go-feature-flag-web Provider for OpenFeature

About this provider

GO Feature Flag provider allows you to connect to your GO Feature Flag instance with the @openfeature/web-sdk.

The main difference between this provider and @openfeature/go-feature-flag-provider is that it uses a static evaluation context.
This provider is more sustainable for client-side implementation.

If you want to know more about this pattern, I encourage you to read this blog post.

What is GO Feature Flag?

GO Feature Flag is a simple, complete and lightweight self-hosted feature flag solution 100% Open Source.
Our focus is to avoid any complex infrastructure work to use GO Feature Flag.

This is a complete feature flagging solution with the possibility to target only a group of users, use any types of flags, store your configuration in various location and advanced rollout functionality. You can also collect usage data of your flags and be notified of configuration changes.

Install the provider

npm install @openfeature/go-feature-flag-web-provider @openfeature/web-sdk

How to use the provider?

const evaluationCtx: EvaluationContext = {
  targetingKey: 'user-key',
  email: '[email protected]',
  name: 'John Doe',
};

const goFeatureFlagWebProvider = new GoFeatureFlagWebProvider({
  endpoint: endpoint,
  // ...
}, logger);

await OpenFeature.setContext(evaluationCtx); // Set the static context for OpenFeature
OpenFeature.setProvider(goFeatureFlagWebProvider); // Attach the provider to OpenFeature
const client = await OpenFeature.getClient();

// You can now use the client to use your flags
if(client.getBooleanValue('my-new-feature', false)){
    //...
}

// You can add handlers to know what happen in the provider
client.addHandler(ProviderEvents.Ready, () => { ... });
client.addHandler(ProviderEvents.Error, () => { //... });
client.addHandler(ProviderEvents.Stale, () => { //... });
client.addHandler(ProviderEvents.ConfigurationChanged, () => { //... });

Available options

| Option name | Type | Default | Description | |-------------------------------|--------|----------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | endpoint | string | | endpoint is the URL where your GO Feature Flag server is located. | | apiTimeout | number | 0 = no timeout | (optional) timeout is the time in millisecond we wait for an answer from the server. | | apiKey | string | | (optional) If GO Feature Flag is configured to authenticate the requests, you should provide an API Key to the provider. Please ask the administrator of the relay proxy to provide an API Key. | | websocketRetryInitialDelay | number | 100 | (optional) initial delay in millisecond to wait before retrying to connect the websocket | | websocketRetryDelayMultiplier | number | 2 | (optional) multiplier of websocketRetryInitialDelay after each failure (example: 1st connection retry will be after 100ms, second after 200ms, third after 400ms ...) | | websocketMaxRetries | number | 10 | (optional) maximum number of retries before considering the websocket unreachable |

Reconnection

If the connection to the GO Feature Flag instance fails, the provider will attempt to reconnect with an exponential back-off.
The websocketMaxRetries can be specified to customize reconnect behavior.

Event streaming

The GoFeatureFlagWebProvider receives events from GO Feature Flag with changes. Combined with the event API in the web SDK, this allows for subscription to flag value changes in clients.

client.addHandler(ProviderEvents.ConfigurationChanged, (ctx: EventDetails) => {
  // do something when the configuration has changed.
  // ctx.flagsChanged contains the list of changed flags.
});

Contribute

Building

Run nx package providers-go-feature-flag-web to build the library.

Running unit tests

Run nx test providers-go-feature-flag-web to execute the unit tests via Jest.