Strongly-typed Configuration for .NET Core - with full Dependency Injection support
C# PowerShell
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
ci
docs
examples
src
test
.appveyor.yml
.gitignore
LICENCE.md Initial commit Feb 8, 2017
README.md
Thinktecture.Configuration.sln
Visual Studio.ruleset

README.md

Build status Thinktecture.Configuration

Configuration is one of the most prominent cornerstones in software systems, and especially in distributed systems. And it has been a point for discussions in .NET for quite some time.

In one of our projects we have built a solution that lets different applications in different companies exchange data, although being behind firewalls, using the open source Relay Server. But, to our surprise, one of the features our customer was amazed about was the library I've developed to make configuration easier to handle.

In Thinktecture.Configuration I've taken over the ideas, generalized them and added new features.

The basic idea is that .NET developers should be able to deal with configuration data by just using arbitrary classes.

The library consists of 3 main components: IConfigurationLoader, IConfigurationProvider and IConfigurationSelector. The IConfigurationLoader loads data from storage, e.g. JSON from the file system. The IConfigurationProvider uses IConfigurationSelector to select the correct piece of data like a JSON property and to convert this data to requested configuration.

Architecture

In short, the features of the lib are:

  • The configuration (i.e. the type being used by the application)
    • should be type-safe
    • can be an abstract type (e.g. an interface)
    • don't have to have property setters or other kind of metadata just to be deserializable
    • can be and have properties that are virtually as complex as needed
    • can be injected via DI (dependency injection)
    • can have dependencies that are injected via DI (i.e. via constructor injection)
    • can be changed by "overrides"
  • The usage of the configuration in a developer's code base should be free of any stuff from the configuration library
  • The extension of a configuration by new properties or changing the structure of it should be a one-liner

Use Cases

Thinktecture.Configuration is able to use 2 kind of sources:

Although the latter one seems to be unnecessary it has some benefits over IConfiguration.

  • IConfiguration suffers from information loss. For examples
    • If you are using a JSON file as the source and your configuration contains an array-property then you can't differentiate between (1) the array is empty { "myArray": [] } and (2) array is missing at all { }. The same goes for other reference types, i.e. classes.
    • Furthermore, IConfiguration is missing type information, it is a collection of key-value pairs of type string
    • If one of the keys contain accidentally the character : like my:key then the configuration { "my:key": 42 } will be interpreted as { "my": { "key": 42 } }
  • For deserialization of JSON you can use very powerful library Newtonsoft JSON. For IConfiguration we have the ConfigurationBinder, which is being used by IOptions<T>, but the binder is not extensible because it is a static class.

All in all the IConfiguration is more versatile but using JSON files only is a more powerful approach.

To get a better feeling of the library read the next section about how to Use JSON Files as Data Source or jump right to Use IConfiguration as Data Source

Use JSON Files as Data Source

In this post I'm going to show the capabilities of the library by illustrating it with a few examples. In this concrete example I'am using a JSON file containing the configuration values (with Newtosonf.Json in the background) and Autofac for DI.

But the library is not limited to these. The hints are at the end of this post if you want to use a different DI framework, other storage than the file system (i.e. JSON files) or not use JSON altogether.

The first use case is a bit lengthy to explain the basics. The others will just point out specific features.

Nuget: Install-Package Thinktecture.Configuration.JsonFile.Autofac

1. One file containing one or more configurations

Shown features in this example:

  • one JSON file
  • multiple configurations
  • a configuration doesn't have to be on the root of the JSON file
  • a configuration has dependencies known by DI
  • a configuration gets injected in a component like any other dependency

We start with 2 simple configuration types.

public interface IMyConfiguration
{
    string Value { get; }
}

public interface IOtherConfiguration
{
    TimeSpan Value { get; }  
}

The configuration IMyConfiguration is required by our component MyComponent.

public class MyComponent
{
    public MyComponent(IMyConfiguration config)
    {
    }
}

Configuration file configuration.json

{
    "My":
    {
        "Config": { value: "content" }
    },
    "OtherConfig": { value: "00:00:05" }
}

Now let's setup the code in the executing assembly and configure DI to make MyComponent resolvable along with IMyConfiguration.

var builder = new ContainerBuilder();
builder.RegisterType<MyComponent>().AsSelf();

// IFile is required by JsonFileConfigurationLoader to access the file system
// For more info: https://www.nuget.org/packages/Thinktecture.IO.FileSystem.Abstractions/
builder.RegisterType<FileAdapter>().As<IFile>().SingleInstance();

// register so-called configuration provider that operates on "configuration.json"
builder.RegisterJsonFileConfigurationProvider("./configuration.json");

// register the configuration.
// "My.Config" is the (optional) path into the config JSON structure because our example configuration does not start at the root of the JSON
builder.RegisterJsonFileConfiguration<MyConfiguration>("My.Config")
    .AsImplementedInterfaces() // i.e. as IMyConfiguration
    .SingleInstance(); // The values won't change in my example

// same difference with IOtherConfiguration
builder.RegisterJsonFileConfiguration<OtherConfiguration>("OtherConfig")
    .AsImplementedInterfaces();

var container = builder.Build();

The concrete types MyConfiguration and OtherConfiguration are, as often when working with abstractions, used with DI only. Apart from that, these types won't show up at any other places. The type MyConfiguration has a dependency IFile that gets injected during deserialization.

public class MyConfiguration : IMyConfiguration
{
    public string Value { get; set; }

    public MyConfiguration(IFile file)
    {
        ...
    }
}

public class OtherConfiguration : IOtherConfiguration
{
    public TimeSpan Value { get; set; }  
}

The usage is nothing special

// IMyConfiguration gets injected into MyComponent
var component = container.Resolve<MyComponent>();

// we can resolve IMyConfiguration directly if we want to
var config = container.Resolve<IMyConfiguration>();

2. Nesting

Shown features in this use case:

  • one of the properties of a configuration type is a complex type
  • complex property type can be instantiated by Newtonsoft.Json or DI
  • complex property can be resolved directly if required

In this example IMyConfiguration has a property that is not of a simple type. The concrete types implementing IMyConfiguration and IMyClass consist of property getters and setters only thus left out for brevity.

public interface IMyConfiguration
{
    string Value { get; }
    IMyClass MyClassValue { get; }
}

public interface IMyClass
{
    int Value { get; }  
}

The JSON file looks as following:

{
    "Value": "content",
    "MyClassValue": { "Value": 42 }
}

Having a complex property we can decide whether the type IMyClass is going to be instantiated by Newtonsoft.Json or DI.

With just the following line the type IMyClass is not introduced to the configuration library and is going to be instantiated by Newtonsoft.Json.

builder.RegisterJsonFileConfiguration<MyConfiguration>()
    .AsImplementedInterfaces()
    .SingleInstance();

With the following line we introduce the type to the config lib and DI but the instance of IMyClass cannot be resolved directly.

builder.RegisterJsonFileConfigurationType<MyClass>();

Should IMyClass be resolvable directly then we can use the instance created along with IMyConfiguration or let new instance be created.

// option 1: use the property of IMyConfiguration
builder.Register(context => context.Resolve<IMyConfiguration>().MyClassValue)
    .AsImplementedInterfaces()
    .SingleInstance();

// option 2: let create a new instance 
builder.RegisterJsonFileConfiguration<MyClass>("MyClassValue")
    .AsImplementedInterfaces()
    .SingleInstance();

3. Multiple JSON files

The configurations can be loaded from more than one file.

Configuration types are

public interface IMyConfiguration
{
    string Value { get; }
}

public interface IOtherConfiguration
{
    TimeSpan Value { get; }  
}

File myConfiguration.json

{
    "Value": "content"
}

File otherConfiguration.json

{
    "Value": "00:00:05"
}

Having two files we need a means to distinguish between them when registering the configurations. In this case we use RegisterKeyedJsonFileConfigurationProvider that returns a key that will be passed to RegisterJsonFileConfiguration.

var providerKey = builder.RegisterKeyedJsonFileConfigurationProvider("myConfiguration.json");
builder.RegisterJsonFileConfiguration<MyConfiguration>(providerKey)
    .AsImplementedInterfaces()
    .SingleInstance();

var otherKey = builder.RegisterKeyedJsonFileConfigurationProvider("otherConfiguration.json");
builder.RegisterJsonFileConfiguration<OtherConfiguration>(otherKey)
    .AsImplementedInterfaces()
    .SingleInstance();

4. Overrides

A configuration can be assembled from one base configuration and one or more overrides.

In this case we have two config files. One containing the default values of the configuration and the other containing values to override.

Default values come from baseConfiguration.json

{
    "Value":
    {
        "InnerValue_1": 1,
        "InnerValue_2": 2
    }
}

InnerValue_2 will be changed by the overrides.json

{
    "Value":
    {
        "InnerValue_2": 3
    }
}

The configuration

public interface IMyConfiguration
{
    IInnerConfiguration Value { get; }  
}

public interface IInnerConfiguration
{
    int InnerValue_1 { get; }
    int InnerValue_2 { get; }
}

To specify overrides we need to provide more than one file path when registering the configuration provider. The overrides are applied in the same order they passed to RegisterJsonFileConfigurationProvider.

builder.RegisterJsonFileConfigurationProvider("baseConfiguration.json", "overrides.json");
builder.RegisterJsonFileConfiguration<MyConfiguration>()
    .AsImplementedInterfaces()
    .SingleInstance();

5. Extension of the configuration

Let's add a property to IInnerConfiguration from previous paragraph.

public interface IInnerConfiguration
{
    int InnerValue_1 { get; }
    int InnerValue_2 { get; }
    string NewValue { get; }
}

Add the corresponding property to the JSON file baseConfiguration.json

{
    "Value":
    {
        "InnerValue_1": 1,
        "InnerValue_2": 2,
        "NewValue": "content"
    }
}

That's it.

Use IConfiguration as Data Source

The IConfiguration itself is more or less just a collection of key-value pairs of type string. Using it as-it-is in a library is not recommended because (1) the depevelopers on both side (lib owner and lib consumer) have to know the keys of the configuration values and (2) how to (de)serialize the values.

Thinktecture.Extensions.Configuration is using an instance of IConfiguration as a data source to create and populate strongly-typed configuration classes.

The Library Microsoft.Extensions.Options provides similar solution but it is using a static class from Microsoft.Extensions.Configuration.Binder making it hard to extend. Thinktecture.Extensions.Configuration provides new features like

  • DI support (e.g. for Autofac use nuget package Thinktecture.Extensions.Configuration.Autofac)
  • CultureInfo support, so that localized values like 42,1 with , as decimal separator are parsed correctly to 42.1
  • Honoring of collection indexes, meaning a key-value pair "MyIntCollection:1"-42 is deserialized to new int[]{0, 42} instead of new int[]{42}
  • No polution of own code by helper classes/interfaces (IOptions<T>, IOptionsSnapshot<T>, etc) by default
    • Still with reload support
  • If a property of type int[] (ie. an array) is comming from a JSON file and is set to null, ie. JSON is looking like this { "myIntArray": null } then IOptions<T> throws an error because the provider Microsoft.Extensions.Configuration.Json generates an empty string if it sees null. This incompatibility between the provider Microsoft.Extensions.Configuration.Json and IOptions<T> is very unfortunate. Thinktecture.Extensions.Configuration sets the property to null in this case.

Registering IConfiguration as data source for Thinktecture.Configuration

Nuget: Install-Package Thinktecture.Extensions.Configuration.Autofac when using Autofac

Use the extension method RegisterMicrosoftConfigurationProvider or RegisterKeyedMicrosoftConfigurationProvider when having more than 1 instance of IConfiguration. The latter one is more of an edge case.

IConfiguration config = ...;
ContainerBuilder builder = ...; // Autofac container builder

builder.RegisterMicrosoftConfigurationProvider(config);

Registering of strongly-typed Configurations

Use the extension method RegisterMicrosoftConfiguration to register a class with the DI.

builder.RegisterMicrosoftConfiguration<MyConfiguration>()
    .As<IMyConfiguration>()     // from here on we have standard Autofac registration builder
    .InstancePerLifetimeScope();

If the type IMyConfiguration contains a property of a complex type, say IMyOtherConfiguration ...

public interface IMyConfiguration
{
	int MyValue { get; }
	IMyOtherConfiguration InnerConfiguration { get; }
}

... then it has to be registered with DI as well:

builder.RegisterMicrosoftConfigurationType<MyOtherConfiguration, IMyOtherConfiguration>();

The configuration of type IMyConfiguration can now be resolved

var container = builder.Build();
var myConfig = container.Resolve<IMyConfiguration>();

Working with different frameworks, storages and data models

Using another DI framework

To use a different DI framwork than Autofac use the package Thinktecture.Configuration.JsonFile instead of Thinktecture.Configuration.JsonFile.Autofac and implement the interface IJsonTokenConverter using your favorite DI framework. The converter has just one method TConfiguration Convert<TConfiguration>(JToken token).

Load JSON from other media

To load JToken from other storages than the file system just implement the interface IConfigurationLoader<JToken, JToken>. For example, if the JSON configuration are in a database then inject the database context or a data access layer and select corresponding data rows.

Use different data models

If you are using other data model than JSON then reference the package Thinktecture.Configuration and implement the interfaces IConfigurationLoader<TRawDataIn,TRawDataOut>, IConfigurationProvider<TRawDataIn,TRawDataOut> and IConfigurationSelector<TRawDataIn,TRawDataOut>. It sounds like much but if you look into the code of the corresponding JSON-based classes you will see that the classes are pretty small and trivial.

Some final words...

Although the configuration is an important part of the software development it is not the most exciting one. Therefore, a software developer may be inclined to take shortcuts and work with meaningful hardcoded values. Thinktecture.Configuration gives you the means to work with .NET types without thinking too much how to load and parse the values. This saves time, improves the reusability of the components and the software architecture.