Skip to main content

​
Overview

Datoso supports a plugin-based architecture through seeds, which are specialized plugins that handle different DAT file sources. You can create custom seeds to support new DAT repositories or customize existing functionality.
Datoso requires Python 3.11 or higher for development.

​
Getting Started

To develop a custom seed, you should start with the base seed template:

datoso_seed_base

Official seed template repository with boilerplate code and structure

​
Seed Structure

A Datoso seed is a Python package with a specific structure. The naming convention is: 
datoso_seed_<SEED_NAME>

​
Required Components

Every seed must implement the following components:

​
1. Module Attributes

Your seed module must define these attributes in __init__.py: 
__prefix__ = 'PREFIX'  # Prefix for identification of dats
__description__ = 'Description of your seed'
__version__ = '1.0.0'
__author__ = 'Your Name'

​
2. Fetch Module

The fetch.py module must contain a fetch() function to download DAT files: 
# fetch.py
def fetch():
    """Function to fetch data from the source."""
    # Your implementation here
    pass

​
3. Rules Module

The rules.py module must contain a get_rules() function that returns the rules engine configuration: 
# rules.py
def get_rules():
    """Rules for the rules engine."""
    return {
        # Your rules configuration
    }

​
4. Actions Module

The actions.py module must contain a get_actions() function that defines DAT processing actions: 
# actions.py
def get_actions():
    """Actions to take for dats."""
    return {
        # Your actions configuration
    }

​
Plugin Discovery

Datoso automatically discovers installed seeds using Python’s pkgutil.iter_modules(). Seeds are identified by the naming pattern: 
datoso_seed_*

​
Checking Module Requirements

The doctor command validates your seed installation by checking:
  • Required module attributes (__prefix__, __description__)
  • Required functions (fetch.fetch, rules.get_rules, actions.get_actions)
  • External executable dependencies (if specified in __requirements__)
Example requirements specification: 
__requirements__ = {
    'executables': ['wget', 'curl']
}

​
Installation Methods

​
As a Package

Install your seed as a regular Python package: 
pip install datoso_seed_<SEED_NAME>

​
For Development

Install in editable mode for development: 
# Clone your seed repository
git clone https://github.com/yourusername/datoso_seed_myseed
cd datoso_seed_myseed

# Install in editable mode
pip install -e .

​
With Datoso Extras

You can also make your seed installable as a Datoso extra by adding it to the pyproject.toml: 
[project.optional-dependencies]
myseed = [ "datoso-seed-myseed>=1.0.0" ]

​
Available Official Seeds

The following official seeds are available as reference:
  • fbneo - Final Burn Neo
  • nointro - No-Intro Datomatic
  • redump - Redump
  • pleasuredome - Pleasuredome
  • tdc - Total DOS Collection
  • vpinmame - Visual Pinball
  • whdload - WHDLoad
  • eggman - Eggman Teknoparrot, ALLs.net
Some seeds (md_enhanced, sfc_enhancedcolors, sfc_msu1, sfc_speedhacks, translatedenglish) are marked as deprecated.

​
Managing Seeds

​
List Installed Seeds

View all installed seeds: 
datoso seed installed

​
Seed Details

Get detailed information about a specific seed: 
datoso seed details <seed_name>
This displays:
  • Full module name
  • Version
  • Author
  • Description

​
Validation

​
Doctor Command

Use the doctor command to validate your seed: 
# Check all seeds
datoso doctor

# Check specific seed
datoso doctor <seed_name>
The doctor command verifies:
  • Module is properly installed
  • All required attributes exist
  • All required functions are present
  • External dependencies are available

​
Helper Functions

Datoso provides helper functions in src/datoso/helpers/plugins.py: 
from datoso.helpers.plugins import get_seed, installed_seeds, seed_description

# Get a seed module
seed_module = get_seed('myseed')

# Get all installed seeds
seeds = installed_seeds()

# Get seed description
description = seed_description('datoso_seed_myseed')

​
Best Practices

Always use the datoso_seed_ prefix for your package name to ensure proper discovery.
Ensure your seed has all required attributes and functions. Use datoso doctor to validate.
Implement proper error handling in your fetch and process functions. Log errors appropriately.
Provide clear documentation about what your seed does, what sources it supports, and any special requirements.
Test both fetch and process operations with various DAT files before publishing.

​
Next Steps

Troubleshooting

Common issues and debugging tips

Configuration

Learn about Datoso configuration options

Build docs developers (and LLMs) love

Get started for freeTalk to us