to tell the NServiceBus container how to resolve IRequiresRequestContext like this:
public class RequestContextBootstrapper : INeedInitialization
{
public void Init()
{
Configure.Component<IRequiresRequestContext>( /* get object from ServiceStack */ );
}
}
You can access the ServiceStack container via AppHostBase.Container and use it to resolve your objects as Udi suggested.