In my Asp.net MVC app, I have two methods on a controller, one for when the user first arrives on the view and then one when they submit the form on said view.
public ActionResult Foo() {}
[AcceptVerbs(HttpVerbs.Post)]
public ActionResult Foo(string id, Account accountToFoo) {}
In the second action, there's a custom model binder that's assembling the account object that I'm acting on, though that's really not important. This all works fine in testing locally on a server.
We try to be pretty good about writing unit tests to test all our different views are properly getting routed to, including those that are HTTP POST. To do so, we've been using mvccontrib's test helper.
Testing gets have been super simple
"~/account/foo/myusername".
Route().
ShouldMapTo<AccountController>(c => c.Foo("myusername"));
My question is in testing POST routes, how do I write the lambda that I would use to verify the post is receiving accurate values, similar to the GET test above?
For a POST, it looks something like:
"~/account/foo".
WithMethod(HttpVerbs.Post).
ShouldMapTo<AccountController>(a => something_something);
It's the something_something portion of my lambda that I'm having trouble with. Using arbitrary values doesn't work ("a => a.Foo(0, new Account()"). How would I specify the expected values as part of the test?
EDIT I was hoping there was something akin to the way Moq has lambdas for statements such as foo.Setup(s => s.Foo(It.IsAny(), It.Is(i => i > 32)) and so on. Even I have to explicitly supply the values, that's workable--I just can't seem to grok the desired structure to pass those explicit values.
Here's an example. Assuming you have the following action:
public AccountController : Controller
{
[AcceptVerbs(HttpVerbs.Post)]
public ActionResult Foo(string id)
{
return View();
}
}
And the following route registered:
RouteTable.Routes.MapRoute(
"Default",
"{controller}/{action}/{id}",
new { controller = "home", action = "index", id = "" }
);
You could test it like this:
var routeData = "~/account/foo".WithMethod(HttpVerbs.Post);
routeData.Values["id"] = "123";
routeData.ShouldMapTo<AccountController>(c => c.Foo("123"));
Some tweaking might be necessary to include the second Account argument you have.
Using the mvccontrib helper syntax:
"~/account/foo".WithMethod(HttpVerbs.Post).ShouldMapTo<AccountController>(a => a.foo(null));
You pass null as the Foo(string id, Account accountToFoo) method is never executed as part of the routing test.
Related
I have this simple API controller in a NetCore 2.2 web game that is supposed to return a list of monsters based on the dungeonID(Guid).
So I use the URL, passing in the guid of the dungeonID, to that controller like this:
https://localhost:44361/MonsterList/GetMonsters/2f14c8gf-2e7e-466a-bcbg-f4440e92b3dg
But when I step through the code, I just see all zeroes for the dungeonID:
public async Task<JsonResult> GetMonsters(Guid dungeonID)
{
var monsters = await _context.MonsterList.Where(c => c.DungeonID == (dungeonID)).ToListAsync();
return Json(monsters);
}
This returns nothing because, for reasons I don't know, dungeonID is always all zeroes.
But this does work if I hard-code in the dungeonID:
https://localhost:44361/MonsterList/GetMonsters
public async Task<JsonResult> GetMonsters()
{
var monsters = await _context.MonsterList.Where(c => c.DungeonID == Guid.Parse("2f14c8gf-2e7e-466a-bcbg-f4440e92b3dg")).ToListAsync();
return Json(monsters);
}
I've seen lots of posts similiar to mine, like these:
asp.net webapi 2 post parameter is always null
Post parameter is always null
But nothing seems to work.
How do I add the ability to pass in a Guid parameter?
Thanks!
One point is 2f14c8gf-2e7e-466a-bcbg-f4440e92b3dg is not a GUID/UUID . Just try with a correct GUID :
https://localhost:44384/api/values/GetMonsters/2e6ae748-10c2-4e23-84c3-9d3db7c09631
Being kind of a newb to MVC 4 (or really any of the MVC's for ASP.NET) I cant help but feel theres more to the URL helper than what I'm seeing.
Basically I've read the tutorials on populating the attributes in a controllers methods using a query string in the URL.
I dont liek query strings though and prefer a sectioned "folder" like style.
Without much further adu, this is the sample URL:
http://something.com/DataTypes/Search/searchString
this approach is actually pretty safe as there will only ever be single worded searches
I have tried in the DataTypes controller
[HttpGet]
public ActionResult Search(String q)
{
ViewBag.ProvidedQuery = q;
return View();
}
and a few other small variations, right now im just trying to get the string to show up in the view but I dont seem to be getting anything there.
Is there anyway to inject the 3rd string in the url into an attribute?
If not, which URL helper class am I supposed to use to acquire the string data in the URL? Even if I have to parse the whole URL manually so be it, i just need to acquire that 3rd element in the URL as a string
Extremely n00b question im sure, but either im not finding this simple guide somewhere, or im not searching google correctly...
What you're missing is that the default route parameter name is "id". So you want to do this:
[HttpGet]
public ActionResult Search(String id)
{
ViewBag.ProvidedQuery = id;
return View();
}
If you don't want to use the variable name id, then you can modify your Route to look like this:
routes.MapRoute(
name: "Search",
url: "DataTypes/Search/{searchString}",
defaults: new { controller = "DataTypes", action = "Search",
searchString = UrlParameter.Optional });
If you don't want the string to be optional, then you can remove the last field from the defaults object.
you can use RouteTable.Routes.GetRouteData(new HttpContextWrapper(httpContext)) to get the routedata
String URL to RouteValueDictionary
You need to look at the routing in the Global.asax.cs. For example for your case you could add a route to the routes collection like this:
routes.MapRoute("Search",
"DataTypes/Search/{q}",
new { controller = "DataTypes", action = "Search" }
);
Then the q parameter will automatically get mapped to your action. The default controller mapping is likely mapping it to "id".
I have a route:
routes.MapRoute("ResetPasswordConfirm", "reset-password-confirm", new { controller = "Membership", action = "ResetPasswordConfirm" });
and the code
public ActionResult ResetPasswordConfirm(int userid, string key)
{
// ...
}
in my application. So that i have url to be executed like this:
http://localhost/reset-password-confirm?userid=1&key=bla_bla_something
That is absolutely okay, until someone decides to go to
http://localhost/reset-password-confirm
...and look what will happen. ASP.NET will generate predictable error:
The parameters dictionary contains a null entry for parameter 'userid' of non-nullable type 'System.Int32'...
It could be done also by a search robot trying to grab all the possible urls. It's okay, but generates a lot of errors during usage of application in the wild. I want to avoid that, but feel uncomfortable with writing a stub for every possible case for such kind of errors.
Is there any elegant way of doing that? Thanks.
Another way is to handle global errors, just set <customErrors mode="On"></customErrors> on your web.config and create an Error.cshtml on your Shared view folder. MVC3 templates actually include that page.
On the other hand, if you want to be more specific, you should try Action Filters, that's a cool way to handle errors.
[HandleError(View = "YourErrorView", ExceptionType=typeof(NullReferenceException))]
public ActionResult ResetPasswordConfirm(int? userid, string key)
{
if (!userid.HasValue)
throw new NullReferenceException();
// ...
}
Use nullables for your parameters, i.e.:
public ActionResult ResetPasswordConfirm(int? userid, string key)
While reading about mcv3 I came across an attribute name called [ActionName]. It actually gives a new name to the action method. I tested a scenario which made me think; how are the internals working. When I have the following two action methods in my controller class
[ActionName("Test")]
public ActionResult Index()
{
return View();
}
[ActionName("Index")]
public ActionResult Test()
{
return View();
}
I thought this will end up in some kind of infinite loop or will give some ambiguity exception. But the same works fine and the second method is called when i give this url http://mysite:1234/mycontroller
What made MVC engine to choose the second method and not the first?
Any idea why this happens?
Phil Haack has a post on this matter: How a method becomes an action
In short: the ControllerActionInvoker uses reflection to find a method matches the action name.
The ActionNameAttribute redefines the name of the method.
Also be aware that the name of your View matches the ActionName, not the MethodName: the method Index will search for a view with the name "Test"
This is the magic of the Routing engine. Somewhere within the global.asax.cs file there would be routing patterns defined, mostly which defaults to
routes.MapRoute(
"Default", // Route name
"{controller}/{action}/{id}", // URL with parameters
new { controller = "Home", action = "Index", id = "" } // Parameter defaults
);
This is a routing pattern defined for your application. The action name attribute maps to the 'action' parameter within the parameter collection (3rd parameter for MapRoute).
In your case if you map the action 'Index' to method 'Test'. It should call Test() method. I am not sure whether it is still calling Index() for you. In fact the routing engine does not care about the method name if it finds the ActionName attribute over your public method.
ActionNameAttribute it represents an attribute that is used for the name of an action. If it is not present the name of the method is used.
The populated catList is always Count=0 when the code jumps to CreateProduct() so I take it it does not get delivered.
Considering RouteValueDictionary does not do this ? Any other way?
public ActionResult GetCats(int CatID)
{
List<Category> catList = new List<Category>();
if (CatID >= 0 )
{
catList = context.Categories.Where(x => x.PCATID == CatID).ToList();
}
return RedirectToAction("CreateProduct", "Admin", new { catList });
}
public ActionResult CreateProduct(List<Category> catList) { }
You are actually trying to use controllers to do data access.
Move the "GetCats" data retrieval into your business layer (Service object, Repository, whatever suits you).
Then, CreateProduct will need to be there twice (2 signatures). One with no parameters in which you are going to call "GetCats" from your business layer and send it to the view.
The other implementation is going to be the flagged with the HttpPostAttribute and will contain in parameters all the necessary information to create a cat.
That's all. Simple and easy.
You could place any items that you need in TempData then call RedirectToAction.
RedirectToAction simply returns a "302" code to the browser with the URL to redirect to. When this happens your browser performs a GET with that URL.
Your RouteValues need to be simple. You can't really pass complex object or collections using a route value.
if you don't care about the browser url changing you could just
return CreateProduct(catList)