Is there a way to run a scenario multiple times based on the data sets specified in a yaml/json file ?

[ayankumaar]

I want to specify data sets in a yaml file and along with that I would like to specify the specs and scenarios within it to run. The actual data sets will be specified in a csv file. The yaml file will just be used to specify which data set to be picked. For each data set specified I would like to run each scenario once. Is there any way in gauge to achieve this ?

Here is a sample yaml file that I have created for this purpose.

TestInfo.yaml

command:
Creating Sample Spec for Automation Demo(spec): Register a user Automation Demo(scenario)
Search for a string(spec): Search for a string in google(scenario)
dataset:
DS1, DS2, DS3, DS4
---------------------------------------- End of Yaml File---------------------------------------

My requirement is to run scenario 1- Register a user Automation Demo four times (one time each with DS1,DS2,DS3,DS4)
and then run scenario 2- Search for a string in google four times (one time each with DS1,DS2,DS3,DS4)

My configuration is given below:
Gauge version: 1.1.4

Plugins

dotnet (0.1.9)
html-report (4.0.8)
screenshot (0.0.1)

I am using gauge with C#.

I have already looked into the following issues:
#1240 and #1242. From these two I came across the concept of config.yaml file. But, didn't find any documentation or example on that. If there is a way to achieve my requirement using that then I would like to see an example on that.

[zabil]

There is no direct support for this. But you could write a script to parse the YAML to generate csv's and a command or script to run scenarios by name using --scenario flag. I am afraid there's no example at the moment.

[ayankumaar]

@zabil Is there an example for this ? That will help me a lot.

[zabil]

There's no example for this at the moment. A quick google points to a csv reader like https://joshclose.github.io/CsvHelper/ for c# which you could use for reading csv files.

[ayankumaar]

@zabil I am yet to find a solution for my requirement. That is why I am reaching out to you again.

I have written few hooks to trigger execution of my gauge test cases. The hooks get executed in a particular order. For ex: BeforeSuite ->BeforeSpec -> BeforeScenario -> BeforeStep -> AfterStep -> AfterScenario -> AfterSpec -> AfterSuite.
Now, by default these hooks get executed once per execution.

Is there any way to execute these hooks multiple times in the same order ? If there is a way to do this then that will solve my problem.

[zabil]

Is there any way to execute these hooks multiple times in the same order ? If there is a way to do this then that will solve my problem.

I don't think this is possible

[sriv]

Hooks get executed only once per lifecycle, (i.e BeforeSuite/AfterSuite - once per run, BeforeSpec/AfterSpec - once per spec etc). However, you can have multiple hooks, ex: you can have two BeforeSpec hooks.

There is no guarantee on the order of hooks, but you can use tags to filter them out for certain cases.

In your situation, a hack would be to extract hook logic to a common method, and call that in multiple marker methods with the right annotation. ex:

private void commonHookStuff() {
}

@BeforeSpec
public void FirstHook() {
    commonHookStuff();
}

@BeforeSpec
public void SecondHook() {
    commonHookStuff();
}

But do note that in your situation, the csv files need to be present before gauge attempts it's execution lifecycle. Hence you cannot use an execution hook to populate data that would be used by gauge to parse the spec. It is important to note that Gauge's lifecycle looks roughly like this: parse -> validate -> execute. The hooks sit in execute phase, whereas the CSV files are read in the parse phase.