Architecture

Directory Structure

High-Level Directory Structure

arduinoCopy code<root dir>/
├── ats/
│   ├── apis/
│   ├── database/
│   ├── exceptions/
│   ├── exchanges/
│   ├── indicators/
│   ├── logs/
│   ├── saved_states/
│   ├── state/
│   ├── static/
│   ├── strategies/
│   ├── templates/
│   ├── tests/
│   ├── trading_job/
│   ├── utils/
│   ├── __init__.py
│   ├── main.py
│   ├── test.py
├── custom_modules/
│   ├── exchanges/
│   ├── fees/
│   ├── indicators/
│   ├── strategies/
├── .env
├── environment.yml
├── requirements.txt
├── setup.sh
├── start.sh
└── README.md

---

Core Application: ats/

The ats folder contains the core application code, organized into multiple subdirectories for modularity and maintainability.

Key Folders

• apis/ Handles API endpoints, requests to external APIs, and response processing logic.

• database/ Manages database interactions, including models, schemas, connections, and CRUD operations.

• exceptions/ Custom exception classes for handling specific errors throughout the application.

• exchanges/ Connectors and integrations for interacting with various financial or crypto exchanges. Each built-in exchange resides in exchanges/exchanges/<exchange_name> and implements an Exchange class extending the BaseExchange class.

• indicators/ Implementations of technical indicators used in trading strategies, such as Moving Averages, RSI, or MACD. Each indicator must extend the BaseIndicator class.

• logs/ Contains application logs for debugging and monitoring performance.

• saved_states/ Stores serialized states or checkpoint data to maintain consistency across runs or allow resuming processes.

• state/ Serializable classes extending the SimpleState class for managing and persisting application states.

• static/ Static assets like CSS, JavaScript, and images for the trading UI.

• strategies/ Built-in trading strategies, each implemented as a class that extends the BaseStrategy class. Strategies should reside in strategies/<strategy_name> and must define a Strategy class in a strategy.py file.

• templates/ Template files (e.g., HTML files) for rendering dynamic content in the trading UI.

• tests/ Unit tests, integration tests, and other test-related files to verify application correctness.

• trading_job/ Core logic and classes for managing and executing trading jobs.

• utils/ Utility functions and helper modules for reusable functionalities like file operations, data transformations, and logging.

---

Custom Modules: custom_modules/

The custom_modules directory is dedicated to user-defined customizations. It allows users to add their own exchange connectors, strategies, indicators, and fee models. These modules override the core ones in the ats folder if there is a naming conflict.

Key Folders

• exchanges/ Custom exchange implementations. Each exchange must:

  • Be placed in custom_modules/exchanges/<exchange_name>.py.

  • Implement an Exchange class extending the BaseExchange class.

• fees/ Custom fee models for defining unique fee structures. Each fee model must:

  • Be placed in custom_modules/fees/<fee_model_name>.py.

  • Implement a class extending the BaseFees class.

• indicators/ Custom technical indicators. Each indicator must:

  • Be placed in custom_modules/indicators/<indicator_name>.py.

  • Implement a class extending the BaseIndicator class.

• strategies/ Custom trading strategies. Each strategy must:

  • Be placed in custom_modules/strategies/<strategy_name>/strategy.py.

  • Implement a Strategy class extending the BaseStrategy class.

---

Root-Level Files

• .env Contains environment-specific variables, such as API keys and configurations.

• environment.yml Specifies dependencies for setting up the project environment with Conda.

• requirements.txt Python dependencies for the application, used with pip.

• setup.sh A shell script for initializing and configuring the application environment.

• start.sh A shell script for starting the application.