#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
蚌埠市淮上区人民政府 - 通知公告 爬虫
CMS: 龙讯科技 (Lonsun)，知道创宇CloudWAF（间歇性）
列表: /xwzx/tzgg/index.html (page1), /content/column/6797061?pageIndex=N (N>=2)
       15条/页，共约277页4150条
详情: /xwzx/tzgg/{id}.html
       正文: div.ls-article-info.j-fontContent (段落/表格/附件链接)
       标题: meta ArticleTitle / h1.ls-article-title
       日期: meta PubDate / span发布时间
日跑: --max-pages 1 (增量取最新页)
"""

import sys, os, re, time, sqlite3, argparse, json, subprocess
from datetime import datetime
from urllib.parse import urljoin
import requests
from bs4 import BeautifulSoup
import warnings
warnings.filterwarnings("ignore")

BASE_URL = "https://www.huaishang.gov.cn"
SITE_NAME = "蚌埠市淮上区人民政府"
COLUMN_URL = "https://www.huaishang.gov.cn/xwzx/tzgg/index.html"
API_URL = "https://www.huaishang.gov.cn/content/column/6797061?pageIndex={}"
PAGE_SIZE = 15
GROUP = "安徽省"
DB_PATH = "/mnt/data/search.db"

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
    "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
    "Accept-Language": "zh-CN,zh;q=0.9",
}


def log(msg):
    print(f"[{datetime.now().strftime('%H:%M:%S')}] {msg}")


def get_page(url, session=None, retries=5):
    """获取页面，含WAF重试"""
    for i in range(retries):
        try:
            s = session or requests.Session()
            resp = s.get(url, headers=dict(HEADERS, Referer="https://www.huaishang.gov.cn/"), timeout=30)
            if resp.status_code == 200:
                resp.encoding = 'utf-8'
                return resp.text
            if resp.status_code == 403:
                log(f"  CloudWAF触发 403，等待 {3*(i+1)}s 重试 ({i+1}/{retries})")
                time.sleep(3 * (i + 1))
                continue
            log(f"  HTTP {resp.status_code}，重试 ({i+1}/{retries})")
            time.sleep(2)
        except Exception as e:
            log(f"  请求异常: {e}，重试 ({i+1}/{retries})")
            time.sleep(3)
    return None


def parse_list(html):
    """解析列表页，返回 [{url, title, date}]"""
    items = []
    soup = BeautifulSoup(html, 'html.parser')
    ul = soup.select_one('ul.doc_list')
    if not ul:
        log("  ⚠️ 未找到 ul.doc_list")
        return items

    for li in ul.find_all('li', recursive=False):
        if 'lm_line' in (li.get('class') or []):
            continue
        a = li.find('a')
        if not a or not a.get('href'):
            continue
        title = (a.get('title') or '').strip()
        if not title:
            span = a.find('span')
            if span:
                title = span.get_text(strip=True)
        if not title:
            continue
        href = a['href'].strip()
        if not href.startswith('http'):
            href = urljoin(BASE_URL, href)
        date_span = li.find('span', class_='date')
        pub_date = (date_span.get_text(strip=True) or '') if date_span else ''
        items.append({'url': href, 'title': title, 'pub_date': pub_date})
    return items


def fetch_detail(url, session=None):
    """获取详情页信息"""
    html = get_page(url, session)
    if not html:
        return None

    soup = BeautifulSoup(html, 'html.parser')

    # 标题
    title = ''
    mt = soup.find('meta', attrs={'name': 'ArticleTitle'})
    if mt and mt.get('content'):
        title = mt['content'].strip()
    if not title:
        h1 = soup.select_one('h1.ls-article-title')
        if h1:
            title = h1.get_text(strip=True)

    # 日期
    pub_date = ''
    md = soup.find('meta', attrs={'name': 'PubDate'})
    if md and md.get('content'):
        pub_date = md['content'].strip()

    # 来源
    source = ''
    ms = soup.find('meta', attrs={'name': 'ContentSource'})
    if ms and ms.get('content'):
        source = ms['content'].strip()

    # 正文
    content_div = soup.select_one('div.ls-article-info.j-fontContent')

    attachments = []
    content_text = ''
    has_table = False

    if content_div:
        html_str = str(content_div)
        # 提取附件
        for a in content_div.find_all('a'):
            href = a.get('href', '')
            text = a.get_text(strip=True) or '附件'
            if re.search(r'\.(pdf|doc|docx|xls|xlsx|zip|rar|7z)(\?|#|$)', href, re.I):
                if not href.startswith('http'):
                    href = urljoin(BASE_URL, href)
                attachments.append({'url': href, 'text': text})

        has_table = bool(content_div.find('table'))

        # 生成纯文本
        parts = []
        for elem in content_div.children:
            name = getattr(elem, 'name', None)
            if name in ('script', 'style'):
                continue
            if name == 'table':
                rows = []
                for tr in elem.find_all('tr'):
                    cells = [td.get_text(strip=True) for td in tr.find_all(['td', 'th'])]
                    rows.append(' | '.join(cells))
                if rows:
                    parts.append('\n'.join(rows))
            elif name == 'p':
                txt = elem.get_text('\n', strip=True)
                if txt:
                    parts.append(txt)
            elif name == 'div':
                txt = elem.get_text('\n', strip=True)
                if txt:
                    parts.append(txt)
            elif name is None:
                txt = str(elem).strip()
                if txt:
                    parts.append(txt)

        content_text = '\n\n'.join(p for p in parts if p)
        content_text = re.sub(r'\n{3,}', '\n\n', content_text).strip()

    # 纯附件页
    if not content_text and attachments:
        links = '\n'.join(f"[{a['text']}]({a['url']})" for a in attachments)
        content_text = f"附件下载：\n{links}"

    return {
        'title': title,
        'pub_date': pub_date,
        'source': source,
        'content_text': content_text,
        'has_table': has_table,
        'attachments': attachments,
    }


def init_db():
    """初始化数据库连接"""
    conn = sqlite3.connect(DB_PATH)
    c = conn.cursor()
    return conn, c


def insert_article(conn, cur, data):
    """插入文章到gov_raw，同步写gov_search"""
    try:
        content = data.get('content', '') or ''
        summary = content[:200] if content else ''
        attachments = data.get('attachments', '') or ''
        has_table = data.get('has_table', 0)

        cur.execute('''INSERT OR IGNORE INTO gov_raw
            (page_url, title, publish_date, content, site_name, source_url,
             group_name, industry, attachments, has_table, summary)
            VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)''', (
            data['url'],
            data['title'],
            data.get('pub_date', ''),
            content,
            data.get('site_name', SITE_NAME),
            data.get('url', ''),
            data.get('group_name', GROUP),
            'other',
            attachments,
            has_table,
            summary,
        ))
        if cur.rowcount == 0:
            return False  # 已存在

        rowid = cur.lastrowid
        # 同步写 FTS
        try:
            cur.execute('INSERT INTO gov_search (rowid, title, site_name, summary) VALUES (?, ?, ?, ?)',
                        (rowid, data['title'], data.get('site_name', SITE_NAME), summary))
        except sqlite3.OperationalError:
            conn.commit()
            _fts_fallback(rowid, data['title'], data.get('site_name', SITE_NAME), summary)
        conn.commit()
        return True
    except Exception as e:
        log(f"  入库异常: {e}")
        conn.rollback()
        return False


def _fts_fallback(rowid, title, site_name, summary):
    """FTS损坏时用subprocess写入"""
    try:
        subprocess.run([
            'sqlite3', DB_PATH,
            f"INSERT OR REPLACE INTO gov_search (rowid, title, site_name, summary) VALUES ({rowid}, '{title.replace(chr(39), chr(39)+chr(39))}', '{site_name.replace(chr(39), chr(39)+chr(39))}', '{summary.replace(chr(39), chr(39)+chr(39))}')"
        ], capture_output=True, timeout=10)
    except Exception as e:
        log(f"  FTS fallback失败: {e}")


def crawl(max_pages=5, local=False):
    """主爬虫"""
    global DB_PATH
    db_path_local = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'huaishang_test.db') if local else DB_PATH
    if local:
        DB_PATH = db_path_local

    log(f"开始爬取 {SITE_NAME} - 通知公告，{max_pages}页")
    if local:
        log(f"本地测试模式，DB: {db_path_local}")

    conn, cur = init_db()
    session = requests.Session()
    session.headers.update(HEADERS)

    # 第1阶段：收集所有列表页URL
    all_articles = []
    for page in range(1, max_pages + 1):
        url = COLUMN_URL if page == 1 else API_URL.format(page)
        log(f"列表页 {page}: {url}")
        html = get_page(url, session)
        if not html:
            log(f"  ❌ 第{page}页获取失败，停止翻页")
            break
        items = parse_list(html)
        if not items:
            log(f"  ⚠️ 第{page}页无文章，停止翻页")
            break
        log(f"  ✅ {len(items)} 条")
        all_articles.extend(items)
        time.sleep(1.5)

    log(f"\n总计 {len(all_articles)} 条待爬详情")

    # 第2阶段：爬详情入库
    new_count = 0
    skip_count = 0
    for idx, art in enumerate(all_articles, 1):
        log(f"[{idx}/{len(all_articles)}] {art['title'][:50]}...")
        detail = fetch_detail(art['url'], session)
        if not detail:
            log(f"  ❌ 详情获取失败")
            skip_count += 1
            time.sleep(1)
            continue

        final_title = art['title'] or detail['title']
        final_date = art['pub_date'] or detail['pub_date']
        attachments_str = json.dumps(detail.get('attachments', []), ensure_ascii=False)

        ok = insert_article(conn, cur, {
            'title': final_title,
            'url': art['url'],
            'pub_date': final_date,
            'content': detail['content_text'],
            'site_name': SITE_NAME,
            'group_name': GROUP,
            'attachments': attachments_str,
            'has_table': 1 if detail.get('has_table') else 0,
        })
        if ok:
            new_count += 1
            log(f"  ✅ 入库")
        else:
            skip_count += 1
            log(f"  ⏭️ 已存在")
        time.sleep(0.8)

    conn.close()
    log(f"\n{'='*50}")
    log(f"完成! 新增: {new_count}, 跳过: {skip_count}")
    return new_count


if __name__ == '__main__':
    parser = argparse.ArgumentParser(description='蚌埠市淮上区通知公告爬虫')
    parser.add_argument('--max-pages', type=int, default=5, help='最大爬取页数')
    parser.add_argument('--local', action='store_true', help='本地测试模式')
    args = parser.parse_args()
    crawl(max_pages=args.max_pages, local=args.local)