Skip to content

code_style.md

Latest commit

 

History

History
391 lines (292 loc) · 10.5 KB

code_style.md

File metadata and controls

391 lines (292 loc) · 10.5 KB

Code Style and Formatting Standards

Overview

Consistent code formatting ensures readability, maintainability, and reduces merge conflicts. All Python code must follow these standards.

Automation: CI and local hooks enforce Ruff (lint, format, import sorting) and mypy on the public CI source paths reported by uv run python -m infrastructure.project.public_scope source-paths. Mirror commands live in root CLAUDE.md; hook IDs are declared in /.pre-commit-config.yaml.

Python Code Style (PEP 8)

Core PEP 8 Rules

Line Length

# ✅ GOOD: Under 88 characters (Ruff default line length)
result = some_function_with_long_name(argument_one, argument_two)

# ❌ BAD: Too long
result = some_function_with_a_very_long_name_that_exceeds_the_limit(argument_one, argument_two)

Indentation

# ✅ GOOD: 4 spaces
def function():
    if condition:
        return True
    else:
        return False

# ❌ BAD: Mixed tabs/spaces
def function():
	if condition:  # Tab
	    return True  # Spaces

Imports

# ✅ GOOD: Standard library first, then third-party, then local
import os
import sys
from pathlib import Path

import numpy as np
import pandas as pd
import pytest

from infrastructure.core.logging.utils import get_logger
from project.src.data_processing import clean_data

# ❌ BAD: Not grouped or ordered
from project.src.data_processing import clean_data
import pytest
import os
import numpy as np

Blank Lines

# ✅ GOOD: Logical grouping
import os

def function_one():
    pass

def function_two():
    pass

class MyClass:
    pass

# ❌ BAD: Too many or missing blank lines
import os
def function_one():
    pass
def function_two():
    pass

class MyClass:
    pass

Naming Conventions

# ✅ GOOD: snake_case for functions/variables
def process_data():
    raw_data = load_data()
    processed_data = clean_data(raw_data)
    return processed_data

# ✅ GOOD: PascalCase for classes
class DataProcessor:
    pass

# ✅ GOOD: UPPER_CASE for constants
MAX_RETRIES = 3
DEFAULT_TIMEOUT = 30

# ❌ BAD: camelCase or inconsistent
def processData():
    rawData = load_data()
    processedData = cleanData(rawData)

String Quotes

# ✅ GOOD: Double quotes for strings
message = "Hello, world!"

# ✅ GOOD: Single quotes for contractions or apostrophes
error_msg = "Don't do that"

# ❌ BAD: Inconsistent
message = "Hello, world!'

Formatting and linting (Ruff)

Ruff subsumes the roles historically filled by separate formatters (Black), import sorters (isort), and many flake8 rules. Configuration lives under [tool.ruff] in the repository root pyproject.toml.

Commands (CI parity)

uv run python -m infrastructure.project.public_scope source-paths | xargs uvx ruff check --fix
uv run python -m infrastructure.project.public_scope source-paths | xargs uvx ruff format
uv run python -m infrastructure.project.public_scope source-paths | xargs uv run mypy

Line length

Keep wrapping aligned with the 88-character default unless pyproject.toml overrides Ruff.

Legacy tooling (optional locally)

Black, isort, and flake8 are not the CI source of truth for this repository; use them only if you intentionally duplicate checks outside Ruff.

Pre-commit Hooks

pre-commit Configuration

The repository ships .pre-commit-config.yaml at the repo root. Do not paste alternate Black/isort/flake8-only configs here — extend the existing hooks instead.

Illustrative hook IDs (authoritative YAML lives at repo root):

# See /.pre-commit-config.yaml — commit-stage hooks include:
#   ruff-ci  → uvx ruff check … && uvx ruff format …
#   mypy-ci  → uv run mypy …

Installing pre-commit

# Install pre-commit
uv tool install pre-commit

# Install hooks
pre-commit install

# Run on all files
pre-commit run --all-files

# Run specific hook (see .pre-commit-config.yaml for ids)
pre-commit run ruff-ci --all-files

Docstring Standards

Google Style Docstrings

def process_data(data: List[Dict], threshold: float = 0.5) -> Dict[str, Any]:
    """Process input data and filter based on threshold.

    This function takes raw data, applies filtering criteria,
    and returns processed results with statistics.

    Args:
        data: List of dictionaries containing data points.
            Each dict should have 'value' and 'category' keys.
        threshold: Minimum value threshold for filtering.
            Values below this are excluded. Defaults to 0.5.

    Returns:
        Dictionary containing:
        - 'filtered_data': List of filtered data points
        - 'statistics': Dict with count, mean, std stats
        - 'excluded_count': Number of excluded items

    Raises:
        ValueError: If data is empty or threshold is invalid.
        TypeError: If data format is incorrect.

    Example:
        >>> data = [{'value': 0.8, 'category': 'A'}, {'value': 0.3, 'category': 'B'}]
        >>> result = process_data(data, threshold=0.5)
        >>> result['filtered_data']
        [{'value': 0.8, 'category': 'A'}]
    """

NumPy Style Docstrings

def process_data(data, threshold=0.5):
    """Process input data and filter based on threshold.

    Parameters
    ----------
    data : list of dict
        List of dictionaries containing data points.
        Each dict should have 'value' and 'category' keys.
    threshold : float, optional
        Minimum value threshold for filtering. Default is 0.5.

    Returns
    -------
    dict
        Dictionary containing:
        - 'filtered_data': List of filtered data points
        - 'statistics': Dict with count, mean, std stats
        - 'excluded_count': Number of excluded items

    Raises
    ------
    ValueError
        If data is empty or threshold is invalid.
    TypeError
        If data format is incorrect.

    Examples
    --------
    >>> data = [{'value': 0.8, 'category': 'A'}, {'value': 0.3, 'category': 'B'}]
    >>> result = process_data(data, threshold=0.5)
    >>> result['filtered_data']
    [{'value': 0.8, 'category': 'A'}]
    """

File Organization Standards

Module Structure

# __init__.py - Public API exports
"""Data processing module."""

from .core import process_data, validate_input
from .config import DataConfig

__all__ = ["process_data", "validate_input", "DataConfig"]

# core.py - Main implementation
"""Core data processing functionality."""

from infrastructure.core.logging.utils import get_logger

logger = get_logger(__name__)

def process_data(data):
    """Process data."""
    logger.debug(f"Processing {len(data)} items")
    # Implementation...

# config.py - Configuration (optional)
"""Configuration for data processing."""

from dataclasses import dataclass

@dataclass
class DataConfig:
    threshold: float = 0.5

# cli.py - CLI interface (optional)
"""Command-line interface."""

import argparse

def main():
    parser = argparse.ArgumentParser()
    # CLI implementation...

Test File Organization

# tests/test_module.py
"""Tests for module."""

import pytest
from module import function_to_test

def test_basic_functionality():
    """Test basic case."""
    result = function_to_test("input")
    assert result == "expected"

def test_edge_cases():
    """Test edge cases."""
    assert function_to_test("") is None
    assert function_to_test(None) is None

def test_error_conditions():
    """Test error handling."""
    with pytest.raises(ValueError):
        function_to_test("invalid")

Best Practices

Do's ✅

  • ✅ Use Ruff (uvx ruff check, uvx ruff format) for lint/format/import sorting — CI parity
  • ✅ Use mypy (uv run mypy) on the public CI source paths
  • ✅ Use pre-commit hooks (ruff-ci, mypy-ci) to automate checks
  • ✅ Follow PEP 8 guidelines
  • ✅ Use descriptive names for variables and functions
  • ✅ Keep functions small (under 50 lines)
  • ✅ Use consistent quotes (double for strings)
  • ✅ Add docstrings to all public functions
  • ✅ Use type hints on all function signatures

Don'ts ❌

  • ❌ Mix tabs and spaces for indentation
  • ❌ Use lines longer than 88 characters
  • ❌ Import with wildcards (from module import *)
  • ❌ Use single-letter variable names (except in comprehensions)
  • ❌ Skip blank lines between logical sections
  • ❌ Use inconsistent naming (mix camelCase and snake_case)
  • ❌ Hardcode magic numbers without constants
  • ❌ Skip error handling in user-facing functions
  • ❌ Use print() instead of logging
  • ❌ Commit unformatted code

Tool Integration

VS Code Settings

Add to .vscode/settings.json:

{
  "[python]": {
    "editor.defaultFormatter": "charliermarsh.ruff",
    "editor.formatOnSave": true,
    "editor.codeActionsOnSave": {
      "source.fixAll.ruff": "explicit",
      "source.organizeImports.ruff": "explicit"
    }
  },
  "python.analysis.typeCheckingMode": "basic"
}

Use uv run mypy from the terminal for CI-parity typing; enable mypy or basedpyright in the IDE if you want inline diagnostics.

PyCharm / IntelliJ

  • Editor → Code Style → Python: Line length 88 to match Ruff.
  • Enable Ruff plugin (or configure external tool uvx ruff) for format/on-save.
  • Run uv run mypy as an external tool or use bundled typing support.

Enforcement

CI/CD Checks

Mirror .github/workflows/ci.yml lint job locally:

uv run python -m infrastructure.project.public_scope source-paths | xargs uvx ruff check
uv run python -m infrastructure.project.public_scope source-paths | xargs uvx ruff format --check
uv run python -m infrastructure.project.public_scope source-paths | xargs uv run mypy

Development Workflow

  1. Write code following style guidelines
  2. Run pre-commit to format and lint
  3. Fix any issues flagged by tools
  4. Run tests to ensure functionality
  5. Commit with confidence

See Also