The Cybtans command-line interface (CLI) is a cross-platform code generation tool for developing microservice based applications. Use the tool for generating microservice's projects structure, app service contracts, data transfer objects, controllers, client libraries for .NET and Typescript among other utilities.

First download cybtans-cli-x64 the cybtans-cli tool from the latest release. Extract it and additionally add the directory where you extract the executable to your PATH.

Let's create a solution file for this example

 dotnet new sln -n MySolution

Next let's create a microservice for managing Product's data

cybtans-cli service -n Products -o ./Products -sln .\MySolution.sln

The tool generates the following project structures under the Products folder

• Products.Client: The microservice client library for C#
• Products.Models The microservice data transfer objects ,request and response messages
• Products.RestApi The microservice Rest API interface. Defines endpoints for calling the microservice operations
• Products.Services The microservice business logic
• Products.Services.Tests The microservice business logic tests
• Proto The microservice protobuff definitions

The cli leverage the Google Protocol Buffer syntax for defining services ,request and response message. From those definitions client libraries, business services ,dtos and controllers are generated.

Next let's add some definitions to the Product.proto file inside the generated Proto directory

syntax = "proto3";

package Products;

message Product{
	int32 id = 1;
	string name = 2 [required = true];
	string description = 3;
	float price = 4 [optional = true];
	string pictureFileName = 5;
	string pictureUrl = 6;
	int32 brandId = 7;
	int32 catalogId = 8;
	int32 avalaibleStock = 9;
	int32 RestockThreshold = 10;
	datetime createDate = 11;
	datetime updateDate = 12 [optional = true];

	Catalog catalog = 13;
	Brand brand = 14;

	repeated Comment comments = 15;
}

message Catalog {
	int32 id = 1;
	string name = 2;
}

message Brand {
	int32 id = 1;
	string name = 2;
}

message Comment {
	int32 id = 1;
	string text = 2;
	string username = 3;
	int32 userId = 4;
	datetime date = 5;
	int8 rating = 6;
}

message GetProductListRequest {
	string filter = 1;
	string sort = 2;
	int32 skip = 3;
	int32 take = 4 [default = 50];
}

message GetProductRequest {
	int32 id = 1;
}

message UpdateProductRequest {
	int32 id = 1;
	Product Product = 2 [optional = true];
}

message DeleteRequest{
	int32 id = 1;
}

message GetProductListResponse {
	repeated Product items = 1;
	int32 page = 2;
	int32 totalPages = 3;
}


service ProductsService {
	option (prefix) ="api/products";

	rpc GetProducts(GetProductListRequest) returns (GetProductListResponse){		
		option method = "GET";
	};

	rpc GetProduct(GetProductRequest) returns (Product){	
		option template = "{id}"; 
		option method = "GET";
	};

	rpc CreateProduct(Product) returns (Product){			
		option method = "POST";
	};

	rpc UpdateProduct(UpdateProductRequest) returns (Product){			
		option template = "{id}"; 
		option method = "PUT";
	};

	rpc DeleteProduct(DeleteRequest) returns (void){		
		option template = "{id}"; 
		option method = "DELETE";
	};
}

There is a file cybtans.json in the microservice root directory that provides configurations settings for the cybtans-cli in order to generate code from the proto file. To generate code from the proto just open a terminal in the solution's root directory and run

cybtans-cli .

The cybtans-cli search recursively into all the subdirectories for the cybtans.json files and run the proto generator from those settings

Now run the Products.RestApi project and navigate in a browser to https://localhost:5001/swagger ,you will see the swagger page containing the descriptions for endpoints of your service's operations/rpcs.

The endpoints are not implemented so it's up to you to implement the service's interface and register the implementation with ServiceCollection for dependency injection.

Use the BinarySerializer to serialize/deserialize objects to and from bytes arrays or stream. The binary serializer is very fast and when used with objects that implements IReflectorMetadataProvider is super fast. It also generates less than a half of the bytes that JSON generates ,thus improving the bandwidth for sending and receiving messages.

Nuget Console.

Install-Package Cybtans.Serialization -Version 1.0.6

DotNet Core CLI.

dotnet add package Cybtans.Serialization --version 1.0.6

    byte[] buffer = BinaryConvert.Serialize(modelA);
    Model result = BinaryConvert.Deserialize<Model>(buffer);

    //Using Unicode encoding
    var serializer = new BinarySerializer(Encoding.Unicode);
    buffer = serializer.Serialize(model);
    result = serializer.Deserialize<Model>(buffer);

BenchmarkDotNet=v0.12.1, OS=Windows 10.0.18363.836 (1909/November2018Update/19H2)
Intel Core i5-5200U CPU 2.20GHz (Broadwell), 1 CPU, 4 logical and 2 physical cores
.NET Core SDK=3.1.201
  [Host]     : .NET Core 3.1.3 (CoreCLR 4.700.20.11803, CoreFX 4.700.20.12001), X64 RyuJIT
  Job-LYLIXC : .NET Core 3.1.3 (CoreCLR 4.700.20.11803, CoreFX 4.700.20.12001), X64 RyuJIT

Runtime=.NET Core 3.1  RunStrategy=Throughput