Arden logoArden

Quick Start

This guide walks through the fastest ways to get useful output from Arden.

The goal is not just to run one file, but to understand the normal workflow shape:

  1. run a single file
  2. inspect it with check, fmt, and profile
  3. create a project
  4. move into project-aware info, test, and build

1. Run A Single File

Create hello.arden:

import std.io.*;

function main(): None {
    println("Hello, Arden!");
    return None;
}

Run it:

arden run hello.arden

Useful companion commands:

arden check hello.arden
arden compile hello.arden
arden fmt hello.arden
arden lint hello.arden
arden profile hello.arden

What each one is for:

  • run builds and executes the file
  • check stops before native code generation
  • compile emits a native artifact without running it
  • fmt normalizes formatting
  • lint reports static findings
  • profile gives a one-run timing summary

2. Look At A Slightly Larger Program

Once the hello-world path works, try something with a loop and state:

import std.io.*;

function main(): None {
    name: String = "Arden";
    mut counter: Integer = 0;

    println("Starting count for {name}...");

    while (counter < 5) {
        counter += 1;
        println("Count: {counter}");
    }

    println("Done!");
    return None;
}

Run it:

arden run program.arden

If you want more than this, jump straight to:

  • examples/04_control_flow.arden
  • examples/10_ownership.arden
  • examples/14_async.arden

Reference index:

3. Create A Project

arden new my_project
cd my_project
arden run

That generates:

  • arden.toml
  • src/main.arden
  • README.md

Inspect the resolved config:

arden info

At that point you are using project mode rather than ad-hoc single-file compilation.

4. Understand arden.toml

The generated project config is the source of truth for:

  • project name and version
  • entry file
  • full source file list
  • output file name
  • optimization level
  • output kind

Minimal example:

name = "my_project"
version = "0.1.0"
entry = "src/main.arden"
files = ["src/main.arden"]
output = "my_project"
opt_level = "3"
output_kind = "bin"

As you add files, update files = [...] explicitly. Arden project mode does not rely on vague directory scanning.

Reference:

5. Pass Program Arguments

arden run forwards trailing arguments to the compiled program.

arden run hello.arden one two three

Inside Arden, read them with:

  • Args.count()
  • Args.get(index)

Reference:

6. Use The Built-In Tooling

Arden bundles several common workflows:

arden fmt .
arden fix src/main.arden
arden test
arden bench hello.arden --iterations 5
arden bindgen sample.h -o bindings.arden

That matters because once you are in a project you do not need separate unrelated wrappers for formatting, testing, and simple performance checks.

Reference:

7. Use A Shebang Script

On Unix-like systems:

#!/usr/bin/env arden
import std.io.*;

function main(): None {
    println("hello");
    return None;
}

This is useful for quick tooling scripts where creating a whole project would be overkill.

8. Common First-Day Workflow

A realistic first session often looks like this:

arden new scratchpad
cd scratchpad
arden info
arden run
arden check --timings
arden test
arden build --release

That sequence exercises the main project commands without requiring any extra repo-specific setup.

If Something Fails Early

The most common setup problems are toolchain related:

  • LLVM or Clang missing
  • linker mismatch (mold on Linux, lld on macOS/Windows)
  • arden not on your shell path yet

Use:

arden --help
arden info

and compare your machine against:

Next Steps

NextEditor Setup