Official Python implementation?

[Zearin]

On PyPI (Python’s package index), there are some packages that mention Cucumber or Gherkin, but they all seem to be…

• …out of date
• …make no use of the official Gherkin parser for Python
• …a “clone” or “inspired by” spin-off of the Cucumber project (“NIH” strikes again! Grr…)

Is there no official Python implementation of Cucumber?

If not, I’d like to take a crack at writing one. I’m not a brilliant programmer, but I think I could write a working implementation that uses the JSON output from the existing Python implementation of cucumber/gherkin3.

If I do,

• Would you be willing to adopt it as an official implementation?
• Are there any requirements for the implementation? (E.g, “The package should be structured into these separate modules”, or “It must follow these coding standards”, etc.)

[aslakhellesoy]

Great!!

I'm planning to release microcuke later this week - a Cucumber reference implementation in 500 SLOC.

The main purpose of microcuke is to provide a very simple reference implementation that
can be ported to a new programming language in about a day. Think of it as an aid to
developers who wish to implement Cucumber for a new programming language.

Microcuke is written in classic JavaScript (ES5), because that's a language most
programmers can at least read.

Time permitting I'll also publish an early draft of a spec.

Before we can adopt something as an official implementation we need to make sure there is someone willing to maintain it. Ideally I'd like to have at least 2-3 people dedicating time on it for a few months before blessing it as official.

Hang in there while I clean up microcuke and the spec.

[Zearin]

Great!!

Awesome! :)

(Does it matter that I’m only focusing on Gherkin3? I don’t think I can do the older, more complicated implementation style, where there’s a lot more mixing/coupling between Cucumber and Gherkin…)

I'm planning to release microcuke later this week - a Cucumber reference implementation in 500 SLOC.

The main purpose of microcuke is to provide a very simple reference implementation that
can be ported to a new programming language in about a day. Think of it as an aid to
developers who wish to implement Cucumber for a new programming language.

That is an awesome idea.

For projects that span multiple programming languages, I bet something like this would be a huge benefit to many of them.

Microcuke is written in classic JavaScript (ES5), because that's a language most
programmers can at least read.

Good choice! (I don’t really enjoy JavaScript, but your reasoning is totally solid. :P)

Time permitting I'll also publish an early draft of a spec.

Before we can adopt something as an official implementation we need to make sure there is someone willing to maintain it. Ideally I'd like to have at least 2-3 people dedicating time on it for a few months before blessing it as official.

I can commit to maintain this for roughly a year. Around that time, I’m planning to sell an apartment for the first time, buy a house for the first time, move, and lots of other disruptive stuff. I will almost certainly need to step down at that time.

That okay with you?

Hang in there while I clean up microcuke and the spec.

Okay!

:D

[aslakhellesoy]

Does it matter that I’m only focusing on Gherkin3?

Not at all. We all want to retire Gherkin2.

I bet something like this would be a huge benefit to many of them.

I hope so. Having something tiny, but functional will help build consistent implementations.

That okay with you?

Wow you've got your life planned! :-) Yeah that's not a problem at all as long as there is someone else to build it.

The best way to ensure more than one people are contributing to a project is to:

1. Don't write all the code yourself - actively try to recruit other people who can help
2. TDD is so it's easy for other people to join and modify the code

[Zearin]

I'm planning to release microcuke later this week - a Cucumber reference implementation in 500 SLOC.

How goes it? :)

[aslakhellesoy]

Really well actually! I'll push what I have soon. Still a little rough around the edges.

[Zearin]

Really well actually! I'll push what I have soon. Still a little rough around the edges.

« rubbing hands together in anticipation… »

[aslakhellesoy]

Did you miss the microcuke announcement on the cukes list/twitter/gitter?

What are you waiting for?

[yigithanyilmaz]

As I cmd + F ed I realized nobody mentioned Behave. It is a really identical to cucumber has a good support and community. We have used it in some projects b4 willing to use it again if the client wants to stick to python.

[Zearin]

I’ve been working on a Python implementation based on microcuke in my spare time. I’ve made a great deal of progress, but I’ve had trouble converting a few of the JavaScript-isms to Python.

By design, microcuke is minimalist, and most of it is really easy to understand! :D

However, there are a few places where it is difficult to know what to do for a conversion. Examples:

• EventEmitter is a JS staple, but what should someone using microcuke as a reference do if the target language has no equivalent?
  • Aside: Indeed, Python’s standard library has no built-in equivalent. In my implementation, I’ve been using a naïve 3rd party implementation of the EventEmitter class as a starting point (with plans to move to a more powerful—but trickier to implement—async implementation of EventEmitter for Python 3).
• The TestStep.execute(…) method has a world argument...
  • I don’t know if world is just a requirement of JS’s scoping rules, or something that related to microcuke’s use of Promises.
  • Regarding the comment: We're returning a promise of a boolean here, because JavaScript. Most languages should simply return a boolean and forget about the promise stuff - simpler! That is somewhat helpful, but I still don’t know what the world argument is for—and if (or how) it relates to Promises…
• The tests for TestStep have me scratching my head...
  • Once again, I’m not sure of the role of the world variable here. It seems like it's just filler, and not relevant to the tests themselves. Is this correct?
• Anything glue-related isn’t really working yet. I haven’t yet determined the best syntax for writing glue code in Python yet. (There are a few possibilities; must weigh the pros and cons of each.)

@aslakhellesoy:

• Could you clarify any of the above for me?
• Is there any time in the next couple of weeks when I could chat (video or text) with you about some of JS-isms?

[aslakhellesoy]

This is exciting!

I've been working on a Java port of Microcuke (to serve as a basis for the next Cucumber-JVM, using Gherkin 3) and I've run into the same issues as you ;-). The JavaScript-isms are not trivial to translate to a different language.

I'll create new issues in microcuke to address the points you're bringing up here. It's easier to track these issues individually.

[aslakhellesoy]

Just ping me on https://gitter.im/cucumber/chat to see if I'm around. My work hours are 9am-4pm GMT, and I'm usually busy mobbing before lunch, except Fridays which I try to dedicate fully to open source.

[Zearin]

w00t! It is encouraging to know that I’m not alone! :)

To be clear, I think choosing JavaScript was absolutely the right choice. I think it is just that some JS-isms need to be rewritten generically, for the sake of other implementors.

[aslakhellesoy]

Completely agree about rewriting jsisms!

[Zearin]

Good progress on Python implementation. Stay tuned...

[aslakhellesoy]

Nice!

[webmaven]

Ping.

[Zearin]

@webmaven I’ve been working on the Python implementation. :) (Not ready yet, but it’s close.)

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in a week if no further activity occurs.

[stale]

This issue has been automatically closed because of inactivity. You can support the Cucumber core team on opencollective.

[webmaven]

@Zearin, anything to share?

[Zearin]

Whoa… It’s been a little while. :)

I haven’t been able to work on it in a while, but I think I might be able to pick it up again in the coming months. I would still love to see an official Python implementation.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.