#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
git-author-stats

统计 Git 仓库中每个作者的提交次数并生成排名。

用法:
    git_author_stats.py <仓库路径> [--branch 分支] [--top N] [--percent]
"""

import argparse
import os
import subprocess
import sys
from collections import Counter


def run_git_log(repo_path, branch=None):
    """调用 git log 获取作者列表。

    使用 --format=%aN 只输出作者名字，每行一个。
    返回作者名字的列表。
    """
    cmd = ["git", "-C", repo_path, "log", "--format=%aN"]
    if branch:
        cmd.append(branch)

    try:
        result = subprocess.run(
            cmd,
            stdout=subprocess.PIPE,
            stderr=subprocess.PIPE,
            check=True,
            text=True,
            encoding="utf-8",
            errors="replace",
        )
    except FileNotFoundError:
        print("错误: 未找到 git 命令，请先安装 Git。", file=sys.stderr)
        sys.exit(1)
    except subprocess.CalledProcessError as e:
        print("错误: 执行 git log 失败。", file=sys.stderr)
        if e.stderr:
            print(e.stderr.strip(), file=sys.stderr)
        sys.exit(1)

    authors = [line.strip() for line in result.stdout.splitlines() if line.strip()]
    return authors


def count_authors(authors):
    """统计每位作者的提交次数，返回按次数降序排列的列表。"""
    counter = Counter(authors)
    # 按提交次数降序，次数相同按作者名字升序，保证结果稳定
    return sorted(counter.items(), key=lambda item: (-item[1], item[0]))


def format_table(rows, total, show_percent):
    """格式化输出表格。"""
    # 计算列宽
    name_width = max([len(name) for name, _ in rows] + [len("作者")]) + 2
    rank_width = max(len(str(len(rows))), len("排名")) + 2
    count_width = max([len(str(c)) for _, c in rows] + [len("提交次数")]) + 2

    header = (
        f"{'排名':<{rank_width}}"
        f"{'作者':<{name_width}}"
        f"{'提交次数':<{count_width}}"
    )
    if show_percent:
        header += "占比"

    lines = [header, "-" * (rank_width + name_width + count_width + (10 if show_percent else 0))]

    for idx, (name, count) in enumerate(rows, start=1):
        line = (
            f"{idx:<{rank_width}}"
            f"{name:<{name_width}}"
            f"{count:<{count_width}}"
        )
        if show_percent:
            percent = (count / total * 100) if total else 0.0
            line += f"{percent:.2f}%"
        lines.append(line)

    return "\n".join(lines)


def parse_args(argv=None):
    parser = argparse.ArgumentParser(
        prog="git-author-stats",
        description="统计 Git 仓库中每个作者的提交次数并生成排名。",
    )
    parser.add_argument("repo_path", help="Git 仓库路径")
    parser.add_argument(
        "-b", "--branch",
        default=None,
        help="指定分支（默认当前分支）",
    )
    parser.add_argument(
        "-t", "--top",
        type=int,
        default=None,
        help="只显示前 N 名作者",
    )
    parser.add_argument(
        "-p", "--percent",
        action="store_true",
        help="显示每位作者的提交占比",
    )
    return parser.parse_args(argv)


def main(argv=None):
    args = parse_args(argv)

    repo_path = os.path.abspath(args.repo_path)
    if not os.path.isdir(repo_path):
        print(f"错误: 路径不存在或不是目录: {args.repo_path}", file=sys.stderr)
        sys.exit(1)

    # 检查是否是 Git 仓库
    if not os.path.isdir(os.path.join(repo_path, ".git")):
        # 也可能是在 worktree 中，.git 是文件
        git_marker = os.path.join(repo_path, ".git")
        if not os.path.exists(git_marker):
            print(f"错误: 该路径不是 Git 仓库: {args.repo_path}", file=sys.stderr)
            sys.exit(1)

    authors = run_git_log(repo_path, args.branch)
    if not authors:
        print("该仓库没有任何提交记录。")
        return

    ranking = count_authors(authors)
    total = len(authors)

    if args.top is not None and args.top > 0:
        ranking = ranking[: args.top]

    branch_display = args.branch if args.branch else "(当前分支)"
    print(f"仓库: {repo_path}")
    print(f"分支: {branch_display}")
    print(f"总提交数: {total}")
    print()
    print(format_table(ranking, total, args.percent))


if __name__ == "__main__":
    main()