How to run a collection against a CSV of variables?

[satisfice]

I want to do data-driven API testing. To do this I want to target a particular request or collection and give it a CSV of variables to iterate through. Performing one request for each line in the CSV and collecting the results or performing assertions as needed.

I can do this via the CLI, I suppose, by driving the CLI with Python. But if I do that I might as well do everything with Python. I was hoping there is a way to do this in the Bruno GUI but I'm not seeing anything there or in the documentation about it.

[pove]

Big one. With that we will attract not only Postman/Insomnia users but also JMeter ones

[satisfice]

As a tester, this is the most obvious thing I ever want to do with an API: run a route through a zillion variations. And with such a feature, a completely non-technical tester can focus on curating a nice CSV in Excel or Google Sheets. Postman can do this, of course, but it does it in a weird and clunky way.

[christopher-nash]

My usecase is slightly different, but the need is basically the same. This is something I end up doing about once every other month and is one of the main features I look for in an API tool like Postman or Bruno.

I store collections of admin endpoints to our services. Sometimes a bug will be introduced somewhere and I want to do some sort of build operation against multiple IDs - maybe a bulk data fix up or restarting a bunch of workflows with specific ids. I can identify the ones I want to interact with and export them as a CSV. I then want to run a single request for each line in the CSV (exactly as described above).

And same sort of thing goes for me - I could run things via JMeter or a Jupyter notebook with Python or something, but having it built into the tool was SO nice with postman. Its one of the things holding me back from switching over to Bruno.

[csjohnstone]

I use running a series of collection requests to show how our API works passing in the field values from a CSV then use tests to store response values to run the next in a series of steps. We are a API first commerce cloud platform and use Postman and CSV files to load test environments with products, customers, run complex checkout scenarios, etc. Being able to pass in data via a CSV makes this a breeze in PM.

[kjackey1]

Definitely in the above group, before we can settle on Bruno as our one stop API testing toolset we need the ability to have a data driver setup.

[hakim-adamik]

Same here, I have data-driven testing as a requirement, and I'm currently facing two roadblocks:

• I need to write and maintain the data as JSON, not CSV. Is there a reason why Bruno's GUI handles both JSON and CSV, but the CLI only handles CSV?... Would be great to just add the same option to the CLI 🙏 this is currently the only thing preventing me from using Bruno for test automation of my whole codebase.

• Each of the endpoints of my API takes a different body structure, so with a dataset being only usable on a collection, this means I have to define each of my endpoints as a single collection, which is pretty cumbersome. It would be a lot more convenient that a dataset can be used on a single request. It seems to me this is a pretty standard use-case, as different endpoints rarely use the same body structure?

[rreyn-bruno]

Hey @hakim-adamik - as a partial update, we released support for using JSON data for a collection run via the CLI in v1.35.0.

[rreyn-bruno]

@satisfice - I realize it's been quite some time since you opened this discussion, but you are indeed able to use a CSV (or JSON file) inside of the GUI or the CLI to iterate through data.