Git statistics over commits across a yearly time window.

git_stats.py

# /// script
# requires-python = ">=3.11"
# dependencies = []
# ///
# uv run git_stats.py 2025 <optional author>

import subprocess
import sys
from collections import defaultdict

def get_detailed_author_stats(year, target_author):
    # Format: HASH | DATE | ADDED | REMOVED
    cmd = [
        "git", "log",
        f"--since={year}-01-01",
        f"--until={year}-12-31",
        f"--author={target_author}",
        "--numstat",
        "--pretty=format:COMMIT:%h|%ad",
        "--date=short",
        "--",
        "**/test_*.py",
        # "**/tests/*.py",
    ]

    try:
        result = subprocess.run(cmd, capture_output=True, text=True, check=True)

        commits = []
        current_commit = None

        for line in result.stdout.splitlines():
            line = line.strip()
            if not line: continue

            if line.startswith("COMMIT:"):
                details = line.replace("COMMIT:", "").split("|")
                current_commit = {"hash": details[0], "date": details[1], "added": 0, "removed": 0}
                commits.append(current_commit)
                continue

            if current_commit:
                parts = line.split()
                if len(parts) >= 3:
                    try:
                        current_commit["added"] += int(parts[0])
                        current_commit["removed"] += int(parts[1])
                    except ValueError:
                        continue

        # Sort: (Added + Removed) descending
        commits.sort(key=lambda x: x["added"] + x["removed"], reverse=True)

        print(f"\n--- Commit Detail for '{target_author}' in {year} ---")
        print(f"{'Hash':<8} | {'Date':<12} | {'Added':>8} | {'Removed':>8}")
        print("-" * 45)

        for c in commits:
            if c["added"] + c["removed"] > 0:
                print(f"{c['hash']:<8} | {c['date']:<12} | {c['added']:>8} | {c['removed']:>8}")

    except subprocess.CalledProcessError as e:
        print(f"Error executing git: {e}")

def get_summary_stats(year):
    # (Existing summary logic from previous step)
    cmd = [
        "git", "log",
        f"--since={year}-01-01",
        f"--until={year}-12-31",
        "--numstat",
        "--pretty=format:AUTHOR:%aN",
        "--",
        "**/test_*.py",
        # "**/tests/*.py",
    ]
    result = subprocess.run(cmd, capture_output=True, text=True, check=True)
    stats = defaultdict(lambda: {"added": 0, "removed": 0})
    current_author = None
    for line in result.stdout.splitlines():
        if line.startswith("AUTHOR:"):
            current_author = line.replace("AUTHOR:", "")
        elif current_author and line.strip():
            parts = line.split()
            if len(parts) >= 3:
                try:
                    stats[current_author]["added"] += int(parts[0])
                    stats[current_author]["removed"] += int(parts[1])
                except ValueError: continue

    print(f"\n--- Summary for {year} ---")
    print(f"{'Author':<25} | {'Added':>10} | {'Removed':>10}")
    print("-" * 51)
    for author, counts in sorted(stats.items(), key=lambda x: x[1]['added'], reverse=True):
        print(f"{author[:25]:<25} | {counts['added']:>10} | {counts['removed']:>10}")

if __name__ == "__main__":
    year = sys.argv[1] if len(sys.argv) > 1 else "2025"
    author = sys.argv[2] if len(sys.argv) > 2 else None

    if author:
        get_detailed_author_stats(year, author)
    else:
        get_summary_stats(year)