Add Language Support

Goal

Add fully functioning support for a new programming language to the current course repository.

Procedure

1. Environment Setup

Ensure all necessary tools are installed before proceeding:

1. Bun: Check if Bun is installed. If not, run curl -fsSL https://bun.sh/install | bash and source the shell configuration (e.g., source ~/.bashrc).
2. Docker: Ensure Docker is installed. Start and verify the daemon:
   • sudo service docker start
   • sudo docker info
3. course-sdk: Check if course-sdk is installed. If not, run bun install and make install in the repository root to compile the SDK.

2. Initialization

Execute this SDK command to scaffold the basic directory structure: course-sdk add-language <LANGUAGE>

3. Pattern Analysis & Implementation

You must ensure the new language follows the exact implementation patterns of existing languages in this course.

1. Read the starter_templates directory of 2-3 existing working languages (e.g., starter_templates/go, starter_templates/python) to understand the challenge logic, SDKs, dependencies, and patterns.
2. Resolve dependencies for the new language: If existing starters use a vendor SDK (e.g. OpenAI), first check whether that vendor provides an official client for the target language. Use the vendor's official SDK for that language if it exists; only then fall back to a community or "most similar" SDK.
3. Read the newly generated starter file for the target language (in starter_templates/<LANGUAGE>).
4. Edit the target language's source code to implement the necessary boilerplate and "Stage 1" logic.
   • Constraint: STOP after Stage 1. Do not implement Stage 2, Stage 3, or any further logic.
   • Constraint: Prefer the same vendor's official SDK in the new language. If no official SDK exists for that language, use the most similar community SDK or dependency.
   • Constraint: Keep it minimal and matching the simplicity of other starter templates and don't complicate it with things not present in the other starter templates (e.g. extra functions when others have only one main function)
   • Constraint: Ignore any "DON'T EDIT THIS!" comments in the starter template.
   • Constraint: Do NOT manually create or edit files in compiled_starters/ or solutions/. These are all auto-generated by course-sdk compile and course-sdk test.
   • Constraint (Argument Parsing): Starter code must be simple and surprise-free, especially for users new to the platform. When a challenge requires parsing command-line arguments:
     • If the language has a stdlib argument parsing module (e.g., Python's argparse, Go's flag, Rust's clap if already a dependency), use it.
     • Otherwise, use the simplest possible inline check — a direct comparison, not a loop. For example:

       if sys.argv[1] != "-E":
          print("Expected first argument to be '-E'")
          exit(0)
       

     • Never use manual loop-based argument parsing (e.g., iterating over args with for / while to match flags). This adds unnecessary complexity and confuses newcomers.

4. The Iteration Loop (Compile & Test)

Repeat this process until tests pass:

1. Compile: Run sudo -E course-sdk compile <LANGUAGE>.
   • If it fails: Read the error output, analyze the source code, fix the syntax/build error, and retry.
2. Test: Run sudo -E course-sdk test <LANGUAGE>. Do not timeout.
   • Constraint: Verify ONLY the first stage. If Stage 1 passes, the task is complete. Do not attempt to fix or run tests for subsequent stages.
   • If it fails: Compare the expected output vs. actual output. Check how other languages handle this specific test case. Adjust the code and retry.

5. Final Verification

Once sudo -E course-sdk test <LANGUAGE> passes Stage 1, run a final comparison:

• Ensure no logic for Stage 2 or later is present in the solutions.
• Check if the code style is clean.
• Ensure no unnecessary files were created.
• Confirm the implementation matches the simplicity of other starter templates.
• Verify that any argument parsing follows the simplicity constraint: stdlib parser or simple inline check — no loop-based parsing.

Converted and distributed by TomeVault — claim your Tome and manage your conversions.