AGENT.md

Python Usage Preferences

• Code Structure & Organization:

  • Script + Variable Configuration: By default, CLI is not used unless specifically requested. Configuration items are defined as constants or variables at the top of the script.
  • Modular Design: Prefers to break down the code into small, reusable modules.

• Import Style:

  • Unified Import Method: Uses from pathlib import Path, avoiding import pathlib.

• String Handling:

  • String Format: Prefers double quotes (") over single quotes (').

• Type Hinting:

  • Type Annotations: Prefers type hints. Non-primitive types are always enclosed in double quotes.

• Object-Oriented Programming:

  • Class Syntax: Prefers using class for encapsulation of complex logic.

• Code Style:

  • Clean, Readable Code: Encapsulates complex logic into functions and classes. Frequently adds detailed comments.

• Error Handling:

  • Exception Handling: Uses try-except blocks for exceptions with appropriate logging.

• File & Data Operations:

  • Path Management: Prefers Pathlib for handling file paths.
  • Directory Operations: Uses methods like Path.mkdir() and Path.exists() to check and create directories.

• Testing & Debugging:

  • Simple Test Cases: Prefers simple functional tests added after the __main__ block.

• Performance Optimization:

  • Execution Efficiency: Focused on optimizing code execution, especially with large datasets.
  • Libraries: Frequently uses Polars and Pandas.

• Libraries & Frameworks:

  • Data Analysis & Processing: Prefers Pandas, Polars, and SQLAlchemy.
  • Task Scheduling: Uses Apache Airflow for task automation.
  • Machine Learning: Utilizes scikit-learn and XGBoost when needed.

• Asynchronous & Concurrent Processing:

  • Asyncio & Multi-threading: Prefers asyncio and multi-threading for concurrent processing.