[devkit] Improved undefined example(s) by luke-hill · Pull Request #278 · cucumber/compatibility-kit

luke-hill · 2026-06-01T21:10:29Z

🤔 What's changed?

Update some of the undefined examples to include multi-snippet functionality, so as to prove the cucumber implementor is working well when the snippet generation is > 1 per scenario

⚡️ What's your motivation?

See above

🏷️ What kind of change is this?

🏦 Refactoring/debt/DX (improvement to code design, tooling, etc. without changing behaviour)
💥 Breaking change (incompatible changes to the API)

♻️ Anything particular you want feedback on?

📋 Checklist:

I agree to respect and uphold the Cucumber Community Code of Conduct
I've changed the behaviour of the code
- I have added/updated tests to cover my changes.
My change requires a change to the documentation.
- I have updated the documentation accordingly.
Users should know about my change
- I have added an entry to the "Unreleased" section of the CHANGELOG, linking to this pull request.

This text was originally generated from a template, then edited by hand. You can modify the template here.

mpkorstanje · 2026-06-01T21:21:42Z

I think it would be better if the existing tests didn't change and one undefined-multiple was added instead. It allows for a more gradual implementation of functionality and won't break any existing tests.

luke-hill · 2026-06-01T21:23:30Z

@davidjgoss I think there's a bug here in fake-cucumber. It looks like when we have multiple steps that are undefined that whilst the snippet generation is working, the status of the subsequent steps is being reported as UNDEFINED. But I believe they should be SKIPPED

This doesn't affect the purpose of the tests here per se. But I think it's still wrong as/when I come to implement these once I'm at cckv30+

luke-hill · 2026-06-01T21:24:57Z

closes #276

luke-hill · 2026-06-01T22:53:35Z

I think it would be better if the existing tests didn't change and one undefined-multiple was added instead. It allows for a more gradual implementation of functionality and won't break any existing tests.

Will make this change in a couple of days. Notwithstanding the fake-cucumber issue as I think this is a bug here

davidjgoss · 2026-06-02T19:45:44Z

@davidjgoss I think there's a bug here in fake-cucumber. It looks like when we have multiple steps that are undefined that whilst the snippet generation is working, the status of the subsequent steps is being reported as UNDEFINED. But I believe they should be SKIPPED

That was an intentional change in #239, and IIRC it was pointed out by @brasmusson that it works that way already in cucumber-ruby.

luke-hill · 2026-06-03T09:00:49Z

I think at that time cucumber-ruby was somewhere between cckv20 and v22 compliant. It's currently cckv24 compliant and I'm working on the v26 compliance.

Are we saying that the current implementation is correct, or that we need to make a change. I'm a bit confused sorry

davidjgoss · 2026-06-03T14:10:52Z

Are we saying that the current implementation is correct, or that we need to make a change. I'm a bit confused sorry

The current CCK behaviour is correct.

…ble and one for singular

… that are longer

…ped then undefined again (It should stay skipped)

mpkorstanje · 2026-06-04T00:00:12Z

+
+  Scenario: Steps before undefined steps are executed
+    Given an implemented step
+    And an implemented step


I think we can remove this line. I don't think it meaningfully tests anything related to undefined steps.

mpkorstanje · 2026-06-04T00:03:06Z

@@ -0,0 +1,18 @@
+Feature: Examples Tables - With Many Undefined Steps


I don't think this case is needed at all. Example tables are sugar for repeating the same scenario, which is already covered by the undefined-multiple scenario.

This was a migration from something left out on the initial implementation

mpkorstanje · 2026-06-04T00:04:11Z

Is this rename needed?

Not strictly, but it was a bug which was picked up (or will be picked up), so figure in a breaking change (Which this will be), adding this is fine.

mpkorstanje · 2026-06-04T00:07:06Z

+    Given an implemented step
+    And an implemented step
+    And a step that is yet to be defined
+    And another step that is also yet to be defined


This line looks like we're testing multiple scenarios in on scenario.

We can have two scenarios here, one with one step after an undefined step and one with multiple.

It's also not quite clear what we're testing here.

Either a second step after an undefined step will be skipped (not undefined) or a second undefined step after an undefined step is also undefined (not skipped). It should be clear from reading the scenario what will happen.

Agreed. Will refactor this

luke-hill

Not sure when I'll get to this. But put some responses and some things I need to fix (Notes to self).

luke-hill · 2026-06-10T09:57:19Z

@@ -0,0 +1,18 @@
+Feature: Examples Tables - With Many Undefined Steps


This was a migration from something left out on the initial implementation

luke-hill · 2026-06-10T09:57:47Z

Not strictly, but it was a bug which was picked up (or will be picked up), so figure in a breaking change (Which this will be), adding this is fine.

luke-hill · 2026-06-10T09:58:36Z

+    Given an implemented step
+    And an implemented step
+    And a step that is yet to be defined
+    And another step that is also yet to be defined


Agreed. Will refactor this

luke-hill · 2026-06-10T09:58:43Z

+
+  Scenario: Steps before undefined steps are executed
+    Given an implemented step
+    And an implemented step


luke-hill added 2 commits June 1, 2026 22:08

Add in new scenarios with multi-snippet functionality

ddfb4fb

Add in new generation ndjson

d5b3b40

luke-hill added 4 commits June 3, 2026 18:46

Convert examples tables undefined into 2 split ones - one for enumera…

2273e14

…ble and one for singular

Add in undefined enumerable feature and split out undefined scenarios…

ed192af

… that are longer

Add in new indeterminate scenario that flip-flops for undefined, skip…

0e0f8af

…ped then undefined again (It should stay skipped)

Regenerate all NDJSON files

3010ca2

luke-hill commented Jun 3, 2026

View reviewed changes

Comment thread devkit/samples/undefined-enumerable/undefined-enumerable.feature Outdated

luke-hill added 2 commits June 3, 2026 18:59

Fix naming and scenario def

07d3b39

Regenerate NDJSON for new names

9324356

mpkorstanje reviewed Jun 4, 2026

View reviewed changes

luke-hill commented Jun 10, 2026

View reviewed changes

		@@ -0,0 +1,18 @@
		Feature: Examples Tables - With Many Undefined Steps

Uh oh!

Conversation

luke-hill commented Jun 1, 2026

🤔 What's changed?

⚡️ What's your motivation?

🏷️ What kind of change is this?

♻️ Anything particular you want feedback on?

📋 Checklist:

Uh oh!

mpkorstanje commented Jun 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

luke-hill commented Jun 1, 2026

Uh oh!

luke-hill commented Jun 1, 2026

Uh oh!

luke-hill commented Jun 1, 2026

Uh oh!

davidjgoss commented Jun 2, 2026

Uh oh!

luke-hill commented Jun 3, 2026

Uh oh!

davidjgoss commented Jun 3, 2026

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

luke-hill left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

mpkorstanje commented Jun 1, 2026 •

edited

Loading