I added wcf services end point in asp.net core 2.0 to connected services and then I try to use that but with client there is only functions which ended with ..async
I don't want to use ...async.But there is no function without .async
What is problem with this?What should I do?
instead of using that
var response = SystemClient.SearchCountriesAsync(....
I want to use that
var response = SystemClient.SearchCountries(...
but it give that error
Error CS1061 'SystemClient' does not contain a definition for 'SearchCountries' and no extension method 'SearchCountries' accepting a first argument of type 'SystemClient' could be found (are you missing a using directive or an assembly reference?)
Your client does not expose synchronous method but that shouldn't be a problem for you.
Instead of asynchronously calling the method just do this:
response = SystemClient.SearchAirportsAsync(credentials, SystemHelperLanguageTypes.English, SystemHelperSearchTypes.CodeSearch, "ist").Result;
This will call the method synchronously as it will block the call. Check John Skeets answer here.
That being said I would recomend you use the async method that is provided. To support that you would have to change the Action signature to this:
public async Task<IActionResullt> Index()
{
SystemClient SystemClient = new SystemClient();
Credential credential = new Credential();
credential.UserName = "username";
credential.UserPassword = "****";
var response1 = await SystemClient.SearchCountriesAsync(credentials, SystemHelperLanguageTypes.English, SystemHelperSearchTypes.CodeSearch, "TR");
var response = await SystemClient.SearchAirportsAsync(credentials, SystemHelperLanguageTypes.English, SystemHelperSearchTypes.CodeSearch, "ist");
//Do whatever you do with those responses
ViewBag.Language = "ar";
return View();
}
There is a way to generate synchronous methods in your .NET core project in Visual Studio 2019.
Wizard that adds WCF web service reference to your .NET core project has an option Generate Synchronous Operations in the third step, Client Options:
Make sure you check it as it is unchecked by default.
Related
I am playing with the IdentityServer4. Part of that I am trying to build a client using IdentityModel 5.1.0 and trying to use following piece of code available here
// request token
var tokenClient = new TokenClient(disco.TokenEndpoint, "ro.client", "secret");
var tokenResponse = await tokenClient.RequestResourceOwnerPasswordAsync("alice", "password", "api1");
if (tokenResponse.IsError)
{
Console.WriteLine(tokenResponse.Error);
return;
}
Console.WriteLine(tokenResponse.Json);
Console.WriteLine("\n\n");
But this is giving me following error.
error CS1729: 'TokenClient' does not contain a constructor that takes 3 arguments
From the docs, it looks like that page is only applicable to Core 1.0. When I change the documentation to 3.1.0, I get
Sorry This pages does not exist yet
Does this mean that ResourceOwnerPassword flow is not supported for the .NET Core 3.1?
Ctrl + clicking on the method takes to to its signature, where you can find out the specific parameters that the method expects.
Browsing the repo, I've found this snippet on using the password credentials token request:
var response = await _client.RequestPasswordTokenAsync(new PasswordTokenRequest
{
ClientId = "client",
UserName = "user",
Password = "password",
Scope = "scope",
Resource = { "resource1", "resource2" }
});
another overload:
var response = await tokenClient.RequestPasswordTokenAsync(userName: "user", password: "password", scope: "scope");
Or see the actual method definition or another helper.
A useful tip: popular packages usually have a lot of tests. You can check them out to learn how to use the library.
I created 2 Azure Function Apps, both setup with Authentication/Authorization so an AD App was created for both. I would like to setup AD Auth from one Function to the other using MSI. I setup the client Function with Managed Service Identity using an ARM template. I created a simple test function to get the access token and it returns: Microsoft.Azure.Services.AppAuthentication: Token response is not in the expected format.
try {
var azureServiceTokenProvider = new AzureServiceTokenProvider();
string accessToken = await azureServiceTokenProvider.GetAccessTokenAsync("https://myapp-registration-westus-dev.azurewebsites.net/");
log.Info($"Access Token: {accessToken}");
return req.CreateResponse(new {token = accessToken});
}
catch(Exception ex) {
log.Error("Error", ex);
throw;
}
Yes, there is a way to do this. I'll explain at a high level, and then add an item to the MSI documentation backlog to write a proper tutorial for this.
What you want to do is follow this Azure AD authentication sample, but only configure and implement the parts for the TodoListService: https://github.com/Azure-Samples/active-directory-dotnet-daemon.
The role of the TodoListDaemon will be played by a Managed Service Identity instead. So you don't need to register the TodoListDaemon app in Azure AD as instructed in the readme. Just enable MSI on your VM/App Service/Function.
In your code client side code, when you make the call to MSI (on a VM or in a Function or App Service), supply the TodoListService's AppID URI as the resource parameter. MSI will fetch a token for that audience for you.
The code in the TodoListService example will show you how to validate that token when you receive it.
So essentially, what you want to do is register an App in Azure AD, give it an AppID URI, and use that AppID URI as the resource parameter when you make the call to MSI. Then validate the token you receive at your service/receiving side.
Please check that the resource id used "https://myapp-registration-westus-dev.azurewebsites.net/" is accurate. I followed steps here to setup Azure AD authentication, and used the same code as you, and was able to get a token.
https://learn.microsoft.com/en-us/azure/app-service/app-service-mobile-how-to-configure-active-directory-authentication
You can also run this code to check the exact error returned by MSI. Do post the error if it does not help resolve the issue.
HttpClient client = new HttpClient();
client.DefaultRequestHeaders.Add("Secret", Environment.GetEnvironmentVariable("MSI_SECRET"));
var response = await client.GetAsync(String.Format("{0}/?resource={1}&api-version={2}", Environment.GetEnvironmentVariable("MSI_ENDPOINT"), "https://myapp-registration-westus-dev.azurewebsites.net/", "2017-09-01"));
string msiResponse = await response.Content.ReadAsStringAsync().ConfigureAwait(false);
log.Info($"MSI Response: {msiResponse}");
Update:-
This project.json file and run.csx file work for me. Note: The project.json refers to .NET 4.6, and as per Azure Functions documentation (link in comments), .NET 4.6 is the only supported version as of now. You do not need to upload the referenced assembly again. Most probably, incorrect manual upload of netstandard assembly, instead of net452 is causing your issue.
Only the .NET Framework 4.6 is supported, so make sure that your
project.json file specifies net46 as shown here. When you upload a
project.json file, the runtime gets the packages and automatically
adds references to the package assemblies. You don't need to add #r
"AssemblyName" directives. To use the types defined in the NuGet
packages, add the required using statements to your run.csx file.
project.json
{
"frameworks": {
"net46":{
"dependencies": {
"Microsoft.Azure.Services.AppAuthentication": "1.0.0-preview"
}
}
}
}
run.csx
using Microsoft.Azure.Services.AppAuthentication;
public static async Task<HttpResponseMessage> Run(HttpRequestMessage req, TraceWriter log)
{
try
{
var azureServiceTokenProvider = new AzureServiceTokenProvider();
string accessToken = await azureServiceTokenProvider.GetAccessTokenAsync("https://vault.azure.net/");
log.Info($"Access Token: {accessToken}");
return req.CreateResponse(new {token = accessToken});
}
catch(Exception ex)
{
log.Error("Error", ex);
throw;
}
}
I have a spring boot application , which have a spring MVC controller. I am trying to version my rest api using Accept header.
The following is how my Controller looks like
RestController
#RequestMapping(value = "/private/")
public class AppleController {
private final AppleService appleService;
public AppleController(AppleService appleService) {
this.appleService = appleService;
}
#GetMapping(value = "apples/{id}", produces = "application/json; v=1.0",
headers = "Accept=application/json; v=1.0")
public ResponseEntity getByappleId(#PathVariable("id") Long appleId) {
System.out.println("version1");
GetByappleIdResponse response = appleService.findByappleId(appleId);
return new ResponseEntity<>(response, HttpStatus.OK);
}
#GetMapping(value = "apples/{id}", produces = "application/json; v=2.0",
headers = "Accept=application/json; v=2.0")
public ResponseEntity getByappleId2(#PathVariable("id") Long appleId) {
System.out.println("version2");
GetByappleIdResponse response = appleService.findByappleId2(appleId);
return new ResponseEntity<>(response, HttpStatus.OK);
}
Irrespective of the version that I am passing in the Accept header when calling the API always "getByappleId" method is called, hence only version 1 response is returned.
Is there anything wrong in my controller ?
There are many options to implement versioning of REST API:
suggested in the comments approach for manually routing your request;
making version as a part of your Accept header value, f.e.:
(headers = "Accept=application/vnd.name.v1+json")
(headers = "Accept=application/vnd.name.v2+json")
making version as a part of your mapping:
#GetMapping("apples/v1/{id})"
#GetMapping("apples/v2/{id})
So you need to decide which way to go. Some useful links:
Versioning a REST API
Best practices for API versioning?
As described in this answer: https://stackoverflow.com/a/34427044/258813 (and mentioned in the comments) Spring does not support routing using the headers like that.
If you want to support routing via a version header, I would recommend a custom routing condition and annotation - certainly if you are building a large API, it will result in less code and a more elegant solution.
You would define some annotation like #ApiVersion(1) that you can add to any method that is also a request mapping and then add the custom routing condition and it will behave correctly.
I have described using custom routing conditions and annotations (based on subdomains - but that could easily be switched to check headers instead) here: http://automateddeveloper.blogspot.co.uk/2014/12/spring-mvc-custom-routing-conditions.html
So, I'm trying to create a sample where there are the following components/features:
A hangfire server OWIN self-hosted from a Windows Service
SignalR notifications when jobs are completed
Github Project
I can get the tasks queued and performed, but I'm having a hard time sorting out how to then notify the clients (all currently, just until I get it working well) of when the task/job is completed.
My current issue is that I want the SignalR hub to be located in the "core" library SampleCore, but I don't see how to "register it" when starting the webapp SampleWeb. One way I've gotten around that is to create a hub class NotificationHubProxy that inherits the actual hub and that works fine for simple stuff (sending messages from one client to all).
In NotifyTaskComplete, I believe I can get the hub context and then send the message like so:
private void NotifyTaskComplete(int taskId)
{
try
{
var hubContext = GlobalHost.ConnectionManager.GetHubContext<NotificationHub>();
if (hubContext != null)
{
hubContext.Clients.All.sendMessage(string.Format("Task {0} completed.", taskId));
}
}
catch (Exception ex)
{
}
}
BUT, I can't do that if NotificationHubProxy is the class being used as it's part of the SampleWeb library and referencing it from SampleCore would lead to a circular reference.
I know the major issue is the hub in the external assembly, but I can't for the life of me find a relevant sample that's using SignalR or MVC5 or setup in this particular way.
Any ideas?
So, the solution was to do the following two things:
I had to use the SignalR .NET client from the SampleCore assembly to create a HubConnection, to create a HubProxy to "NotificationHub" and use that to Invoke the "SendMessage" method - like so:
private void NotifyTaskComplete(string hostUrl, int taskId)
{
var hubConnection = new HubConnection(hostUrl);
var hub = hubConnection.CreateHubProxy("NotificationHub");
hubConnection.Start().Wait();
hub.Invoke("SendMessage", taskId.ToString()).Wait();
}
BUT, as part of creating that HubConnection - I needed to know the url to the OWIN instance. I decided to pass that a parameter to the task, retrieving it like:
private string GetHostAddress()
{
var request = this.HttpContext.Request;
return string.Format("{0}://{1}", request.Url.Scheme, request.Url.Authority);
}
The solution to having a Hub located in an external assembly is that the assembly needs to be loaded before the SignalR routing is setup, like so:
AppDomain.CurrentDomain.Load(typeof(SampleCore.NotificationHub).Assembly.FullName);
app.MapSignalR();
This solution for this part came from here.
I'm developing an user tracking solution using SignalR, as a fun project to learn SignalR, for ASP.NET MVC applications.
Currently i can track logged users and how long are they on a specific page. If they move to another page i track that also and the timer that SignalR is updating resets... Many other features are implemented or partially implemented.
The problem i'm facing is how to get the full url Controller/Action/Parameters
inside SignalR hub?
When i use HttpContext.Current.Request.Url the url is always /signalr/connect.
NOTE:
var hub = $.connection.myHub;
$.connection.hub.start();
is in the _Layout.cshtml.
UPDATE:
I've tried to use
var location = '#HttpContext.Current.Request.Url';
var hub = $.connection.myHub;
$.connection.hub.start().done(function () {
hub.setLocation(location);
});
And the location is passed correctly but I need it on the Connect() task not later.
Is it possible to do this?
UPDATE 2:
This approach doesn't work
var hub = $.connection.myHub;
$.connection.hub.start(function(){hub.setLocation(location)});
as the Connect() is called before.
In my hub i have several methods but i would like pass a value (in my case a location) to the Connect(), is that possible?
public class MyHub : Hub, IDisconnect, IConnected
{
public Task Connect()
{
//do stuff here
//and i would like to have the **location** value
}
public Task Disconnect()
{
//do stuff here
}
}
Update 3
Use QueryString to pass data before the Connect() occurs.
var location = '#HttpContext.Current.Request.Url';
var hub = $.connection.myHub;
$.connection.hub.qs = "location= + location;
$.connection.hub.start();
Passing data like your location value to Connect() is possible via a querystring parameter: SignalR: How to send data to IConnected.Connect()
Using query-string is not very secure, cause a hacker can forge JS code and send you wrong location breaking whatever logic you have behind it.
You can try to get this from owin-enviromment variables
var underlyingHttpContext =
Context.Request.Environment[typeof(HttpContextBase).FullName] as HttpContextBase;
Then extract whatever you need.
It will work on IIS, for non-IIS hosting look for other OWIN stuff https://github.com/aspnet/AspNetKatana/wiki/OWIN-Keys
You could pass it from your client js call to your hub as a parameter.