Plugin is a JavaScript package that extends built-in React Native CLI features. It can provide an array of additional commands to run or platforms to target.
For example, react-native-windows
package is a plugin that provides run-windows
command and windows
platform.
Details of this particular integration as well as how to provide an additional platform for React Native were described in a dedicated section
about platforms.
Except for React Native dependencies, where configuration is implicit, each package needs to have a react-native.config.js
at the root folder in order to be discovered by the CLI as a plugin.
module.exports = {
commands: [
{
name: 'foo-command',
func: () => console.log('It worked'),
},
],
};
Above is an example of a plugin that exports a command named
foo-command
that can be executed withnpx react-native foo-command
and logs "It worked" and exits.
At the startup, React Native CLI reads configuration from all dependencies listed in package.json
and reduces them into a single configuration.
At the end, an array of commands concatenated from all plugins is passed on to the CLI to be loaded after built-in commands.
Also, you can provide your own configuration by creating a react-native.config.js
file in the root of the project, which overrides the configuration from plugins.
See
healthChecks
for information on how plugins can provide additional health checks fornpx react-native doctor
.
type Command = {
name: string;
description?: string;
func: (argv: Array<string>, config: ConfigT, args: Object) => ?Promise<void>;
options?: Array<{
name: string;
description?: string;
parse?: (val: string) => any;
default?:
| string
| boolean
| number
| ((config: ConfigT) => string | boolean | number);
}>;
examples?: Array<{
desc: string;
cmd: string;
}>;
};
Note:
ConfigT
is described inconfiguration
section
A name that will be used in order to run the command.
Note: If you want your command to accept additional arguments, make sure to include them in the name.
For example, my-command <argument>
will require an argument to be provided and will throw a validation error otherwise. Alternatively, my-command [argument]
will accept an argument, but will not throw when run without it. In that case, make sure to check for its presence.
Function that will be run when this command is executed. Receives an array of arguments, in order they were provided, a config object (see configuration
section) and options, that were passed to your command.
You can return a Promise if your command is async.
All errors are handled by the built-in logger. Prefer throwing instead of implementing your own logging mechanism.
An array of options that your command accepts.
Name of the option.
For example, a --reset-cache
option will result in a resetCache: true
or resetCache: false
present in the options
object - passed to a command function as a last argument.
Just like with a command name, your option can require a value (e.g. --port <port>
) or accept an optional one (e.g. --host [host]
). In this case, you may find default
value useful.
Note: Almost all names are accepted, but there's an exception for
--version
, we consider it a reserved word and do not guarantee that it will work as expected, so avoid using it. (see this for further details)
Optional description of your option. When provided, will be used to output a better help information.
Parsing function that can be used to transform a raw (string) option as passed by the user into a format expected by your function.
Default value for the option when not provided. Can be either a primitive value or a function, that receives a configuration and returns a primitive.
Useful when you want to use project settings as default value for your option.
An array of example usage of the command to be printed to the user.
String that describes this particular usage.
A command with arguments and options (if applicable) that can be run in order to achieve the desired goal.