#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""Fetch 前漢演義 chapters from zh.wikisource.org, convert to zh-cn, write Jekyll posts."""

from __future__ import annotations

import argparse
import json
import re
import sys
import time
import urllib.parse
import urllib.request
from pathlib import Path
from typing import List, Optional, Tuple

from bs4 import BeautifulSoup
import zhconv

UA = "KwanWaiPangBlogBot/1.0 (https://github.com/KwanWaiPang/kwanwaipang.github.io; zhconv+read-only)"

ROOT = Path(__file__).resolve().parents[1]
OUT_DIR = ROOT / "_posts" / "前汉演义"


def api_parse(page: str) -> str:
    q = urllib.parse.urlencode(
        {"action": "parse", "page": page, "prop": "text", "format": "json"}
    )
    url = f"https://zh.wikisource.org/w/api.php?{q}"
    req = urllib.request.Request(url, headers={"User-Agent": UA})
    with urllib.request.urlopen(req, timeout=60) as resp:
        data = json.loads(resp.read().decode("utf-8"))
    if "error" in data:
        raise RuntimeError(data["error"])
    return data["parse"]["text"]["*"]


def _parse_nav_subtitle(nav_text: str) -> Optional[str]:
    """从导航块文字解析副标题；维基文库各回版式略有出入。"""
    flat = re.sub(r"\s+", " ", nav_text.replace("\n", " "))
    patterns = [
        r"前漢演義\s*(第.+?回)\s*(.+?)\s*作者",
        r"前漢演義\s*(第.+?回)\s*(.+?)(?=\s*[◄►])",
        r"前漢演義\s*(第.+?回)\s*(.+?)(?=\s*→|\s*←)",
        r"前漢演義\s*(第.+?回)\s*(.+?)(?=\s*姊妹计划)",
    ]
    for pat in patterns:
        m = re.search(pat, flat)
        if m:
            return m.group(2).strip()
    return None


def extract_title_and_paragraphs(html: str) -> Tuple[str, List[str]]:
    soup = BeautifulSoup(html, "lxml")
    po = soup.select_one(".mw-parser-output")
    if not po:
        raise RuntimeError("no .mw-parser-output")

    subtitle = ""
    paragraphs: List[str] = []

    for child in po.children:
        name = getattr(child, "name", None)
        if not name:
            continue
        if name in ("div", "table") and not subtitle:
            nav_text = child.get_text(" ", strip=True)
            if "前漢演義" in nav_text and "回" in nav_text:
                sub = _parse_nav_subtitle(nav_text)
                if sub:
                    subtitle = sub
            continue
        if name == "p":
            t = child.get_text()
            if not t or not t.strip():
                continue
            paragraphs.append(t.strip())

    if not subtitle:
        blob = po.get_text(" ", strip=True)[:1200]
        subtitle = _parse_nav_subtitle(blob) or ""

    if not subtitle:
        raise RuntimeError("subtitle not found")

    return subtitle, paragraphs


def to_cn(s: str) -> str:
    return zhconv.convert(s, "zh-cn")


def cn_num(n: int) -> str:
    """阿拉伯数字转中文数字（用于 Jekyll title 中的回次，1–100）。"""
    digits = "零一二三四五六七八九"
    if n < 10:
        return digits[n]
    if n == 10:
        return "十"
    if n < 20:
        return "十" + digits[n - 10]
    if n < 100:
        tens, ones = divmod(n, 10)
        return digits[tens] + "十" + (digits[ones] if ones else "")
    if n == 100:
        return "一百"
    return str(n)


def build_markdown(chapter: int, subtitle_tw: str, paras_tw: List[str]) -> str:
    subtitle = to_cn(subtitle_tw)
    body_lines = []
    for p in paras_tw:
        body_lines.append(to_cn(p))
    body = "\n\n".join("　　" + ln.replace("\n", "") for ln in body_lines)

    title = f'读书笔记之——《前汉演义》第{cn_num(chapter)}回：{subtitle}'
    wikilink = f"https://zh.wikisource.org/wiki/%E5%89%8D%E6%BC%A2%E6%BC%94%E7%BE%A9/%E7%AC%AC{chapter:03d}%E5%9B%9E"

    return f"""---
layout: post
title: "{title}"
date: 2026-04-18
tags: [Books]
comments: true
author: kwanwaipang
# toc: false
excerpt: ""
---

<!-- * 参考资料：[蔡东藩《前汉演义》](https://zh.wikisource.org/wiki/%E5%89%8D%E6%BC%A2%E6%BC%94%E7%BE%A9) · [维基文库：本回全文]({wikilink}) -->


---

# 原文

{body}

---

# 译文

（译文生成中）
"""


def main(argv=None) -> None:
    p = argparse.ArgumentParser(description="Fetch 前漢演義 chapters from zh.wikisource.org")
    p.add_argument(
        "--start",
        type=int,
        default=26,
        help="起始回次（含）。默认 26 与历史批处理一致；全书请用 1",
    )
    p.add_argument(
        "--end",
        type=int,
        default=67,
        help="结束回次（含）。默认 67；全书至 100 请写 100",
    )
    args = p.parse_args(argv)
    start, end = args.start, args.end
    if start > end:
        print("error: --start must be <= --end", file=sys.stderr)
        sys.exit(1)

    OUT_DIR.mkdir(parents=True, exist_ok=True)
    for n in range(start, end + 1):
        page = f"前漢演義/第{n:03d}回"
        print("fetch", page)
        html = api_parse(page)
        sub, paras = extract_title_and_paragraphs(html)
        md = build_markdown(n, sub, paras)
        path = OUT_DIR / f"2026-04-18-前汉演义{n:02d}.md"
        path.write_text(md, encoding="utf-8")
        time.sleep(0.35)


if __name__ == "__main__":
    main()