Feature/Cucumber Messages #233
Conversation
|
FYI, github Build is failing b/c one produced component (FileSinkPlugin) gets published only to my local Nuget store. It can then not be found by the build pipeline in github. |
|
First end-to-end working scenario is operational.
Next:
|
clrudolphi
force-pushed
the
feature-CucumberMessages
branch
from
364e466
to
1f2f1e6
Compare
|
Build is failing b/c I have a private build of Cucumber Messages nuget package. When the next release of that comes out, I will patch up the nuget references in this branch. |
This is fine. We have feed https://www.myget.org/feed/Packages/reqnroll-unstable where we could publish interim releases (of dependencies as well) and I think this feed is added to the nuget sources, so the build would be able to resolve it though them. But I never tested this model yet. |
…zing events as Cucumber Messages.
…ecutionEvents to one or more Message Sinks. Default implementation of one Sink that serializes Messages to an ndjson file per feature. Not finished: hooking up the TestThreadExecution event handlers.
…mber Messages Publisher and Broker to the initialization sequence provided by the Plugin architecture.
…est project. The File Sink is hooked in as a nuget package. Next: exploratory debugging to confirm the event listeners are wired up correctly at Feature start time.
Need to figure out why trace logging also not working.
…check of whether it had been initialized (temp). (Temp) - added Debugger.Launch to FileSinkPlugin to force launch of debugger.
…Publisher from the FileSinkPlugIn completely.
…the use of a BlockingCollection. Eliminated base class for Sinks. Eliminated use of Async for MessageSinks. Sinks have to ensure their own thread-safety and non-blocking behavior. May want to revisit that.
…ke Messages (meta, Source, GherkinDocument, and Pickles).
… information into Messages. First example is transforming StepBinding classes into messages of type StepDefinition.
…pport for CucumberMessages for that feature.
…e, then all are published.
…ioState emit sequences of Envelopes for consumption by the Publisher.
…fter scenario completion so that we have full binding Match info available to create the TestCase message.
… simple scenarios.
minimal and pending scenarios added.
… which I don't think apply to the .NET implementation. Need to add support for images and documents as embedded resources to be added to generated test projects before attachments and hooks can be implemented and tested. Not yet added infrastructure to compare output ndjon with expected.
…ature Tracker needs them. Eliminated the explicit use of the PickleStepSequence from the generated code; reducing the footprint of the changes required during generation.
…n EventListener rather than registering event handler delegates (should be marginally faster).
…ns get published as StepDefinition messages. Restructured test project to split CCK tests from basic integration tests.
…e. Found this to be necessary when Features are run in parallel in to order to keep the written messages from clobbering each other.
… version without the build hash portion.
The External Data sample needed adjustment to properly fetch its csv file from Resources given the new project structure.
…ost) invocations of it. Made Publisher, Broker, and FileSink async/await. Made FeatureTracker async, but not Scenario and Step trackers (tbd).
…lisher, Broker, and Sink). Returned FeatureTracker back to normal, sync methods.
…rectly serialized.
…at a FeatureFinishedEvent is received but not all Scenarios are yet marked as Finished, then the entire Feature success status should be set to False.
…enarios in parallel
…when two Features were rewriting simultaneously.
|
I believe this is now ready for a shakedown and code review. I've fixed the issues that I'm aware of and completed my punch-list. Documentation is updated. I've tested it with multiple Features running in parallel and desk-checked it for thread safety for the upcoming Scenario parallel execution model. |
|
It looks like Cucumber Messages schemas are changing in an upcoming release PR102. This will require some changes to logic within the Messages implementation. |
…hmentAdded events as Obsolete.
…ature class start-up. Modified FeatureInfo constructor to accept the FeatureMessages object as optional parameter. Had to modify a Runtime test to accomodate that change. Also changed Generation such that we're more thoroughly emitting the 'global::' prefix on Reqnroll types (for Messages related types). Marked the Non-Compliant CCK test scenarios as [Ignore].
…was failing now that is called well before the Messages configuration system is spun up.
…ce when throwing an 'already resolved' registration exception.
Draft PR to provide a space for reviews and suggestions on this Feature.
🤔 What's changed?
Adds support for creation of Cucumber Messages during the execution of Reqnroll scenarios.
Changes are in three main area:
FeatureGeneration: as a feature file is processed and used to generate C# code, the Gherkin toolchain is used to create the first few Messages for the feature. These are attached to the FeatureInfo object.
CucumberMessagePublication: written as an embedded plugin, this component listens for ExecutionEvents and converts them, and their attached state, to Cucumber Messages. These are 'published' to one or more sinks.
FileSinkPlugin: a traditional runtime plugin, this plugin registers with the Publishing component to receive the Messages as they are released by the Publisher and stored to a file (one per feature file).
Messages are strings serialized from the CucumberMessage Envelope format and stored one per line in an ".ndjson" result file.
⚡️ What's your motivation?
To provide future compatibility with the broader Cucumber eco-system of tooling.
To provide the input to a future implementation of a Reqnroll replacement to LivingDocs.
🏷️ What kind of change is this?
♻️ Anything particular you want feedback on?
Anything.
📋 Checklist: