#!/usr/bin/env python3
"""
知乎回答全文抓取脚本 (bypass zse-ck)
用法:
  python3 zhihu_fetch.py <answer_url或answer_id> [output.txt]
示例:
  python3 zhihu_fetch.py https://www.zhihu.com/question/24326030/answer/2001766056285467052
  python3 zhihu_fetch.py 2001766056285467052 /tmp/out.txt

原理: 用 playwright 真实浏览器渲染, 先访问首页种 cookie(d_c0), 再访问回答页,
等待 zse-ck JS 验证完成, 点击"阅读全文"展开, 提取 .RichText 正文。
需要: pip3 install playwright --break-system-packages
浏览器: /root/.cache/ms-playwright/chromium-1217/chrome-linux64/chrome (已存在)
"""
import sys, re, json, argparse
from playwright.sync_api import sync_playwright

CHROME = "/root/.cache/ms-playwright/chromium-1217/chrome-linux64/chrome"
UA = ("Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
      "(KHTML, like Gecko) Chrome/126.0.0.0 Safari/537.36")

def extract_answer_id(url_or_id: str) -> str:
    m = re.search(r"/answer/(\d+)", url_or_id)
    if m:
        return m.group(1)
    if url_or_id.isdigit():
        return url_or_id
    raise ValueError("无法解析 answer id: " + url_or_id)

def fetch(url_or_id: str, timeout_ms: int = 45000) -> dict:
    aid = extract_answer_id(url_or_id)
    # 从链接中提取 question id（如有），否则用占位符（知乎按 answer_id 定位，question id 不影响）
    mq = re.search(r"/question/(\d+)", url_or_id)
    qid = mq.group(1) if mq else "0"
    url = f"https://www.zhihu.com/question/{qid}/answer/{aid}"
    # 尝试通过 API 拿元数据(快), 正文用浏览器
    result = {"answer_id": aid, "url": url, "text": "", "html_len": 0, "error": None}
    with sync_playwright() as p:
        browser = p.chromium.launch(
            executable_path=CHROME,
            headless=True,
            args=["--no-sandbox", "--disable-dev-shm-usage",
                  "--disable-blink-features=AutomationControlled"],
        )
        ctx = browser.new_context(user_agent=UA, viewport={"width": 1280, "height": 900}, locale="zh-CN")
        page = ctx.new_page()
        try:
            # 第一步: 访问首页种 cookie (d_c0 匿名token)
            page.goto("https://www.zhihu.com/", timeout=30000, wait_until="domcontentloaded")
            page.wait_for_timeout(5000)
        except Exception as e:
            result["error"] = f"home: {e}"
        try:
            # 第二步: 访问回答页
            page.goto(url, timeout=timeout_ms, wait_until="domcontentloaded")
            page.wait_for_timeout(12000)
            # 第三步: 点"阅读全文"展开
            for sel in ["button:has-text('阅读全文')", "button:has-text('展开')",
                        ".RichContent-collapsedText"]:
                try:
                    btn = page.query_selector(sel)
                    if btn:
                        btn.click(timeout=3000)
                        page.wait_for_timeout(3000)
                except Exception:
                    pass
            # 第四步: 提取正文
            rich = page.query_selector(".RichText, .RichContent-inner, .Post-RichTextContainer")
            if rich:
                result["text"] = rich.inner_text()
            html = page.content()
            result["html_len"] = len(html)
        except Exception as e:
            result["error"] = f"page: {e}"
        browser.close()
    return result

if __name__ == "__main__":
    ap = argparse.ArgumentParser()
    ap.add_argument("url_or_id", help="知乎回答链接或 answer_id")
    ap.add_argument("output", nargs="?", default=None, help="输出文件路径(默认打印到stdout)")
    args = ap.parse_args()
    r = fetch(args.url_or_id)
    if r["error"]:
        print("WARN:", r["error"], file=sys.stderr)
    if r["text"]:
        if args.output:
            with open(args.output, "w", encoding="utf-8") as f:
                f.write(r["text"])
            print(f"已保存 {len(r['text'])} 字符 -> {args.output}", file=sys.stderr)
        else:
            print(r["text"])
    else:
        print("FAILED: 未提取到正文, html_len=", r["html_len"], file=sys.stderr)
        sys.exit(1)