This is an example of what I want to achieve, however I want to do my own custom attribute that also feeds itself from something other than the request url. In the case of HttpGet/HttpPost these built-in attributes obviously have to look at the http request method, but is there truly no way to make Url.Action() resolve the correct url then?
[HttpGet("mygeturl")]
[HttpPost("myposturl")]
public ActionResult IndexAsync()
{
// correct result: I get '/mygeturl' back
var getUrl = Url.Action("Index");
// wrong result: It adds a ?method=POST query param instead of returning '/myposturl'
var postUrl = Url.Action("Index", new { method = "POST" });
return View();
}
I've looked at the aspnet core source code and I truly can't find a feature that would work here. All the LinkGenerator source code seems to require routedata values but routedata always seems to require to be in the url somewhere, either in the path or in the query string. But even if I add the routedata value programmatically, it won't be in time for the action selection or the linkgenerator doesn't care.
In theory what I need is to pass something to the UrlHelper/LinkGenerator and have it understand that I want the url back out that I defined in my custom attribute, in this case the HttpPost (but I'll make my own attribute).
If I have an Apex function that is named authorize() that just gets a username, password, and session token, and another function called getURL('id#', 'key'), that takes an id# for the record as a string and a key for the image to return as a string as parameters. getURL calls the authorize function inside it in order to get the credentials for its callout. The authorize is a post request, and the getURL is a get request.
I am trying to figure out how to test both of these callouts just so I can make sure that getURL is returning the proper JSON as a response. It doesn't even have to be the URL yet which is its intention eventually. But I just need to test it to make sure these callouts are working and that I am getting a response back for the 75% code coverage that it needs.
I made a multiRequestMock class that looks like this:
public class MultiRequestMock implements HttpCalloutMock {
Map<String, HttpCalloutMock> requests;
public MultiRequestMock(Map<String, HttpCalloutMock> requests) {
this.requests = requests;
}
public HTTPResponse respond(HTTPRequest req) {
HttpCalloutMock mock = requests.get(req.getEndpoint());
if (mock != null) {
return mock.respond(req);
} else {
throw new MyCustomException('HTTP callout not supported for test methods');
}
}
public void addRequestMock(String url, HttpCalloutMock mock) {
requests.put(url, mock);
}
}
I then began to write a calloutTest.cls file but wasn't sure how to use this mock class in order to test my original functions. Any clarity or assistance on this would be helpful Thank you.
I believe in your calloutTest class you use Test.setMock(HttpCalloutMock.class, new MultiRequestMock(mapOfRequests)); then call the getUrl and/or authorize methods and instead of the request really executing the response returned will be that which is specified in the response(HttpRequest) method you have implemented in the MultiRequestMock class. That is basically how I see it working, for more info and an example you can see this resource on testing callout classes. This will get you the code coverage you need but unfortunately cannot check you are getting the correct JSON response. For this, you may be able to use the dev console and Execute Anonymous?
You may want to look at simplifying your HttpCalloutMock Implementation and think about removing the map from the constructor as this class really only needs to return a simple response then your calloutTest class can be where you make sure the returned response is correct.
Hope this helps
I have two POST methods and one DELETE in my resource. They have same path.
I annotated one of POSTs with #DefaultMethod, so when someone doesn't send correct Accept header, correct method will be selected. But this causes that when DELETE is called, cxf selects POST instead of correct delete method. Is there any workaround for this?
CXF version: 3.1.17
#DefaultMethod
#POST
#Consumes(MeasurementMediaType.MEASUREMENT_TYPE)
#Produces(MeasurementMediaType.MEASUREMENT_TYPE)
public Response post(MeasurementRepresentation measurementRepresentation, #HeaderParam(value = HttpHeaders.ACCEPT) String acceptHeader) URISyntaxException {
...
}
#POST
#Consumes(MEASUREMENT_COLLECTION_TYPE)
#Produces(MEASUREMENT_COLLECTION_TYPE)
public Response post(MeasurementCollectionRepresentation measurementCollectionRepresentation, #HeaderParam(value = HttpHeaders.ACCEPT) String acceptHeader) {
...
}
#DELETE
public Response delete(
#QueryParam("fragmentType") String fragmentType,
#QueryParam("source") String source,
#QueryParam("dateFrom") DateTime dateFrom,
#QueryParam("dateTo") DateTime dateTo,
#QueryParam("type") String type) {
...
}
java.lang.NullPointerException
at com.cumulocity.measurement.rest.resources.MeasurementCollectionResource.post(MeasurementCollectionResource.java:280)
Two things:
1) DefaultMethod appears not to refer to the default method to select, but the default HTTPMethod. So it's essentially overriding the httpmethod of your call. This is a CXF extension to JAX-RS, so you may be able to ask the CXF team to update the functionality or create a new annotation for your use case.
2) If I understand you correctly, you would like the first method to be called if someone sends the body of {"Hello" : "World"}? Wouldn't you then get errors when trying to construct your MeasurementRepresentation? If they send a bad request, why not let CXF respond with an appropriate HTTP error code?
In previous versions of asp.net, we could use
#Request.Url.AbsoluteUri
But it seems it's changed. How can we do that in asp.net core 1.0?
You have to get the host and path separately.
#Context.Request.Host
#Context.Request.Path
You need scheme, host, path and queryString
#string.Format("{0}://{1}{2}{3}", Context.Request.Scheme, Context.Request.Host, Context.Request.Path, Context.Request.QueryString)
or using new C#6 feature "String interpolation"
#($"{Context.Request.Scheme}://{Context.Request.Host}{Context.Request.Path}{Context.Request.QueryString}")
You can use the extension method of Request:
Request.GetDisplayUrl()
This was apparently always possible in .net core 1.0 with Microsoft.AspNetCore.Http.Extensions, which adds extension to HttpRequest to get full URL; GetEncodedUrl.
e.g. from razor view:
#using Microsoft.AspNetCore.Http.Extensions
...
Link to myself
Since 2.0, also have relative path and query GetEncodedPathAndQuery.
Use the AbsoluteUri property of the Uri, with .Net core you have to build the Uri from request like this,
var location = new Uri($"{Request.Scheme}://{Request.Host}{Request.Path}{Request.QueryString}");
var url = location.AbsoluteUri;
e.g. if the request url is 'http://www.contoso.com/catalog/shownew.htm?date=today' this will return the same url.
You can consider to use this extension method (from Microsoft.AspNetCore.Http.Extensions namespace:
#Context.Request.GetDisplayUrl()
For some my projects i prefer more flexible solution. There are two extensions methods.
1) First method creates Uri object from incoming request data (with some variants through optional parameters).
2) Second method receives Uri object and returns string in following format (with no trailing slash): Scheme_Host_Port
public static Uri GetUri(this HttpRequest request, bool addPath = true, bool addQuery = true)
{
var uriBuilder = new UriBuilder
{
Scheme = request.Scheme,
Host = request.Host.Host,
Port = request.Host.Port.GetValueOrDefault(80),
Path = addPath ? request.Path.ToString() : default(string),
Query = addQuery ? request.QueryString.ToString() : default(string)
};
return uriBuilder.Uri;
}
public static string HostWithNoSlash(this Uri uri)
{
return uri.GetComponents(UriComponents.SchemeAndServer, UriFormat.UriEscaped);
}
Usage:
//before >> https://localhost:44304/information/about?param1=a¶m2=b
Request.GetUri(addQuery: false);
//after >> https://localhost:44304/information/about
//before >> https://localhost:44304/information/about?param1=a¶m2=b
new Uri("https://localhost:44304/information/about?param1=a¶m2=b").GetHostWithNoSlash();
//after >> https://localhost:44304
There is a clean way to get the current URL from a Razor page or PageModel class. That is:
Url.PageLink()
Please note that I meant, the "ASP.NET Core Razor Pages", not the MVC.
I use this method when I want to print the canonical URL meta tag in the ASP.NET Core razor pages. But there is a catch. It will give you the URL which is supposed to be the right URL for that page. Let me explain.
Say, you have defined a route named "id" for your page and therefore, your URL should look like
http://example.com/product?id=34
The Url.PageLink() will give you exactly that URL as shown above.
Now, if the user adds anything extra on that URL, say,
http://example.com/product?id=34&somethingElse
Then, you will not get that "somethingElse" from this method. And that is why it is exactly good for printing canonical URL meta tag in the HTML page.
The accepted answer helped me, as did the comment for it from #padigan but if you want to include the query-string parameters as was the case for me then try this:
#Context.Request.PathBase#Context.Request.GetEncodedPathAndQuery()
And you will need to add #using Microsoft.AspNetCore.Http.Extensions in the view in order for the GetEncodedPathAndQuery() method to be available.
public string BuildAbsolute(PathString path, QueryString query = default(QueryString), FragmentString fragment = default(FragmentString))
{
var rq = httpContext.Request;
return Microsoft.AspNetCore.Http.Extensions.UriHelper.BuildAbsolute(rq.Scheme, rq.Host, rq.PathBase, path, query, fragment);
}
If you're looking to also get the port number out of the request you'll need to access it through the Request.Host property for AspNet Core.
The Request.Host property is not simply a string but, instead, an object that holds both the host domain and the port number. If the port number is specifically written out in the URL (i.e. "https://example.com:8080/path/to/resource"), then calling Request.Host will give you the host domain and the port number like so: "example.com:8080".
If you only want the value for the host domain or only want the value for the port number then you can access those properties individually (i.e. Request.Host.Host or Request.Host.Port).
var returnUrl = string.IsNullOrEmpty(Context.Request.Path) ? "~/" : $"~{Context.Request.Path.Value}{Context.Request.QueryString}";
You may want to get the URL to use it on the razor side, there is an alternative way to get the home app URL:
Url.Content("~/.....")
Example
In the following example, I wanted to generate a QR code and display it in an img tag src, because of using custom route annotation in the Action, I can't use #URL.Action so as an alternative solution I use ~ like this:
<script>
$("#imgCode").attr("src", "#(Url.Content("~/generateQr/"))"+ code);
</script>
Controller Side
[Route("/generateQr/{code}")]
...
ILSpy show how it was done in Microsoft.Owin.dll.
// Microsoft.Owin.OwinRequest
using System;
/// <summary>
/// Gets the uniform resource identifier (URI) associated with the request.
/// </summary>
/// <returns>The uniform resource identifier (URI) associated with the request.</returns>
public virtual Uri Uri => new Uri(string.Concat(Scheme, Uri.SchemeDelimiter, Host, PathBase, Path, QueryString));
I wonder why they removed this property.
I run a web service where I convert a file from one file format into another. The conversion logic is already functioning but now, I want to query this logic via Jersey. Whenever file upload via Jersey is addressed in tutorials / questions, people describe how to do this using multipart form data. I do however simply want to send and return a single file and skip the overhead of sending multiple parts. (The webservice is triggered by another machine which I control so there is no HTML form involved.)
My question is how would I achieve something like the following:
#POST
#Path("{sessionId"}
#Consumes("image/png")
#Produces("application/pdf")
public Response put(#PathParam("sessionId") String sessionId,
#WhatToPutHere InputStream uploadedFileStream) {
return BusinessLogic.convert(uploadedFile); // returns StreamingOutput - works!
}
How do I get hold of the uploadedFileStream (It should be some annotation, I guess which is of course not #WhatToPutHere). I figured out how to directly return a file via StreamingOutput.
Thanks for any help!
You do not have to put anything in the second param of the function; just leave it un-annoted.
The only thing you have to be carefull is to "name" the resource:
The resource should have an URI like: someSite/someRESTEndPoint/myResourceId so the function should be:
#POST
#Path("{myResourceId}")
#Consumes("image/png")
#Produces("application/pdf")
public Response put(#PathParam("myResourceId") String myResourceId,
InputStream uploadedFileStream) {
return BusinessLogic.convert(uploadedFileStream);
}
If you want to use some kind of SessionID, I'd prefer to use a Header Param... something like:
#POST
#Path("{myResourceId}")
#Consumes("image/png")
#Produces("application/pdf")
public Response put(#HeaderParam("sessionId") String sessionId,
#PathParam("myResourceId") String myResourceId,
InputStream uploadedFileStream) {
return BusinessLogic.convert(uploadedFileStream);
}