I am trying to write a Gatling script where I read a starting number from a CSV file and loop through, say 10 times. In each iteration, I want to increment the value of the parameter.
It looks like some Scala or Java math is needed but could not find information on how to do it or how and where to combine Gatling EL with Scala or Java.
Appreciate any help or direction.
var numloop = new java.util.concurrent.atomic.AtomicInteger(0)
val scn = scenario("Scenario Name")
.asLongAs(_=> numloop.getAndIncrement() <3, exitASAP = false){
feed(csv("ids.csv")) //read ${ID} from the file
.exec(http("request")
.get("""http://finance.yahoo.com/q?s=${ID}""")
.headers(headers_1))
.pause(284 milliseconds)
//How to increment ID for the next iteration and pass in the .get method?
}
You copy-pasted this code from Gatling's Google Group but this use case was very specific.
Did you first properly read the documentation regarding loops? What's your use case and how doesn't it fit with basic loops?
Edit: So the question is: how do I get a unique id per loop iteration and per virtual user?
You can compute one for the loop index and a virtual user id. Session already has a unique ID but it's a String UUID, so it's not very convenient for what you want to do.
// first, let's build a Feeder that set an numeric id:
val userIdFeeder = Iterator.from(0).map(i => Map("userId" -> i))
val iterations = 1000
// set this userId to every virtual user
feed(userIdFeeder)
// loop and define the loop index
.repeat(iterations, "index") {
// set an new attribute named "id"
exec{ session =>
val userId = session("userId").as[Int]
val index = session("index").as[Int]
val id = iterations * userId + index
session.set("id", id)
}
// use id attribute, for example with EL ${id}
}
Here is my answer to this:
Problem Statement: I had to repeat the gatling execution for configured set of times, and my step name has to be dynamic.
object UrlVerifier {
val count = new java.util.concurrent.atomic.AtomicInteger(0)
val baseUrl = Params.applicationBaseUrl
val accessUrl = repeat(Params.noOfPagesToBeVisited,"index") {
exec(session=> {
val randomUrls: List[String] = UrlFeeder.getUrlsToBeTested()
session.set("index", count.getAndIncrement).set("pageToTest", randomUrls(session("index").as[Int]))
}
).
exec(http("Accessing Page ${pageToTest}")
.get(baseUrl+"${pageToTest}")
.check(status.is(200))).pause(Params.timeToPauseInSeconds)
}
So basically UrlFeeder give me list of String (urls to be tested) and in the exec, we are using count (AtomicInteger), and using this we are populating a variable named 'index' whose value will start from 0 and will be getAndIncremented in each iteration. This 'index' variable is the one which will be used within repeat() loop as we are specifying the name of counterVariable to be used as 'index'
Hope it helps others as well.
Related
I am using Gatling 3.6.1 and I am trying to repeat the request 10 times for the next 10 products from the feeder file. This is what I tried:
feed(products, 10)
.repeat(10, "index") {
exec(session => {
val index = session("index").as[Int]
val counter = index + 1
session.set("counter", counter)
})
.exec(productIdsRequest())
}
private def productIdsRequest() = {
http("ProductId${counter}")
.get(path + "products/${product_code${counter}}")
.check(jsonPath("$..code").count.gt(2))
}
I am having trouble getting the counter value to my API URL.
I would like to have something like
products/${product_code1},
products/${product_code2} etc.
But instead, I get the error 'nested attribute definition is not allowed'
So basically I would like that every request gets called with one product from the feeder (in the batch of 10 products)
Can you please help?
Thanks!
Disclaimer: I don't know how realized your feeder products.
If I clearly understand you - just need to move .repeat on high level:
.repeat(10, "counter") {
feed(product)
.exec(http("ProductId ${counter}")
.get("products/${product_code}")
.check(jsonPath("$..code").count.gt(2)))
}
I am trying to send a REST API call to retrieve a lot of data. Now this data is returned in JSON format and is limited to 2000 records each call. However, if there are more than 2000 records then there is a key called nextRecordsUrl with a link to an endPoint with the next 2000 records.
And this pattern continues until there are less than 2000 records in a single call and then nextRecordsUrl is undefined.
I have this loop that essentially pushes the data to an array and then calls the endpoint listed in the nextRecordsUrl key.
do {
for (var i in arrLeads.records) {
let data = arrLeads.records[i];
let createDate = new GMT(data.CreatedDate, "dd-MM-YYYY");
let fAssocDate = new GMT(data['First_Campaign_assoc_date__c'],"dd-MM-YYYY");
let lAssocDate = new GMT(data['Last_Campaign_assoc_date__c'], "dd-MM-YYYY");
let convDate = new GMT(data.ConvertedDate, "dd-MM-YYYY");
leads.push([data.FirstName, data.LastName, data.Id, data.Email, data.Title, data['hs_contact_id__c'], createDate, data.Company, data.OwnerId, data.Country, data['Region_by_manager__c'], data.Status, data.Industry, data['Reason_for_Disqualifying_Nurture__c'], data['Reason_for_Disqualifying_Dead__c'], data['Lead_Type__c'], data['First_Campaign_Name__c'], data['First_CampaignType__c'], fAssocDate, data['Last_Campaign_Name__c'], lAssocDate, data['Last_Campaign_Type__c'], convDate, data.ConvertedAccountId, data.ConvertedContactId, data.ConvertedOpportunityId]);
}
var arrLeads = getQueryWithFullEndPoint(arrLeads.nextRecordsUrl);
} while (arrLeads.nextRecordsUrl != null && arrLeads.nextRecordsUrl != undefined);
I used to use the regular while loop, but it caused obvious problems in that it wouldn't run at all if the initial call had an empty nextRecordsUrl field.
But this also has an issue. While the first iteration works well, the last one does not because it makes the call on the next iteration before the loop checks the nextRecordsUrl field.
So essentially, it will loop through all the records normally. But when it gets to the last one, it has already run the last endPoint and will break the loop, because now the 'nextRecordsUrl key is empty. So the last iteration will be ignored.
I thought of moving the call to after the check, right after the do {, but that will cause problems for the first iteration.
I also thought about duplicating the for loop after the do, but I prefer a cleaner solution that doesn't involve duplicating code.
So how do I write this code in a way that takes into account the first iteration, even if there no second iteration and the last iteration, where it has records but has an empty nextRecordsUrl, without having to double up the code?
I actually managed to solve this with a small tweak in the code.
My issue is that I getting the next batch before I could test the original, and so I was always 1 batch short.
The solutions was actually to define a variable at the beginning of each loop arrLeadsNext = arrLeads.nextRecordsUrl;
and have the while statement check this variable.
So it looks like this:
do {
var arrLeadsNext = arrLeads.nextRecordsUrl;
for (var i in arrLeads.records) {
let data = arrLeads.records[i];
let createDate = new GMT(data.CreatedDate, "dd-MM-YYYY").process();
let fAssocDate = new GMT(data['First_Campaign_assoc_date__c'], "dd-MM-YYYY").process();
let lAssocDate = new GMT(data['Last_Campaign_assoc_date__c'], "dd-MM-YYYY").process();
let convDate = new GMT(data.ConvertedDate, "dd-MM-YYYY").process();
leads.push([data.FirstName, data.LastName, data.Id, data.Email, data.Title, data['hs_contact_id__c'], createDate, data.Company, data.OwnerId, data.Country, data['Region_by_manager__c'], data.Status, data.Industry, data['Reason_for_Disqualifying_Nurture__c'], data['Reason_for_Disqualifying_Dead__c'], data['Lead_Type__c'], data['First_Campaign_Name__c'], data['First_CampaignType__c'], fAssocDate, data['Last_Campaign_Name__c'], lAssocDate, data['Last_Campaign_Type__c'], convDate, data.ConvertedAccountId, data.ConvertedContactId, data.ConvertedOpportunityId]);
}
var arrLeads = (arrLeadsNext != null) ? getQueryWithFullEndPoint(arrLeadsNext) : '';
} while (arrLeadsNext != null && arrLeadsNext != undefined);
So the check at the end hasn't been changed by the next API call for the next batch as it will only change once the next iteration of the loop begins.
I have a set of data in a Google spreadsheet in two columns. One column is a list of article titles and the other is the ID of a hotel that is in that article. Call it list1.
Example data
I would like returned a new list with article titles in one column, and an array of the hotel IDs in that article in the other column. Call it list2.
Example data
There are thousands of lines that this needs to be done for, and so my hope was to use Google Apps Script to help perform this task. My original thinking was to
Create column 1 of list2 which has the unique article titles (no script here, just the G-sheets =unique() formula.
Iterate through the titles in list2, looking for a match in first column of the list1
If there is a match:
retrieve its corresponding value in column 2
push it to an empty array in column two of list2
move onto next row in list1
if no longer a match, loop back to step 2.
I've written the following code. I am currently getting a type error (TypeError: Cannot read property '0' of undefined (line 13, file "Code")), however, I wanted to ask whether this is even a valid approach to the problem?
function getHotelIds() {
var outputSheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('list2');
var lastRow = outputSheet.getLastRow();
var data = outputSheet.getRange(2,1,lastRow,2).getValues();
var workingSheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('list1');
var lastActiveRow = workingSheet.getLastRow();
var itemIDS = [];
for (var i=1; i<=data.length; i++) {
var currentArticle = data[i][0];
var lookupArticle = workingSheet[i][0];
if (currentArticle === lookupArticle) {
var tempValue = [workingSheet[i][1]];
itemIDS.push(tempValue);
}
}
}
Use a simple google sheets formula:
You can use a very simple formula to achieve your goal instead of using long and complicated scripts.
Use =unique(list1!A2:A) in cell A2 of list2 sheet to get the unique hotels.
and then use this formula to all the unique hotels by dragging it down in column B.
=JOIN(",",filter(list1!B:B,list1!A:A=A2))
You got the idea right, but the logic needed some tweaking. The "undefined" error is caused by the workingSheet[i][0]. WorkingSheet is a Sheet object, not an array of data. Also, is not necessary to get the data from list2 (output), it is rather the opposite. You have to get the data from the list1 (source) sheet instead, and iterate over it.
I added a new variable, oldHotel, which will be used to compare each line with the current hotel. If it's different, it means we have reached a different Hotel and the data should be written in list2.
function getHotelIds() {
var outputSheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('list2');
var outLastRow = outputSheet.getLastRow();
var workingSheet = SpreadsheetApp.getActiveSpreadsheet().getSheetByName('list1');
var lastActiveRow = workingSheet.getLastRow();
var sourceValues = workingSheet.getRange("A2:B" + lastActiveRow).getValues();
var itemIDS = [];
var oldHotel = sourceValues[0][0]; //first hotel of the list
for (var i = 0; i < sourceValues.length; i++) {
if (sourceValues[i][0] == oldHotel) {
itemIDS.push(sourceValues[i][1]);
/*When we reach the end of the list, the oldHotel variable will never be different. So the next if condition is needed. Otherwise it wouldn't write down the last Hotel.
*/
if (i == sourceValues.length - 1) {
outputSheet.getRange(outLastRow + 1, 1, 1, 2).setValues([
[sourceValues[i][0], itemIDS.toString()]
]);
}
} else {
outputSheet.getRange(outLastRow + 1, 1, 1, 2).setValues([
[sourceValues[i - 1][0], itemIDS.toString()]
]);
oldHotel = sourceValues[i][0]; //new Hotel will be compared
outLastRow = outputSheet.getLastRow(); //lastrow has updated
itemIDS = []; //clears the array to include the next codes
}
}
}
I also converted the itemIDS array to a String each time, so it's written down in a single cell without issues.
Make sure each column of the Sheet is set to "Plain text" from Format > Number > Plain Text
References
getRange
setValues
toString()
I am using a postman to automate apis.
Now I am using following request , lets say :-
{
"customerId": "{{currentClientId}}"
}
Where clientid is a dynamic variable whose value is substituted dynamically as 1 , 2, 3,4 so on..
I call this request multiple times using setNextRequest call in this eg lets say 10.This is being done using a counter variable. I am initialising the counter in my previous request to 0 and using for loop with value as counter as 10 calling the request 10 times.There is no response in body just successful http code 204.
I want to store all these clientids coming in request into an environment Client array variable so I wrote a following pre-request script:-
counter = pm.environment.get("counter");
ClientArray = pm.environment.get("ClientArray");
ClientArray.push(pm.environment.get("currentClientId"));
pm.environment.set("ClientArray",ClientArray);
In Test Script, wrote following code :-
counter = pm.environment.get("counter");
if(counter<=10) {
console.log("hi");
postman.setNextRequest("Request");
counter++;
pm.environment.set("counter",counter);
console.log("Counter",counter);
}
The above scipts is throwing
TypeError | ClientArray.push is not a function.
Could someone please advice how to achieve this.
When you retrieve a value from an environment variable like you're doing:
ClientArray = pm.environment.get("ClientArray");
You're not getting an array, you're getting a string which is why you're getting that error. You need to treat the variable like a string, append the currentClientId much like you do for the counter. Something like:
var currentClientIds = pm.environment.get("ClientArray");
currentClientIds = currentClientIds + "," + currentClientId
When you're done appending i.e. out of your loop simply take the string and convert it to an array:
var currentClientIds = pm.environment.get("ClientArray");
var idsArr = curentClientIds.split(',');
I have two arrays:
GlobalArray:Array(Int,Array[String]) and SpecificArray:Array(Int,Int).
The first Int in both of them is a key and I would like to get the element corresponding to that key from the GlobalArray.
In pseudocode:
val v1
For each element of SpecificArray
Get the corresponding element from GlobalArray to use its Array[String]
If (sum <= 100)
for each String of the Array
update v1
// ... some calculation
sum += 1
println (v1)
I know using .map() I could go through each position of the SpecificArray, but so far I was able to do this:
SpecificArray.map{x => val in_global = GlobalArray.filter(e => (e._1 == x._1))
// I don't know how to follow
}
How about something like below, I would prefer to for comprehension code which has better readability.
var sum:Int = 0
var result:String = ""
for {
(k1,v1) <- SpecificArray //v1 is the second int from specific array
(k2,values) <- GlobalArray if k1 == k2 //values is the second array from global array
value <- values if sum < 100 //value is one value frome the values
_ = {sum+=1; result += s"(${v1}=${value})"} //Update something here with the v1 and value
} yield ()
println(result)
Note needs more optimization
Convert GlobalArray to Map for faster lookup.
val GlobalMap = GlobalArray.toMap
SpecificArray.flatMap(x => GlobalMap(x._1))
.foldLeft(0)((sum:Int, s:String) => {
if(sum<=100) {
// update v1
// some calculation
}
sum+1
})
If not all keys of SpecificArray is present in GlobalMap then use GlobalMap.getOrElse(x._1, Array())
How sum affects the logic and what exactly is v1 is not clear from your code, but it looks like you do search through GlobalArray many times. If this is so, it makes sense to convert this array into a more search-friendly data structure: Map. You can do it like this
val globalMap = GlobalArray.toMap
and then you may use to join the strings like this
println(SpecificArray.flatMap({case (k,v) => globalMap(k).map(s => (k,v,s))}).toList)
If all you need is strings you may use just
println(SpecificArray.flatMap({case (k,v) => globalMap(k)}).toList)
Note that this code assumes that for every key in the SpecificArray there will be a matching key in the GlobalArray. If this is not the case, you should use some other method to access the Map like getOrElse:
println(SpecificArray.flatMap({case (k,v) => globalMap.getOrElse(k, Array()).map(s => (k,v,s))}).toList)
Update
If sum is actually count and it works for whole "joined" data rather than for each key in the SpecificArray, you may use take instead of it. Code would go like this:
val joined = SpecificArray2.flatMap({case (k,v) => globalMap.getOrElse(k, Array()).map(s => (s,v))})
.take(100) // use take instead of sum
And then you may use joined whatever way you want. And updated demo that builds v1 as joined string of form v1 += String_of_GlobalArray + " = " + 2nd_Int_of_SpecificArray is here. The idea is to use mkString instead of explicit variable update.