Coursier

2.1.0-M5-5-g2cb552ea9

Edit

Overview

The CLI of coursier is the Scala application and artifact manager. It can install Scala applications and setup your Scala development environment. It can also download and cache artifacts from the web.

See Installation for how to install the CLI of coursier.

Once installed, the CLI of coursier provides a number of services:

This page succinctly describes each of these commands. More details about each them are then given in the dedicated documentation page of each command (see links on the left).

Getting started

Install cs and your Scala development environment

Main available commands

install

The install command installs Scala applications in the installation directory configured when installing cs (~/.local/share/coursier/bin by default on Linux):

$ cs install scalafmt
$ scalafmt --version
scalafmt 3.0.6

You may give a specific version number as follows:

$ cs install scalafmt:2.4.2
$ scalafmt --version
scalafmt 2.4.2

If the installation directory is not already on your PATH, you may add it for the current session with

$ eval "$(cs install --env)" # add installation directory in PATH in the current session

See the dedicated page for more details about install. It also presents the related commands list, update and uninstall.

search

If you are not sure what the exact name of the application you want to install is, use the search command:

$ cs search fmt
scalafmt

launch

launch launches applications from their name, or directly from one or more Maven dependencies. The applications need not be installed with install.

$ cs launch scalafmt -- --help
scalafmt 3.0.6
Usage: scalafmt [options] [<file>...]

  -h, --help               prints this usage text
  -v, --version            print version
…

Like the install command, launch accepts an optional version number. A typical use case is to start a Scala REPL with a specific Scala version:

$ cs launch scala:2.12.15
Welcome to Scala 2.12.15 (OpenJDK 64-Bit Server VM, Java 1.8.0_292).
Type in expressions for evaluation. Or try :help.

scala>

You may also combine launch with the fetch command to run a Scala REPL with specific dependencies on the classpath:

$ cs launch scala:2.13.7 -- -cp $(cs fetch --scala 2.13.7 -p org.typelevel::cats-core:2.6.0)

In order to launch an application that does not have a published application descriptor, launch accepts Maven coordinates as well:

$ cs launch org.scalameta::scalafmt-cli:2.4.2 -- --help
scalafmt 2.4.2
Usage: scalafmt [options] [<file>...]

  -h, --help               prints this usage text
  -v, --version            print version
…

Arguments can be passed to the application after --.

launch has more options, to specify a main class, pass Java options, etc. See the dedicated page for more details.

bootstrap

bootstrap creates standalone binary launchers from one or more dependencies. The produced launches do not need cs to be installed, so they can be shared, notably as part of a repository.

$ cs bootstrap org.scalameta::scalafmt-cli:2.4.2 -o scalafmt
$ ./scalafmt --version
scalafmt 2.4.2

On Linux, bootstrap does not create Windows .bat files by default. If you use it to produce shareable scripts, it is recommended to add the --bat option in order for Windows users not be stuck:

$ cs bootstrap --bat org.scalameta::scalafmt-cli:2.4.2 -o scalafmt
$ ls
scalafmt scalafmt.bat

bootstrap can generate a variety of launchers, including native ones like GraalVM native images or Scala Native executables. See the dedicated page for more details.

java

The java command manages JVMs. For example, the following command

$ cs java --jvm 11 -version
openjdk version "11.0.6" 2020-01-14
OpenJDK Runtime Environment AdoptOpenJDK (build 11.0.6+10)
OpenJDK 64-Bit Server VM AdoptOpenJDK (build 11.0.6+10, mixed mode)

will automatically download the latest AdoptOpenJDK 11 (in the coursier cache), unpack it (in the managed JVM directory), and run it with -version.

It uses the index of jabba to know where to download JVM archives, and assumes AdoptOpenJDK if only a version is passed.

The java-home command prints the Java home of a JVM, like

$ cs java-home --jvm 9
~/Library/Caches/Coursier/jvm/adopt@1.9.0-0/Contents/Home

One can update profile files (Linux / macOS) or user environment variables (Windows) for a specific JVM with --setup, like

$ cs java --jvm openjdk:1.14 --setup

(This updates JAVA_HOME and PATH.)

You don't have to update these environment variables if you prefer not to. For example, one can just get a Java home with

$ cs java-home --jvm graalvm:20
~/Library/Caches/Coursier/jvm/graalvm@20.0.0/Contents/Home

or set environment variables only for the current session with

$ eval "$(cs java --env --jvm 11)"
$ java -version
openjdk version "11.0.6" 2020-01-14
OpenJDK Runtime Environment AdoptOpenJDK (build 11.0.6+10)
OpenJDK 64-Bit Server VM AdoptOpenJDK (build 11.0.6+10, mixed mode)

These two commands only write things in the coursier cache, and the managed JVM directory, which is a cache too (both of these directories live under ~/.cache/coursier on Linux, ~/Library/Caches/Coursier on macOS).

The dedicated page gives more details about the java and java-home commands.

resolve

resolve lists the transitive dependencies of one or more other dependencies. Use like

$ cs resolve io.circe::circe-generic:0.12.3
com.chuusai:shapeless_2.13:2.3.3:default
io.circe:circe-core_2.13:0.12.3:default
io.circe:circe-generic_2.13:0.12.3:default
io.circe:circe-numbers_2.13:0.12.3:default
org.scala-lang:scala-library:2.13.0:default
org.typelevel:cats-core_2.13:2.0.0:default
org.typelevel:cats-kernel_2.13:2.0.0:default
org.typelevel:cats-macros_2.13:2.0.0:default

Note that this only relies on metadata files (POMs in particular), and doesn't download any JAR.

resolve has more options, to print trees, find which dependency brings another one, etc. See the dedicated page for more details.

fetch

fetch fetches the JARs of one or more dependencies.

$ cs fetch io.circe::circe-generic:0.12.3
/path/to/coursier/cache/https/repo1.maven.org/maven2/io/circe/circe-generic_2.13/0.12.3/circe-generic_2.13-0.12.3.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/org/scala-lang/scala-library/2.13.0/scala-library-2.13.0.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/io/circe/circe-core_2.13/0.12.3/circe-core_2.13-0.12.3.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/com/chuusai/shapeless_2.13/2.3.3/shapeless_2.13-2.3.3.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/io/circe/circe-numbers_2.13/0.12.3/circe-numbers_2.13-0.12.3.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/org/typelevel/cats-core_2.13/2.0.0/cats-core_2.13-2.0.0.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/org/typelevel/cats-macros_2.13/2.0.0/cats-macros_2.13-2.0.0.jar
/path/to/coursier/cache/https/repo1.maven.org/maven2/org/typelevel/cats-kernel_2.13/2.0.0/cats-kernel_2.13-2.0.0.jar

fetch has more options, to join its output with the path separator, fetch source JARs, fetch javadoc, etc. See the dedicated page for more details.

setup

setup is the command used to install a Scala development environment. It ensures that:

  • a JVM is installed on your machine,
  • the installation directory of the install command is in your PATH,
  • the standard Scala applications are installed on your system.

setup is a "repackaging" of features of the java and install commands. One doesn't have to run setup for the other commands to work fine, it's just a convenience.

The features that setup wraps up are not tied to each other. You can rely on the java or java-home commands while managing applications with your preferred package manager. Or you can manage JVMs any other way while using the install command. setup only conveniently sets them up in one go, and it only sets up a JVM if it doesn't find one already installed.

See the dedicated page for more details about what the setup command does.

← OverviewInstallation →