SkillAgentSearch skills...

AWSLambdaSharpTemplate

A template for AWS Lambda functions written in C# with .NET Core

Install / Use

/learn @Kralizek/AWSLambdaSharpTemplate
About this skill

Quality Score

0/100

Supported Platforms

Universal

README

Build status

Project description

Write complex AWS Lambda functions in C#, faster.

This project gives you a strong template that helps you write Lambda functions using C# and .NET Core without giving up on SOLID programming principles.

Overview

This project is designed so that you can create a Lambda function like you create an ASP.NET Core web application. The body of your function becomes your Startup class where the setup and the initialization happens. The real logic of your function is nicely contained into an handler that acts as entry point.

This is what a minimal Lambda function will look like

public class Function : EventFunction<string>
{
    protected override void Configure(IConfigurationBuilder builder)
    {

    }

    protected override void ConfigureLogging(ILoggingBuilder logging, IExecutionEnvironment executionEnvironment)
    {

    }

    protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
    {
        RegisterHandler<EventHandler>(services);
    }
}

The only method required is ConfigureServices that you use to register your handler. You can use ConfigureLogging to add log providers to the provided ILoggingBuilder. Configure allows you to customize where your settings are taken from. Very much like ASP.NET Core web applications, you can use environment variables, configuration files and in-memory collections.

Type of functions and handlers

AWS Lambda supports two types of invocation, RequestResponse and Event. This distinction is mirrored by offering two distinct base classes.

RequestResponse functions

A RequestResponse function represents a function that produces a return value that will be returned to the caller of the function.

To create a RequestResponse function, simply change your class so that it inherits from RequestResponseFunction<TInput, TOutput> where TInput is a class representing the incoming data and TOutput represents the value your function will produce. Unless specific use cases, TOutput should always be your type, never Task<TOutput>.

A RequestResponse function requires an handler that implements IRequestResponseHandler<TInput, TOutput> to be registered in ConfigureServices.

This is the signature of IRequestResponseHandler<TInput, TOutput>

public interface IRequestResponseHandler<TInput, TOutput>
{
    Task<TOutput> HandleAsync(TInput input, ILambdaContext context);
}

Here you can find a sample that shows a RequestResponse function that returns the input string as upper case.

Event functions

An Event function represents a function that is not expected to produce any value.

To create an Event function, simply change your class so that it inherits from EventFunction<TInput> where TInput is a class representing the incoming data.

An Event function requires an handler that implements IEventHandler<TInput> to be registered in ConfigureServices.

This is the signature of IEventHandler<TInput>

public interface IEventHandler<TInput>
{
    Task HandleAsync(TInput input, ILambdaContext context);
}

Here you can find a sample that shows an Event function that accepts an input string and logs it into CloudWatch logs.

The library offers special support for two classes of Event functions: functions that handle SQS and SNS messages.

Event functions handling SNS notifications

AWS Lambda functions can be used to handle SNS notifications.

Since all SNS notifications have the same structure, the package Kralizek.Lambda.Template.Sns can be used to speed up the development of functions that handle SNS notifications.

To do so, create a class that implements the interface INotificationHandler<MyNotification>, then change the ConfigureServices method to look like the following snippet.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseNotificationHandler<MyNotification, MyNotificationHandler>();
}

Parallel execution of SNS notifications

Since a single SNS request can contain multiple messages, you can specify that you want all messages to be processed in parallel by changing the ConfigureServices method like in the snippet below.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseNotificationHandler<MyNotification, MyNotificationHandler>().WithParallelExecution(maxDegreeOfParallelism: 4);
}

You can use the maxDegreeOfParallelism parameter to specify the amount of parallel executions that you desire. By default, the amount of logical processors available is used.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseNotificationHandler<MyNotification, MyNotificationHandler>().WithParallelExecution();
}

Custom serialization of SNS messages

Since each SNS message contains the actual payload as encoded a string, a custom serializer can be specified to replace the default JSON serializer.

To do so, create an implementation of the interface INotificationSerializer and register it in the ConfigureServices method.

public class MyCustomSerializer : INotificationSerializer
{
  public TMessage? Deserialize<TMessage>(string input)
  {
    // implement your deserialization strategy here
  }
}
protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseNotificationHandler<MyNotification, MyNotificationHandler>();

    services.UseCustomNotificationSerializer<MyCustomSerializer>();
}

Event functions handling SQS messages

AWS Lambda functions can be used to handle SQS messages.

Since all SQS messages have the same structure, the package Kralizek.Lambda.Template.Sqs can be used to speed up the development of functions that handle SQS messages.

To do so, create a class that implements the interface IMessageHandler<MyMessage>, then change the ConfigureServices method to look like the following snippet.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseQueueMessageHandler<MyMessage, MyMessageHandler>();
}

Parallel execution of SQS messages

Since a single SQS message can contain multiple messages, you can specify that you want all messages to be processed in parallel by changing the ConfigureServices method like in the snippet below.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseQueueMessageHandler<MyMessage, MyMessageHandler>().WithParallelExecution(maxDegreeOfParallelism: 4);
}

You can use the maxDegreeOfParallelism parameter to specify the amount of parallel executions that you desire. By default, the amount of logical processors available is used.

protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseQueueMessageHandler<MyMessage, MyMessageHandler>().WithParallelExecution();
}

Custom serialization of SQS messages

Since each SQS message contains the actual payload as encoded a string, a custom serializer can be specified to replace the default JSON serializer.

To do so, create an implementation of the interface IMessageSerializer and register it in the ConfigureServices method.

public class MyCustomSerializer : IMessageSerializer
{
  public TMessage? Deserialize<TMessage>(string input)
  {
    // implement your deserialization strategy here
  }
}
protected override void ConfigureServices(IServiceCollection services, IExecutionEnvironment executionEnvironment)
{
    services.UseQueueMessageHandler<MyMessage, MyMessageHandler>()

    services.UseCustomMessageSerializer<MyCustomSerializer>();
}

Creating a new function

The best way to create a new AWS Lambda that uses this structure is to use the dotnet new template provided via NuGet.

  1. Ensure you have the latest .NET Core SDK installed.
  2. Open your console prompt of choice and type dotnet new -i "Kralizek.Lambda.Templates". This will install the latest version of the templates. Depending on your previous usage of dotnet new, it might take some time.
  3. List all available templates by typing dotnet new -all. You will see 4 new entries, all starting with Lambda.
  4. Create a new project using the template of your choice by typing dotnet new {short name of the template} -n NameOfYourProject. E.g. dotnet new lambda-template-event-empty -n Sample
  5. Start hacking!

Here is a list of all the available templates

|Name|Short name|Description| |-|-|-| |Lambda Empty Event Function|lambda-template-event-empty|Creates an Event function with no extra setup| |Lambda Empty RequestResponse Function|lambda-template-requestresponse-empty|Creates a RequestResponse function with no extra setup| |Lambda Boilerplate Event Function|lambda-template-event-boilerplate|Creates an Event function with some boilerplate added| |Lambda Boilerplate RequestResponse Function|lambda-template-requestresponse-boilerplate|Creates a RequestResponse function with some boilerplate added| |Lambda SNS Handler Function|lambda-template-sns-event|Creates a function to handle SNS notifications| |Lambda SQS Handler Function|lambda-template-sqs-event|Creates a function to ha

View on GitHub
GitHub Stars61
CategoryDevelopment
Updated1y ago
Forks25

Languages

C#

Security Score

85/100

Audited on Jul 1, 2024

No findings