Spray.io test response not matching actual output - scala

I'm trying to set up some tests for an API made by a coworker with spray.io, and I'm encountering some odd behavior. When a request results in an error for any reason, we want to return a JSON value along the lines of:
{"status":false,"message":"useful message here"}
This happens just fine in the actual browser. I have navigated to an unhandled route in the web browser, and I get the desired JSON value. So, I want to test this. Now, since I'm new to spray.io, I started off with the very simple test:
"leave GET requests to root path unhandled" in {
Get() ~> myRoute ~> check {
handled must beFalse
}
}
This went fine, no problems. Since it's my first time playing with spray.io, I looked at some of the sample tests for testing false routes, and wrapped myRoute with sealRoute() so I could check the response without failing tests:
"leave GET requests to root path unhandled" in {
Get() ~> sealRoute(myRoute) ~> check {
handled must beTrue
}
}
This also works fine. So, I decided to just make sure the text of the response was usable with this, before I went to the trouble of parsing JSON and verifying individual values:
"leave GET requests to root path unhandled" in {
Get() ~> sealRoute(myRoute) ~> check {
responseAs[String] contains "false"
}
}
This is failing. To investigate, I threw a simple line of code in to log the actual value of responseAs[String] to a file, and I got this:
The requested resource could not be found.
Can anyone tell me what I'm doing wrong? I'm thinking that one of the following is occurring:
responseAs[String] is doing more than taking the exact response and giving it back to me, applying some type of filter along the way
The framework itself is not fully evaluating the query, but rather making a mockup object for the test framework to evaluate, and therefore not executing the desired 'turn errors to json' methods that my co-worker has implemented
I have tried searching google and stack overflow specifically for similar issues, but I'm either not putting in the right queries, or most other people are content to have the default error messages and aren't trying to test them beyond checking handled must beFalse.
Edit - This is the relevant part of the RejectionHandler:
case MissingQueryParamRejection(paramName) :: _=>
respondWithMediaType(`application/json`) {
complete(BadRequest, toJson(Map("status" -> false, "message" -> s"Missing parameter $paramName, request denied")))
}

Okay, so with insight from here and a coworker, the problem has been found:
Basically, the custom RejectionHandler was defined within our custom Actor object, and it wasn't coming into scope in the tests. To resolve this, the following steps were taken:
Moved the definition for the custom RejectionHandler into its own object in a separate file (as it had to do its own imports, it was causing a "encountered unrecoverable cycle resolving import" error)
Imported this new object into both the original file and the test spec.
(fun fact - http://spray.io/documentation/1.2.2/spray-routing/key-concepts/rejections/#rejectionhandler seems to demonstrate the RejectionHandler as a top-level object but you can't have top-level implicit vals in Scala, hence the need for an object)

Related

How to handle a basic form submission with http4s?

I can't believe this isn't in the http4s documentation, and the example code I was able to dig up online (after poking around long enough to discover the UrlForm class) is not working for me.
The relevant bit of code looks like this:
case req # POST -> Root / "compose" =>
req.decode[UrlForm] { ps =>
println("ps.values: " + ps.values)
val content = ps.getFirstOrElse("content",
throw new IllegalStateException("No content given!"))
// Do something with `content`...
}
When submitting the associated form, the IllegalStateException is thrown. ps.values is an empty map (Map()).
I can see (using println) that the Content-Type is application/x-www-form-urlencoded, as expected, and I can see from my browser's Network tab that request "paramaters" (the encoded form values) are being sent properly.
The problem is that I had a filter (javax.servlet.Filter) in place that was calling getParameterMap on the HttpServletRequest. This was draining the InputStream for the request, and it was happening before the request got passed off to the servlet (BlockingHttp4sServlet) instance.
It seems to me the BlockingHttp4sServlet should raise an IllegalStateException (or something more descriptive) when it receives an InputStream with isFinished returning true. (I've filed an issue with the http4s project on Github.)

Adding headers in a Spray.io Test Spec fails to send the request into the route

I'm on Spray 1.3.3. I have a spec that looks like the following:
"Add a collection using a PATCH with correct update hash" in {
Patch("/datastore/collections/FMI", CollectionPatch(addPriceCodes =
Some(List(PriceCode("AMI", "22", None))))) ~>
addHeader(hashHeader, updateHash) ~> routes ~> check {
status should equal(Accepted)
}
}
This looks like it should in the documentation I've seen, and it compiles, but when I run it, I get a 500 back. If I set a breakpoint in the route, I don't ever get inside of it (where I do in other test functions that don't add a header). In IntelliJ I get an error saying it can't resolve the symbol ~> between the addHeader and routes.
What's the right way to add in a header to send to the route?
Turns out what I have listed above works fine. I was getting into my route and was looking at the wrong thing in the exception backtrace.

How to fail a Gatling test from within "exec"?

A Gatling scenario with an exec chain. After a request, returned data is saved. Later it's processed and depending on the processing result, it should either fail or pass the test.
This seems like the simplest possible scenario, yet I can't find any reliable info how to fail a test from within an exec block. assert breaks the scenario and seemingly Gatling (as in: the exception throw doesn't just fail the test).
Example:
// The scenario consists of a single test with two exec creating the execChain
val scn = scenario("MyAwesomeScenario").exec(reportableTest(
// Send the request
exec(http("127.0.0.1/Request").get(requestUrl).check(status.is(200)).check(bodyString.saveAs("MyData")
// Process the data
.exec(session => {
assert(processData(session.attributes("MyData")) == true, "Invalid data");
})
))
Above the scenario somewhere along the line "guardian failed, shutting down system".
Now this seems a useful, often-used thing to do - I'm possibly missing something simple. How to do it?
You have to abide by Gatling APIs.
With checks, you don't "fail" the test, but the request. If you're looking for failing the whole test, you should have a look at the Assertions API and the Jenkins plugin.
You can only perform a Check at the request site, not later. One of the very good reasons is that if you store the bodyString in the Sessions like you're doing, you'll end using a lot of memory and maybe crashing (still referenced, so not garbage collectable). You have to perform your processData in the check, typically in the transform optional step.
were you looking for something like
.exec(http("getRequest")
.get("/request/123")
.headers(headers)
.check(status.is(200))
.check(jsonPath("$.request_id").is("123")))
Since the edit queue is already full.
This is already resolved in the new version of Gatling. Release 3.4.0
They added
exitHereIf
exitHereIf("${myBoolean}")
exitHereIf(session => true)
Make the user exit the scenario from this point if the condition holds. Condition parameter is an Expression[Boolean].
I implemented something using exitHereIfFailed that sounds like exactly what you were trying to accomplish. I normally use this after a virtual user attempts to sign in.
exitHereIfFailed is used this way
val scn = scenario("MyAwesomeScenario")
.exec(http("Get data from endpoint 1")
.get(request1Url)
.check(status.is(200))
.check(bodyString.saveAs("MyData"))
.check(processData(session.attributes("MyData")).is(true)))
.exitHereIfFailed // If we weren't able to get the data, don't continue
.exec(http("Send the data to endpoint 2")
.post(request2Url)
.body(StringBody("${MyData}"))
This scenario will abort gracefully at exitHereIfFailed if any of the checks prior to exitHereIfFailed have failed.

Why afBedSheet don't see my types?

I'm running a BedSheet application but when I make a HTTP request to a particular route, I get this error in the browser
500 - Internal Server Error
afBeanUtils::TypeNotFoundErr
- Could not find match for Type mypod::MyClass.
Available Values
afBedSheet::FileAsset
afBedSheet::HttpStatus
afBedSheet::MethodCall
afBedSheet::Redirect
afBedSheet::Text
sys::File
sys::InStream
afBeanUtils::TypeNotFoundErr : Could not find match for Type mypod::MyClass.
afBeanUtils::TypeLookup.check (TypeLookup.fan:157)
afBeanUtils::TypeLookup.doFindParent (TypeLookup.fan:140)
However MyClass is there and it is being instantiated by other class. Why is not found?
That Err msg could be a little more informative...
It sounds like your request handler is returning an instance of mypod::MyClass and BedSheet is complaining that it doesn't know what to do with it.
You probably want to be returning an instance of afBedSheet::Text as in:
Obj indexPage() {
return Text.fromHtml("<html>Hello!</html>")
}
If the above doesn't sound right, then I'd need to see some code to be of further help.

Routing based on query parameter in Play framework

My web application will be triggered from an external system. It will call one request path of my app, but uses different query parameters for different kinds of requests.
One of the parameters is the "action" that defines what is to be done. The rest of the params depend on the "action".
So I can get request params like these:
action=sayHello&user=Joe
action=newUser&name=Joe&address=xxx
action=resetPassword
...
I would like to be able to encode it similarly in the routes file for play so it does the query param based routing and as much of the validation of other parameters as possible.
What I have instead is one routing for all of these possibilities with plenty of optional parameters. The action processing it starts with a big pattern match to do dispatch and parameter validation.
Googling and checking SO just popped up plenty of samples where the params are encoded in the request path somehow, so multiple paths are routed to the same action, but I would like the opposite: one path routed to different actions.
One of my colleagues said we could have one "dispatcher" action that would just redirect based on the "action" parameter. It would be a bit more structured then the current solution, but it would not eliminate the long list of optional parameters which should be selectively passed to the next action, so I hope one knows an even better solution :-)
BTW the external system that calls my app is developed by another company and I have no influence on this design, so it's not an option to change the way how my app is triggered.
The single dispatcher action is probably the way to go, and you don't need to specify all of your optional parameters in the route. If action is always there then that's the only one you really need.
GET /someRoute controller.dispatcher(action: String)
Then in your action method you can access request.queryString to get any of the other optional parameters.
Note: I am NOT experienced Scala developer, so maybe presented snippets can be optimized... What's important for you they are valid and working.
So...
You don't need to declare every optional param in the routes file. It is great shortcut for type param's validation and best choice would be convince 'other company' to use API prepared by you... Anyway if you haven't such possibility you can also handle their requests as required.
In general: the dispatcher approach seems to be right in this place, fortunately you don't need to declare all optional params in the routes and pass it between actions/methods as they can be fetched directly from request. In PHP it can be compared to $_GET['action'] and in Java version of Play 2 controller - DynamicForm class - form().bindFromRequest.get("action").
Let's say that you have a route:
GET /dispatcher controllers.Application.dispatcher
In that case your dispatcher action (and additional methods) can look like:
def dispatcher = Action { implicit request =>
request.queryString.get("action").flatMap(_.headOption).getOrElse("invalid") match {
case "sayHello" => sayHelloMethod
case "newUser" => newUserMethod
case _ => BadRequest("Action not allowed!")
}
}
// http://localhost:9000/dispatcher?action=sayHello&name=John
def sayHelloMethod(implicit request: RequestHeader) = {
val name = request.queryString.get("name").flatMap(_.headOption).getOrElse("")
Ok("Hello " + name )
}
// http://localhost:9000/dispatcher?action=newUser&name=John+Doe&address=john#doe.com
def newUserMethod(implicit request: RequestHeader) = {
val name = request.queryString.get("name").flatMap(_.headOption).getOrElse("")
val address = request.queryString.get("address").flatMap(_.headOption).getOrElse("")
Ok("We are creating new user " + name + " with address " + address)
}
Of course you will need to validate incoming types and values 'manually', especially when actions will be operating on the DataBase, anyway biggest part of your problem you have resolved now.