| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
Sorry, something went wrong.
🤖 Devin AI EngineerI'll be helping with this pull request! Here's what you should know: ✅ I will automatically:
Note: I can only respond to comments from users who have write access to this repository. ⚙️ Control Options:
|
Sorry, something went wrong.
| return if @client.initialized? | ||
|
|
||
| state = @client.data_source_status_provider.status.state | ||
| raise "the LaunchDarkly client was unable to initialize; the data source state is #{state}" |
There was a problem hiding this comment.
The OpenFeature SDK will catch this exception? I want to double check that raising is the expectation.
Sorry, something went wrong.
There was a problem hiding this comment.
Yes — raising is the mechanism the OpenFeature Ruby SDK expects. Configuration#init_provider wraps the init call in a rescue, dispatches PROVIDER_ERROR (which moves provider state to ERROR), and only re-raises — as ProviderInitializationError — when the caller used set_provider_and_wait. With set_provider the init runs on its own thread and the exception is swallowed after the error event (v0.6.5 configuration.rb):
def init_provider(provider, context, raise_on_error: false)
...
dispatch_provider_event(provider, ProviderEvent::PROVIDER_READY)
rescue => e
error_code = (e.error_code if e.respond_to?(:error_code) && e.error_code) || Provider::ErrorCode::GENERAL
dispatch_provider_event(provider, ProviderEvent::PROVIDER_ERROR, error_code:, message: e.message)
raise ProviderInitializationError.new(..., original_error: e) if raise_on_error
endSo a failed initialization ends in the ERROR state rather than FATAL, matching the mapping we settled on in the spec: the LaunchDarkly client can still serve whatever data it has, so the OpenFeature SDK should keep calling the provider. Note the SDK also reads error_code off the exception if it exposes one, so we could raise a custom error later if we want something more specific than GENERAL.
Sorry, something went wrong.
Co-Authored-By: rlamb@launchdarkly.com <4955475+kinyoklion@users.noreply.github.com>
Adds a feature matrix to the README mirroring the OpenFeature Ruby SDK's own feature table, so a reader can see at a glance what the provider does and does not support. - One row per OpenFeature SDK feature (providers, targeting, hooks, logging, domains, eventing, shutdown, tracking, transaction context propagation, extending), plus initialization and flag metadata. - Every row is currently supported following the tracking, lifecycle, eventing, wrapper, and flag metadata work; each note says how the provider implements it, or that the OpenFeature SDK provides it with no provider involvement. - The legend keeps room for `⚠️`/`❌` so unsupported rows can state whether the limitation is in the OpenFeature Ruby SDK or is a provider gap. <details> <summary>Implementation details</summary> **Requirements** - [x] I have added test coverage for new or changed functionality - [x] I have followed the repository's pull request submission guidelines - [x] I have validated my changes against all supported platform versions **Related issues** Follow-up to the audit of LaunchDarkly's OpenFeature providers, which produced #27, #28, #29, #30 and #31 in this repository. **Describe the solution you've provided** The matrix reuses the row set and legend of the [OpenFeature Ruby SDK feature table](https://github.com/open-feature/ruby-sdk#-features) so the two can be read side by side, and so the same format can be applied to the other LaunchDarkly providers. Statements about eventing, initialization, shutdown, tracking, and flag metadata were checked against the current implementation in `lib/ldclient-openfeature`. **Describe alternatives you've considered** Writing a LaunchDarkly-specific row set instead of mirroring OpenFeature's — rejected, since the point of the matrix is comparability with the OpenFeature SDK table and with the other providers. **Testing** Documentation only; no code changes. **Additional context** Screenshots and a staging preview do not apply to a README change in a server-side library. </details> Link to Devin session: https://app.devin.ai/sessions/0c452d209ec54b068ba120b4c92b8f6c Requested by: @kinyoklion <!-- CURSOR_SUMMARY --> --- > [!NOTE] > **Overview** > Adds an OpenFeature **feature matrix** to the README so readers can see at a glance what this provider supports. > > The table mirrors the OpenFeature Ruby SDK feature list (providers, targeting, hooks, logging, domains, eventing, initialization, shutdown, tracking, transaction context, extending, flag metadata). Every row is marked supported, with notes on how the provider implements it or that the SDK handles it. The legend leaves room for partial/unsupported rows later. > > <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit 898abf3. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup> <!-- /CURSOR_SUMMARY --> Co-authored-by: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
🤖 I have created a release *beep* *boop* --- ## [0.3.0](0.2.0...0.3.0) (2026-08-25) ### Features * Emit provider events from LaunchDarkly client status ([#29](#29)) ([1b53b0c](1b53b0c)) * Populate OpenFeature flag metadata from the evaluation reason ([#31](#31)) ([e6ef78a](e6ef78a)) * Report provider name and version to LaunchDarkly ([#30](#30)) ([c5dd94d](c5dd94d)) * Support OpenFeature tracking ([#27](#27)) ([709f52d](709f52d)) * Support provider initialization and shutdown ([#28](#28)) ([27df803](27df803)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please). <!-- CURSOR_SUMMARY --> --- > [!NOTE] > **Overview** > **Release 0.3.0** — this PR only bumps version metadata from **0.2.0** to **0.3.0** in `.release-please-manifest.json`, `lib/ldclient-openfeature/version.rb`, and the `SDK_VERSION` example in `PROVENANCE.md`, and adds the **0.3.0** section to `CHANGELOG.md`. > > No provider or SDK logic changes appear in the diff. The new changelog entry documents features already landed on main: OpenFeature **tracking**, provider **initialization/shutdown**, **provider events** from LaunchDarkly client status, **flag metadata** from evaluation reason, and reporting **provider name/version** to LaunchDarkly. > > <sup>Reviewed by [Cursor Bugbot](https://cursor.com/bugbot) for commit 8416ff3. Bugbot is set up for automated code reviews on this repo. Configure [here](https://www.cursor.com/dashboard/bugbot).</sup> <!-- /CURSOR_SUMMARY --> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
| Back | FazBrowse Home | New Git URL |
Implements the OpenFeature provider lifecycle hooks so the OpenFeature SDK can reflect LaunchDarkly initialization failures and release resources.
- Adds Provider#init, which raises when the LaunchDarkly client did not initialize, so the OpenFeature SDK moves the provider to ERROR instead of reporting READY.
- Adds Provider#shutdown, which closes the LaunchDarkly client when the provider is replaced or OpenFeature::SDK.shutdown is called.
- The LaunchDarkly client is still created in the constructor, matching the Java provider; a client cannot be restarted, so a new provider instance is required after shutdown.
Implementation detailsRequirements
Related issues
Part of an audit of the LaunchDarkly OpenFeature providers against the current OpenFeature specification. Provider lifecycle (spec sections 2.4 / 1.7) was unimplemented here: initialization failures were invisible to the OpenFeature SDK, and shutting down the OpenFeature SDK leaked the LaunchDarkly client's connections and event processor.
Describe the solution you've provided
OpenFeature::SDK::Configuration calls init on providers that respond to it and dispatches PROVIDER_READY on success, or PROVIDER_ERROR when init raises; it calls shutdown on providers that respond to it. init uses LDClient#initialized? and includes the data source state from data_source_status_provider in the raised error message for diagnosability.
Describe alternatives you've considered
Deferring LaunchDarkly client construction to init — rejected because it would change the meaning of the existing public client accessor and the constructor's wait_for_seconds argument. Blocking inside init on a data source status listener — rejected as redundant, since the constructor already waits, and later state changes are better delivered as provider events.
Testing
bundle exec rake (RSpec + RuboCop) on Ruby 3.4: 57 examples, 0 failures; 12 files inspected, no offenses. Ruby 3.4 was used because the gemspec requires >= 3.4.
@cursor review
Link to Devin session: https://app.devin.ai/sessions/0c452d209ec54b068ba120b4c92b8f6c
Requested by: @kinyoklion
Note
Overview
Implements OpenFeature provider lifecycle so the SDK can report LaunchDarkly init failures and release resources.
Provider#init no-ops if the client is already initialized; otherwise it raises with the data source state so OpenFeature can move the provider to ERROR instead of READY. Provider#shutdown closes the LaunchDarkly client. Construction still happens in the constructor; a closed client cannot be restarted.
Specs cover successful init, failed init, and shutdown.
Reviewed by Cursor Bugbot for commit bc88a8b. Bugbot is set up for automated code reviews on this repo. Configure here.