RNMLKit provides Expo modules that allow developers to use MLKit native libraries in their Expo apps.
Here are some previews in action:
Face Detection | Object Detection |
---|---|
git clone [email protected]:infinitered/react-native-mlkit.git
yarn
yarn build
Note: MLKit is not supported in the iOS simulator. You will need to use a hardware device.
cd apps/ExampleApp
npx expo run:ios -d
- Create the native project folders
cd apps/ExampleApp
npx expo prebuild
- Open
apps/ExampleApp/ios/ExampleApp.xcworkspace
in XCode - Select the
ExampleApp
target, and select a Team on the "Signing & Capabilities" tab - Select your target hardware device (MLKit is not supported in the simulator at this time)
- Click the play button to build the app
Note: Android support is currently under active development, some modules may not function as intended
cd apps/ExampleApp
# run on hardware device
npx expo run:android -d
The monorepo is organized into the following sections:
- Apps -- holds the example app (
ExampleApp
) that demonstrates how to use the modules. - Modules -- holds the modules themselves. Each module is a separate npm package, and they are published to npm
under the name
@infinitered/[module name]
. - Packages -- holds internal packages that are only used by the modules and are not published to npm. Things like the linter config.
For usage instructions see the readme in each module's directory.
Use yarn build
at the root level to compile typescript in all workspaces. You will need to do this when you make
changes before other modules to be able to see the changes. (The packages have a main field that points to the compiled
code.)
Running yarn build
inside of an expo module workspace will start watch mode for files inside of that module.
Expo SDK | MLKit |
---|---|
^49.0.0 | <= 0.7.7 |
^50.0.0 | ^1.0.0 |
^51.0.0 | ^1.0.0 |
This project follows the all-contributors specification. Contributions of any kind welcome!