I am new to Scala and Gatling.
I am trying to write framework for Load and performance testing using Gatling API in Scala for REST API endpoints.
I have a query regarding one of the code snippet which is supposed to generate signature(calling another method) and save the value in the session.
.exec(session => {
session.set("sign", SignatureGeneration.getSignature(key, ElFileBody("abc.json").toString()))
})
abc.json -
{"device": "${device}"}
In above code getSignature takes arguments (String, String). I want to read the json file and replace ${} value in it with the feeders and send it as String to the method.
While debugging the code I found out, ElFileBody send object as <function1> and not the json content of it.
Solution -
val bodyExpr = ElFileBody("abc.json")
val bodyStr = bodyExpr(session).get
Related
I'm currently trying to mock external server using Wiremock.
One of my external server endpoint takes a payload.
This endpoint is defined as follow :
def sendRequestToMockServer(payload: String) = {
for {
request_entity <- Marshal(payload).to[RequestEntity]
response <- Http().singleRequest(
HttpRequest(
method = HttpMethods.GET,
uri = "http://localhost:9090/login",
entity = request_entity
)
)
} yield {
response
}
}
To mock this endpoint using Wiremock, I have written the following code :
stubFor(
get(urlEqualTo("/login"))
.willReturn(
aResponse()
.withHeader("Content-Type","application/json")
.withBodyFile("wireMockResponse.json")
.withStatus(200)
)
.withRequestBody(matchingJsonPath("requestBody.json"))
)
where I Have defined the request body in the requestBody.json file.
But when I run tests , I keep getting an error indicating that the requested Url is not found.
I'm thinking that the error is related to this line withRequestBody(matchingJsonPath("requestBody.json")), because when I comment it the error disappear.
Any suggestions on how to work around this?
matchingJsonPath does not populate a file at a provided filepath, but instead evaluates the JsonPath provided. See documentation.
I'm not entirely sure there is a way to provide the request body as a .json file. If you copy the contents of the file into the withRequest(equalToJson(_yourJsonHere_)), does it work? If it does, you could get the file contents as a JSON string above the definition and provide it to the function (or I guess, make a function to return a JSON string from a .json file).
Additionally, you could make a custom request matcher that does the parsing for you. I think I'd recommend this only if the above does not work.
Our test automation needs to interact with kafka and we are looking at how we can achieve this with karate.
We have a java class which reads from kafka and puts records in an internal list. We then ask for these records from karate, filter out all messages from background traffic, and return the first message that matches our filter.
So our consumer looks like this (simplified):
// consume.js
function(bootstrapServers, topic, filter, timeout, interval) {
var KafkaLib = Java.type('kafka.KafkaLib')
var records = KafkaLib.getRecords(bootstrapServers, topic)
for (record_id in records) {
// TODO here we want to convert record to a json (and later xml for xml records) so that
// we can access them as 'native' karate data types and use notation like: cat.cat.scores.score[1]
var record = records[record_id]
if (filter(record)) {
karate.log("Record matched: " + record)
return record
}
}
throw "No records found matching the filter: " + filter
}
Records can be json, xml, or plain text, but looking in the json case now.
In this case given that in kafka there is a message like this:
{"correlationId":"b3e6bbc7-e5a6-4b2a-a8f9-a0ddf435de67","text":"Hello world"}
This is loaded as a string in the record variable above.
We want to convert this to json so that a filter like this would work:
* def uuid = java.util.UUID.randomUUID() + ''
# This is what we are publishing to kafka
* def payload = ({ correlationId: uuid, text: "Hello world" })
* def filter = function(m) { return m.correlationId == uuid }
Is there a way to convert a string to a native karate variable in javascript? Might have missed it looking at https://intuit.github.io/karate/#the-karate-object. By the way var jsonRecord = karate.toJson(record) did not work and jsonRecord.uuid was undefined.
Edit: I have made an example of what I am trying to achieve here:
https://github.com/KostasKgr/karate-issues/blob/java_json_interop/src/test/java/examples/consumption/consumption.feature
Many thanks
Sometime ago I had put together a something that could be used to test Kafka from within Karate. Pls see if https://github.com/Sdaas/karate-kafka helps. Happy to enhance / improve if it helps you.
Can you try,
* json payload = { correlationId: uuid, text: "Hello world" }
ref : Type Conversion
for type conversion within javascript ideally karate.toMap(object) or karate.toJson(object) should.
rather than wrapping up everything into one JS function, I would suggest keeping the record invoking part outside the JS and let karate cast it.
* json records = Java.type('kafka.KafkaLib').getRecords(bootstrapServers, topic)
* consume(records, filter, timeout, interval)
As mentioned in the comments of another answer, there is now an enhancement ticket on karate to achieve what was discussed in this thread, see https://github.com/intuit/karate/issues/1202
Until that is in place, I managed to get most of what I wanted concerning JSON by parsing string to json in Java and returning that to karate.
Map<String,Object> result = new ObjectMapper().readValue(record, HashMap.class);
Not sure if the same can be worked around for xml
You can see the workaround in action here:
https://github.com/KostasKgr/karate-issues/blob/java_json_interop_v2/src/test/java/examples/consumption/consumption.feature
Because of Karate's support for Java inter-op you can easily write some "glue" code to connect your existing Kafka systems to Karate test-suites, see the first link below.
Here are a few references:
how to use Java inter-op to listen and wait for events: https://twitter.com/KarateDSL/status/1417023536082812935
the Karate ActiveMQ example: https://github.com/intuit/karate/tree/master/karate-netty#consumer-provider-example
Walmart Labs blog post (Kafka specific): https://medium.com/walmartglobaltech/kafka-automation-using-karate-6a129cfdc210
Karate Kafka (3rd party project / example): https://github.com/Sdaas/karate-kafka
I have a pretty simple scenario to get a random value from a CSV file that contains only one column with header id. Then, I want to paste this value into a body of a POST request. An example of the body of my request looks as follows:
{ "id" : "123" }. So, instead of "123", I need to use a value from a CSV file. Here are my feeder and scenario:
val csvFeeder = csv("src/test/resources/ID.csv").random
val scn = scenario("Test POST request")
.feed(csvFeeder)
.exec(http("Test POST request")
.post(uri)
.header("Content-type", "application/json")
.body(StringBody("{\"id\":\"$id\"}"))
.check(status is 200, responseTimeInMillis lte 2000)
)
This example is the closest implementation I managed to make so far. But it doesn't work as I expect. I can't paste the value from a feeder into $id that is inside a body(). Currently, Gatling sends the following body: {"id":"$id"}
How can I retrieve a value from a feeder and paste it into a body string request?
Found the solution. The code is right. The issue was in the version of artifact I was using. I used version 3.0.0-RC3 of gatling-charts-highcharts, gatling-core & gatling-http and gatling-maven-plugin was 3.0.0. Now I switched to 2.3.1 for gatling-* dependencies and used 2.2.4 for maven plugin. Now it works.
I am building a webservice request that submits some data to a 3rd party service we use:
val promise = WS.url("http://example.com/api/xxx/testers?api_key=%s" format(prefineryAPIKey)).post(requestBody)
val data = promise.value.get.body
The variable requestBody is XML content and I need to send this post with "Content-Type: text/xml". How do I set that inside the webservice? Does it work like Play!'s result by chaining .as("text/xml")?
Thanks
Found it! I had to dig around the API docs and decrypt some of the compiler errors, but basically the above will look like:
val promise = WS
.url("http://example.com/api/xxx/testers?api_key=%s" format(prefineryAPIKey))
.withHeaders("Content-Type" -> "text/xml")
.post(requestBody)
When you call WS you are putting together a WSRequestHolder. The docs for WSRequest are here:
http://www.playframework.org/documentation/api/2.0/scala/index.html#play.api.libs.ws.WS$$WSRequestHolder
i am confused on how to combine the json library in dispatch and lift to parse my json response.
I am apparently a scala newbie.
I have written this code :
val status = {
val httpPackage = http(Status(screenName).timeline)
val json1 = httpPackage
json1
}
Now i am stuck on how to parse the twitter json response
I've tried to use the JsonParser:
val status1 = JsonParser.parse(status)
but got this error:
<console>:38: error: overloaded method value parse with alternatives:
(s: java.io.Reader)net.liftweb.json.JsonAST.JValue<and>
(s: String)net.liftweb.json.JsonAST.JValue
cannot be applied to (http.HttpPackage[List[dispatch.json.JsObject]])
val status1 = JsonParser.parse(status1)
I unsure and can't figure out what to do next in order to iterate through the data, extract it and render it to my web page.
Here's another way to use Dispatch HTTP with Lift-JSON. This example fetches JSON document from google, parses all "titles" from it and prints them.
import dispatch._
import net.liftweb.json.JsonParser
import net.liftweb.json.JsonAST._
object App extends Application {
val http = new Http
val req = :/("www.google.com") / "base" / "feeds" / "snippets" <<? Map("bq" -> "scala", "alt" -> "json")
val json = http(req >- JsonParser.parse)
val titles = for {
JField("title", title) <- json
JField("$t", JString(name)) <- title
} yield name
titles.foreach(println)
}
The error that you are getting back is letting your know that the type of status is neither a String or java.io.Reader. Instead, what you have is a List of already parsed JSON responses as Dispatch has already done all of the hard work in parsing the response into a JSON response. Dispatch has a very compact syntax which is nice when you are used to it but it can be very obtuse initially, especially when you are first approaching Scala. Often times, you'll find that you have to dive into the source code of the library when you are first learning to see what is going on. For instance, if you look into the dispatch-twitter source code, you can see that the timeline method actually performs a JSON extraction on the response:
def timeline = this ># (list ! obj)
What this method is defining is a Dispatch Handler which converts the Response object into a JsonResponse object, and then parses the response into a list of JSON Objects. That's quite a bit going on in one line. You can see the definition for the operand ># in the JsHttp.scala file in the http+json Dispatch module. Dispatch defines lots of Handlers that do a conversion behind the scenes into different types of data which you can then pass to block to work with. Check out the StdOut Walkthrough and the Common Tasks pages for some of the handlers but you'll need to dive into the various modules source code or Scaladoc to see what else is there.
All of this is a long way to get to what you want, which I believe is essentially this:
val statuses = http(Status(screenName).timeline)
statuses.map(Status.text).foreach(println _)
Only instead of doing a println, you can push it out to your web page in whatever way you want. Check out the Status object for some of the various pre-built extractors to pull information out of the status response.