Server-side launchdarkly feature flags for nextjs with client-side experimentation and insights. Detects known bots and excludes them from cient-side instrumentation, while still including them in server-side feature flags.
npm install @carforyou/ld
Include the express middleware in your custom server to fetch flags server-side. Pass a function getLDUser in which you return the LD user object. For example, you could read the user id from an auth token and pass it as the key or generate uuids for public users.
import { getLDRequestHandler } from "@carforyou/ld"
const getLDUser = (req, res, isBot) => ({ key: "user@example.com", anonymous: false })
const ldRequestHandler = getLDRequestHandler(LAUNCH_DARKLY_SDK_KEY, getLDUser)
const server = express()
server.use(ldRequestHandler)
.use(nextHandler)
.listen()
In _app, pass req.ldData to the LDProvider. Feature flags are statically available accross requests and page transitions.
import { LDProvider } from "@carforyou/ld"
static async getInitialProps({ req }) {
return { ldData: req.ldData }
}
render() {
<LDProvider ldClientId={LAUNCH_DARKLY_ID} initialLDData={ldData}>
...
</LDProvider>
}
Then in your components
import React from "react"
import { useFlags } from "@carforyou/ld"
const AppHead: FC = () => {
const flags = useFlags()
return flags.exampleFlag ? "on" : "off"
}
Alternatively, you can also use the context and get hold of some additional information
import React, { useContext } from "react"
import { LDContext } from "@carforyou/ld"
const AppHead: FC = () => {
const { flags, user, isBot } = useContext()
}
npm run build
You can link your local npm package to integrate it with any local project:
npm run link -- <relative_path_to_project>
This ensures that projects react is linked back to build package and prevents errors due to duplicate react instances.
New versions are released on the ci using semantic-release as soon as you merge into master. Please make sure your merge commit message adheres to the corresponding conventions.