Pure Kotlin implementation of a generic command-line parser.
- Declarative: describe what your commands and parameters are
- Platform-agnostic: core library has no platform-specific dependencies and can be used in any Kotlin project
- Hackable: build extensions on top of it however you like
kotlinx-cli
can be used to create user-friendly and flexible command-line interfaces
for Kotlin/JVM, Kotlin/Native, and any other Kotlin console applications.
Program defines what arguments are expected.
kotlinx-cli
will figure out how to parse those, reporting errors if the program arguments are invalid,
and also generate help and usage messages as well.
Note that the library is experimental and the API is subject to change.
The library is published to Maven Central repository.
- Add the Maven Central repository if it is not already there:
repositories {
mavenCentral()
}
In Kotlin multiplatform projects, add the following dependency to a source set (it may be a common or platform specific source set):
kotlin {
sourceSets {
commonMain {
dependencies {
implementation("org.jetbrains.kotlinx:kotlinx-cli:0.3.5")
}
}
}
}
In Kotlin projects, add the following dependency to the dependencies
element of pom.xml
:
<dependency>
<groupId>org.jetbrains.kotlinx</groupId>
<artifactId>kotlinx-cli-jvm</artifactId>
<version>0.3.5</version>
</dependency>
There are 2 base entity: option and argument.
Option - command line entity started with some prefix (-/--) and can have value as next entity in command line string.
Argument - command line entity which role is connected only with its position.
Command line entities can be several types:
- ArgType.Boolean
- ArgType.Int
- ArgType.String
- ArgType.Double
- ArgType.Choice (value can be only from predefined list)
Custom types can be created.
import kotlinx.cli.*
fun produce(result: List<Double>, format: String, outputFileName: String?) {
outputFileName.let {
// Print to file.
...
} ?: run {
// Print to stdout.
...
}
}
fun readFrom(inputFileName: String): String {
...
}
fun calculate(inputData: String, eps: Double, debug: Boolean = false): List<Double> {
...
}
enum class Format {
HTML,
CSV,
PDF
}
fun main(args: Array<String>) {
val parser = ArgParser("example")
val input by parser.option(ArgType.String, shortName = "i", description = "Input file").required()
val output by parser.option(ArgType.String, shortName = "o", description = "Output file name")
val format by parser.option(ArgType.Choice<Format>(), shortName = "f",
description = "Format for output file").default(Format.CSV).multiple()
val stringFormat by parser.option(ArgType.Choice(listOf("html", "csv", "pdf"), { it }), shortName = "sf",
description = "Format as string for output file").default("csv").multiple()
val debug by parser.option(ArgType.Boolean, shortName = "d", description = "Turn on debug mode").default(false)
val eps by parser.option(ArgType.Double, description = "Observational error").default(0.01)
parser.parse(args)
val inputData = readFrom(input)
val result = calculate(inputData, eps, debug)
format.forEach {
produce(result, it, output)
}
}
It's also possible to use arguments in current example.
...
val input by parser.argument(ArgType.String, description = "Input file")
val output by parser.argument(ArgType.String, description = "Output file name").optional()
Auto-generated help message for this example is
Usage: example options_list
Arguments:
input -> Input file { String }
output -> Output file name (optional) { String }
Options:
--format, -f [csv] -> Format for output file { Value should be one of [html, csv, pdf] }
--debug, -d [false] -> Turn on debug mode
--eps [0.01] -> Observational error { Double }
--help, -h -> Usage info
If application has rich command line interface and executes different actions with different arguments, subcommands can be useful.
@file:OptIn(ExperimentalCli::class)
import kotlinx.cli.*
fun main(args: Array<String>) {
val parser = ArgParser("example")
val output by parser.option(ArgType.String, "output", "o", "Output file")
class Summary: Subcommand("summary", "Calculate summary") {
val invert by option(ArgType.Boolean, "invert", "i", "Invert results").default(false)
val addendums by argument(ArgType.Int, "addendums", description = "Addendums").vararg()
var result: Int = 0
override fun execute() {
result = addendums.sum()
result = if (invert!!) -1 * result else result
}
}
class Multiply: Subcommand("mul", "Multiply") {
val numbers by argument(ArgType.Int, description = "Addendums").vararg()
var result: Int = 0
override fun execute() {
result = numbers.reduce{ acc, it -> acc * it }
}
}
val summary = Summary()
val multiple = Multiply()
parser.subcommands(summary, multiple)
parser.parse(args)
}
Then help information will be available for each subcommand separately.
In case of example summary -h
help info will be
Usage: example summary options_list
Arguments:
addendums -> Addendums { Int }
Options:
--invert, -i -> Invert results
--help, -h -> Usage info
In case of example mul -h
help info will be
Usage: example mul options_list
Arguments:
numbers -> Addendums { Int }
Options:
--help, -h -> Usage info
The boolean property strictSubcommandOptionsOrder
defines the allowed order of options and arguments for subcommands.
When it is false
(default), then the main program's options can be specified everywhere, even after the subcommand.
Otherwise, parameters can only be specified after the subcommands where they are defined. For example,
@file:OptIn(ExperimentalCli::class)
import kotlinx.cli.*
fun main(args: Array<String>) {
val parser = ArgParser("example", strictSubcommandOptionsOrder = true)
val output by parser.option(ArgType.String, "output", "o", "Output file")
class Multiply: Subcommand("mul", "Multiply") {
val numbers by argument(ArgType.Int, description = "Addendums").vararg()
var result: Int = 0
override fun execute() {
result = numbers.reduce{ acc, it -> acc * it }
}
}
val multiple = Multiply()
parser.subcommands(summary, multiple)
parser.parse(args)
}
example -o out.txt mul 1 2 3 -o out.txt # OK
example mul 1 2 3 -o out.txt # fail in this case, but OK if strictSubcommandOptionsOrder is false