nanoFramework.WebServer.FileSystem 1.2.70

Welcome to the .NET nanoFramework WebServer repository

Build status

Component	Build Status	NuGet Package
nanoFramework.WebServer
nanoFramework.WebServer.FileSystem

.NET nanoFramework WebServer

This library was coded by Laurent Ellerbach who generously offered it to the .NET nanoFramework project.

This is a simple nanoFramework WebServer. Features:

• Handle multi-thread requests
• Serve static files from any storage using nanoFramework.WebServer.FileSystem NuGet. Requires a target device with support for storage (having System.IO.FileSystem capability).
• Handle parameter in URL
• Possible to have multiple WebServer running at the same time
• supports GET/PUT and any other word
• Supports any type of header
• Supports content in POST
• Reflection for easy usage of controllers and notion of routes
• Helpers to return error code directly facilitating REST API
• HTTPS support
• URL decode/encode

Limitations:

• Does not support any zip in the request or response stream

Usage

You just need to specify a port and a timeout for the queries and add an event handler when a request is incoming. With this first way, you will have an event raised every time you'll receive a request.

using (WebServer server = new WebServer(80, HttpProtocol.Http)
{
    // Add a handler for commands that are received by the server.
    server.CommandReceived += ServerCommandReceived;

    // Start the server.
    server.Start();

    Thread.Sleep(Timeout.Infinite);
}

You can as well pass a controller where you can use decoration for the routes and method supported.

using (WebServer server = new WebServer(80, HttpProtocol.Http, new Type[] { typeof(ControllerPerson), typeof(ControllerTest) }))
{
    // Start the server.
    server.Start();

    Thread.Sleep(Timeout.Infinite);
}

In this case, you're passing 2 classes where you have public methods decorated which will be called every time the route is found.

With the previous example, a very simple and straight forward Test controller will look like that:

public class ControllerTest
{
    [Route("test"), Route("Test2"), Route("tEst42"), Route("TEST")]
    [CaseSensitive]
    [Method("GET")]
    public void RoutePostTest(WebServerEventArgs e)
    {
        string route = $"The route asked is {e.Context.Request.RawUrl.TrimStart('/').Split('/')[0]}";
        e.Context.Response.ContentType = "text/plain";
        WebServer.OutPutStream(e.Context.Response, route);
    }

    [Route("test/any")]
    public void RouteAnyTest(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }
}

In this example, the RoutePostTest will be called every time the called url will be test or Test2 or tEst42 or TEST, the url can be with parameters and the method GET. Be aware that Test won't call the function, neither test/.

The RouteAnyTestis called whenever the url is test/any whatever the method is.

There is a more advance example with simple REST API to get a list of Person and add a Person. Check it in the sample.

[!Important]

By default the routes are not case sensitive and the attribute must be lowercase. If you want to use case sensitive routes like in the previous example, use the attribute CaseSensitive. As in the previous example, you must write the route as you want it to be responded to.

A simple GPIO controller REST API

You will find in simple GPIO controller sample REST API. The controller not case sensitive and is working like this:

• To open the pin 2 as output: http://yoururl/open/2/output
• To open pin 4 as input: http://yoururl/open/4/input
• To write the value high to pin 2: http://yoururl/write/2/high
  • You can use high or 1, it has the same effect and will place the pin in high value
  • You can use low of 0, it has the same effect and will place the pin in low value
• To read the pin 4: http://yoururl/read/4, you will get as a raw text highor lowdepending on the state

Authentication on controllers

Controllers support authentication. 3 types of authentications are currently implemented on controllers only:

• Basic: the classic user and password following the HTTP standard. Usage:
  • [Authentication("Basic")] will use the default credential of the webserver
  • [Authentication("Basic:myuser mypassword")] will use myuser as a user and my password as a password. Note: the user cannot contains spaces.
• APiKey in header: add ApiKey in headers with the API key. Usage:
  • [Authentication("ApiKey")] will use the default credential of the webserver
  • [Authentication("ApiKeyc:akey")] will use akey as ApiKey.
• None: no authentication required. Usage:
  • [Authentication("None")] will use the default credential of the webserver

The Authentication attribute applies to both public Classes an public Methods.

As for the rest of the controller, you can add attributes to define them, override them. The following example gives an idea of what can be done:

[Authentication("Basic")]
class ControllerAuth
{
    [Route("authbasic")]
    public void Basic(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }

    [Route("authbasicspecial")]
    [Authentication("Basic:user2 password")]
    public void Special(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }

    [Authentication("ApiKey:superKey1234")]
    [Route("authapi")]
    public void Key(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }

    [Route("authnone")]
    [Authentication("None")]
    public void None(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }

    [Authentication("ApiKey")]
    [Route("authdefaultapi")]
    public void DefaultApi(WebServerEventArgs e)
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.OK);
    }
}

And you can pass default credentials to the server:

using (WebServer server = new WebServer(80, HttpProtocol.Http, new Type[] { typeof(ControllerPerson), typeof(ControllerTest), typeof(ControllerAuth) }))
{
    // To test authentication with various scenarios
    server.ApiKey = "ATopSecretAPIKey1234";
    server.Credential = new NetworkCredential("topuser", "topPassword");

    // Start the server.
    server.Start();

    Thread.Sleep(Timeout.Infinite);
}

With the previous example the following happens:

• All the controller by default, even when nothing is specified will use the controller credentials. In our case, the Basic authentication with the default user (topuser) and password (topPassword) will be used.
  • When calling http://yoururl/authbasic from a browser, you will be prompted for the user and password, use the default one topuser and topPassword to get access
  • When calling http://yoururl/authnone, you won't be prompted because the authentication has been overridden for no authentication
  • When calling http://yoururl/authbasicspecial, the user and password are different from the defautl ones, user2 and password is the right couple here
• If you would have define in the controller a specific user and password like [Authentication("Basic:myuser mypassword")], then the default one for all the controller would have been myuser and mypassword
• When calling http://yoururl/authapi, you must pass the header ApiKey (case sensitive) with the value superKey1234 to get authorized, this is overridden the default Basic authentication
• When calling http://yoururl/authdefaultapi, the default key ATopSecretAPIKey1234 will be used so you have to pass it in the headers of the request

All up, this is an example to show how to use authentication, it's been defined to allow flexibility.

The webserver supports having multiple authentication methods or credentials for the same route. Each pair of authentication method plus credentials should have its own method in the controller:

class MixedController
{

    [Route("sameroute")]
    [Authentication("Basic")]
    public void Basic(WebServerEventArgs e)
    {
        WebServer.OutPutStream(e.Context.Response, "sameroute: Basic");
    }

    [Authentication("ApiKey:superKey1234")]
    [Route("sameroute")]
    public void Key(WebServerEventArgs e)
    {
        WebServer.OutPutStream(e.Context.Response, "sameroute: API key #1");
    }

    [Authentication("ApiKey:superKey5678")]
    [Route("sameroute")]
    public void Key2(WebServerEventArgs e)
    {
        WebServer.OutPutStream(e.Context.Response, "sameroute: API key #2");
    }

    [Route("sameroute")]
    public void None(WebServerEventArgs e)
    {
        WebServer.OutPutStream(e.Context.Response, "sameroute: Public");
    }
}

The webserver selects the route for a request:

• If there are no matching methods, a not-found response (404) is returned.
• If authentication information is passed in the header of the request, then only methods that require authentication are considered. If one of the method's credentials matches the credentials passed in the request, that method is called. Otherwise a non-authorized response (401) will be returned.
• If no authentication information is passed in the header of the request:
  • If one of the methods does not require authentication, that method is called.
  • Otherwise a non-authorized response (401) will be returned. If one of the methods requires basic authentication, the WWW-Authenticate header is included to request credentials.

The webserver does not support more than one matching method. Calling multiple methods most likely results in an exception as a subsequent method tries to modify a response that is already processed by the first method. The webserver does not know what to do and returns an internal server error (500). The body of the response lists the matching methods.

Having multiple matching methods is considered a programming error. One way this occurs is if two methods in a controller accidentally have the same route. Returning an internal server error with the names of the methods makes it easy to discover the error. It is expected that the error is discovered and fixed in testing. Then the internal error will not occur in the application that is deployed to a device.

Managing incoming queries thru events

Very basic usage is the following:

private static void ServerCommandReceived(object source, WebServerEventArgs e)
{
    var url = e.Context.Request.RawUrl;
    Debug.WriteLine($"Command received: {url}, Method: {e.Context.Request.HttpMethod}");

    if (url.ToLower() == "/sayhello")
    {
        // This is simple raw text returned
        WebServer.OutPutStream(e.Context.Response, "It's working, url is empty, this is just raw text, /sayhello is just returning a raw text");
    }
    else
    {
        WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.NotFound);
    }
}

You can do more advance scenario like returning a full HTML page:

WebServer.OutPutStream(e.Context.Response, "<html><head>" +
    "<title>Hi from nanoFramework Server</title></head><body>You want me to say hello in a real HTML page!<br/><a href='/useinternal'>Generate an internal text.txt file</a><br />" +
    "<a href='/Text.txt'>Download the Text.txt file</a><br>" +
    "Try this url with parameters: <a href='/param.htm?param1=42&second=24&NAme=Ellerbach'>/param.htm?param1=42&second=24&NAme=Ellerbach</a></body></html>");

And can get parameters from a URL a an example from the previous link on the param.html page:

if (url.ToLower().IndexOf("/param.htm") == 0)
{
    // Test with parameters
    var parameters = WebServer.decryptParam(url);
    string toOutput = "<html><head>" +
        "<title>Hi from nanoFramework Server</title></head><body>Here are the parameters of this URL: <br />";
    foreach (var par in parameters)
    {
        toOutput += $"Parameter name: {par.Name}, Value: {par.Value}<br />";
    }
    toOutput += "</body></html>";
    WebServer.OutPutStream(e.Context.Response, toOutput);
}

And server static files:

// E = USB storage
// D = SD Card
// I = Internal storage
// Adjust this based on your configuration
const string DirectoryPath = "I:\\";
string[] _listFiles;

// Gets the list of all files in a specific directory
// See the MountExample for more details if you need to mount an SD card and adjust here
// https://github.com/nanoframework/Samples/blob/main/samples/System.IO.FileSystem/MountExample/Program.cs
_listFiles = Directory.GetFiles(DirectoryPath);
// Remove the root directory
for (int i = 0; i < _listFiles.Length; i++)
{
    _listFiles[i] = _listFiles[i].Substring(DirectoryPath.Length);
}

var fileName = url.Substring(1);
// Note that the file name is case sensitive
// Very simple example serving a static file on an SD card                   
foreach (var file in _listFiles)
{
    if (file == fileName)
    {
        WebServer.SendFileOverHTTP(e.Context.Response, DirectoryPath + file);
        return;
    }
}

WebServer.OutputHttpCode(e.Context.Response, HttpStatusCode.NotFound);

[!Important]

Serving files requires the nanoFramework.WebServer.FileSystem nuget AND that the device supports storage so System.IO.FileSystem.

And also REST API is supported, here is a comprehensive example:

if (url.ToLower().IndexOf("/api/") == 0)
{
    string ret = $"Your request type is: {e.Context.Request.HttpMethod}\r\n";
    ret += $"The request URL is: {e.Context.Request.RawUrl}\r\n";
    var parameters = WebServer.DecodeParam(e.Context.Request.RawUrl);
    if (parameters != null)
    {
        ret += "List of url parameters:\r\n";
        foreach (var param in parameters)
        {
            ret += $"  Parameter name: {param.Name}, value: {param.Value}\r\n";
        }
    }

    if (e.Context.Request.Headers != null)
    {
        ret += $"Number of headers: {e.Context.Request.Headers.Count}\r\n";
    }
    else
    {
        ret += "There is no header in this request\r\n";
    }

    foreach (var head in e.Context.Request.Headers?.AllKeys)
    {
        ret += $"  Header name: {head}, Values:";
        var vals = e.Context.Request.Headers.GetValues(head);
        foreach (var val in vals)
        {
            ret += $"{val} ";
        }

        ret += "\r\n";
    }

    if (e.Context.Request.ContentLength64 > 0)
    {

        ret += $"Size of content: {e.Context.Request.ContentLength64}\r\n";

        var contentTypes = e.Context.Request.Headers?.GetValues("Content-Type");
        var isMultipartForm = contentTypes != null && contentTypes.Length > 0 && contentTypes[0].StartsWith("multipart/form-data;");

        if(isMultipartForm)
        {
            var form = e.Context.Request.ReadForm();
            ret += $"Received a form with {form.Parameters.Length} parameters and {form.Files.Length} files.";
        }
        else 
        {
            var body = e.Context.Request.ReadBody();

            ret += $"Request body hex string representation:\r\n";
            for (int i = 0; i < body.Length; i++)
            {
                ret += body[i].ToString("X") + " ";
            }
        }

    }

    WebServer.OutPutStream(e.Context.Response, ret);
}

This API example is basic but as you get the method, you can choose what to do.

As you get the url, you can check for a specific controller called. And you have the parameters and the content payload!

Notice the extension methods to read the body of the request:

• ReadBody will read the data from the InputStream while the data is flowing in which might be in multiple passes depending on the size of the body
• ReadForm allows to read a multipart/form-data form and returns the text key/value pairs as well as any files in the request

Example of a result with call:

And more! Check the complete example for more about this WebServer!

Using HTTPS

You will need to generate a certificate and keys:

X509Certificate _myWebServerCertificate509 = new X509Certificate2(_myWebServerCrt, _myWebServerPrivateKey, "1234");

// X509 RSA key PEM format 2048 bytes
        // generate with openssl:
        // > openssl req -newkey rsa:2048 -nodes -keyout selfcert.key -x509 -days 365 -out selfcert.crt
        // and paste selfcert.crt content below:
        private const string _myWebServerCrt =
@"-----BEGIN CERTIFICATE-----
MORETEXT
-----END CERTIFICATE-----";

        // this one is generated with the command below. We need a password.
        // > openssl rsa -des3 -in selfcert.key -out selfcertenc.key
        // the one below was encoded with '1234' as the password.
        private const string _myWebServerPrivateKey =
@"-----BEGIN RSA PRIVATE KEY-----
MORETEXTANDENCRYPTED
-----END RSA PRIVATE KEY-----";

using (WebServer server = new WebServer(443, HttpProtocol.Https)
{
    // Add a handler for commands that are received by the server.
    server.CommandReceived += ServerCommandReceived;
    server.HttpsCert = _myWebServerCertificate509;

    server.SslProtocols = System.Net.Security.SslProtocols.Tls | System.Net.Security.SslProtocols.Tls11 | System.Net.Security.SslProtocols.Tls12;
    // Start the server.
    server.Start();

    Thread.Sleep(Timeout.Infinite);
}

[!IMPORTANT] Because the certificate above is not issued from a Certificate Authority it won't be recognized as a valid certificate. If you want to access the nanoFramework device with your browser, for example, you'll have to add the CRT file as a trusted one. On Windows, you just have to double click on the CRT file and then click "Install Certificate...".

You can of course use the routes as defined earlier. Both will work, event or route with the notion of controller.

WebServer status

It is possible to subscribe to an event to get the WebServer status. That can be useful to restart the server, put in place a retry mechanism or equivalent.

server.WebServerStatusChanged += WebServerStatusChanged;

private static void WebServerStatusChanged(object obj, WebServerStatusEventArgs e)
{
    // Do whatever you need like restarting the server
    Debug.WriteLine($"The web server is now {(e.Status == WebServerStatus.Running ? "running" : "stopped" )}");
}

E2E tests

There is a collection of postman tests nanoFramework WebServer E2E Tests.postman_collection.json in WebServerE2ETests which should be used for testing WebServer in real world scenario. Usage is simple:

• Import json file into Postman
• Deploy WebServerE2ETests to your device - copy IP
• Set the base_url variable to match your device IP address
• Choose request you want to test or run whole collection and check tests results.

The WebServerE2ETests project requires the name and credentials for the WiFi access point. That is stored in the WiFi.cs file that is not part of the git repository. Build the WebServerE2ETests to create a template for that file, then change the SSID and credentials. Your credentials will not be part of a commit.

Feedback and documentation

For documentation, providing feedback, issues and finding out how to contribute please refer to the Home repo.

Join our Discord community here.

Credits

The list of contributors to this project can be found at CONTRIBUTORS.

License

The nanoFramework WebServer library is licensed under the MIT license.

Code of Conduct

This project has adopted the code of conduct defined by the Contributor Covenant to clarify expected behaviour in our community. For more information see the .NET Foundation Code of Conduct.

.NET Foundation

This project is supported by the .NET Foundation.