#!/usr/bin/env python3
"""榆社县人民政府 - 通知公告 爬虫"""
import re, requests
from bs4 import BeautifulSoup
from datetime import datetime
import sqlite3, os

BASE_URL = "http://www.yushe.gov.cn"
SITE_NAME = "yushe.gov.cn-通知公告"
DB_PATH = os.getenv("SEARCH_DB", "/root/search.db")
CATEGORY = "eia"
HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
CUTOFF_DATE = "2023-06-16"
CUTOFF_INT = int(CUTOFF_DATE.replace("-", ""))


def get_conn():
    conn = sqlite3.connect(DB_PATH, timeout=60)
    conn.row_factory = sqlite3.Row
    return conn


def clean_text(text):
    if not text:
        return ""
    return re.sub(r'\s+', ' ', text.strip())


def init_session():
    """初始化session, 处理SessionVerify重定向"""
    s = requests.Session()
    r = s.get(f"{BASE_URL}/xwzx/tzgg/", headers=HEADERS, timeout=15, verify=False)
    m = re.search(r"href\s*=\s*'([^']+SessionVerify=[^']+)'", r.text)
    if m:
        s.get(BASE_URL + m.group(1), headers=HEADERS, timeout=15, verify=False)
    return s


def get_list_url(page_num):
    """获取列表页URL, page_num从1开始"""
    if page_num == 1:
        return f"{BASE_URL}/xwzx/tzgg/"
    return f"{BASE_URL}/xwzx/tzgg_{page_num}"


def extract_articles(session, page_num):
    """从列表页提取文章"""
    url = get_list_url(page_num)
    try:
        resp = session.get(url, headers=HEADERS, timeout=30, verify=False)
        resp.raise_for_status()
        resp.encoding = 'utf-8'
    except Exception as e:
        print(f"  [WARN] 列表页({page_num})获取失败: {e}", flush=True)
        return []

    soup = BeautifulSoup(resp.text, 'html.parser')
    articles = []

    for a in soup.find_all('a', class_='tit', href=True):
        href = a['href']
        txt = a.get_text(strip=True)
        if txt and len(txt) > 5:
            full_url = BASE_URL + href if href.startswith('/') else href
            # 取列表页上可见的日期
            parent_li = a.find_parent('li')
            date_str = ''
            if parent_li:
                m = re.search(r'(\d{4}-\d{2}-\d{2})', parent_li.get_text())
                if m:
                    date_str = m.group(1)
            articles.append({
                'url': full_url,
                'title': txt,
                'date_str': date_str,
            })

    return articles


def extract_detail(session, article):
    """获取详情"""
    url = article['url']
    try:
        resp = session.get(url, headers=HEADERS, timeout=30, verify=False)
        resp.raise_for_status()
        resp.encoding = 'utf-8'
    except Exception as e:
        print(f"  [WARN] 详情页获取失败({article['title'][:30]}...): {e}", flush=True)
        return None, None, None, None

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

    # 标题 - 从<title>标签
    title_tag = soup.select_one('title')
    title = ''
    if title_tag:
        title = re.sub(r'_通知公告_新闻中心.*$', '', title_tag.get_text(strip=True)).strip()
        title = re.sub(r'_榆社县人民政府门户网站$', '', title).strip()
    if not title:
        title = article['title']

    # 日期
    pub_date = ''
    for meta in soup.find_all('meta'):
        name = (meta.get('name') or '').lower()
        if name in ['publishdate', 'pubdate', 'date']:
            m = re.match(r'(\d{4}-\d{2}-\d{2})', meta.get('content', ''))
            if m:
                pub_date = m.group(1)
                break
    if not pub_date:
        m = re.search(r'(\d{4}-\d{2}-\d{2})', html)
        if m:
            pub_date = m.group(1)

    # 正文 - .article
    content = ''
    content_el = soup.select_one('.article')
    if not content_el:
        content_el = soup.select_one('.content')
    if content_el:
        content_html = str(content_el)
        # 图片base64内嵌
        img_headers = {**HEADERS, 'Referer': BASE_URL + '/'}
        def embed_img(m):
            src = m.group(1)
            if src.startswith('/'):
                src = BASE_URL + src
            elif src.startswith('..'):
                src = BASE_URL + src[2:]
            elif not src.startswith('http'):
                dir_part = url.rsplit('/', 1)[0]
                src = dir_part + '/' + src
            return f'<img src=\"{src}\"' 
        content_html = re.sub(r'<img[^>]+?\bsrc=[\"\']([^\"\']+)[\"\']', embed_img, content_html)
        content = content_html

    summary = clean_text(content[:200]) if content else ''

    return title, pub_date, content, summary


def main():
    print(f"=== {SITE_NAME} 爬虫 ===", flush=True)
    print(f"数据库: {DB_PATH}", flush=True)

    session = init_session()
    conn = get_conn()
    cur = conn.cursor()

    total_new = 0
    total_skip = 0
    total_error = 0

    for page_num in range(1, 13):  # 1-12
        articles = extract_articles(session, page_num)
        if not articles:
            continue
        print(f"  第{page_num}页: {len(articles)}条文章", flush=True)

        for article in articles:
            # 日期过滤
            if article['date_str']:
                date_int = int(article['date_str'].replace('-', ''))
                if date_int < CUTOFF_INT:
                    total_skip += 1
                    continue

            # 获取详情
            title, pub_date, content, summary = extract_detail(session, article)
            if not title or not content:
                total_skip += 1
                continue
            if not pub_date:
                total_error += 1
                continue

            date_rank = int(pub_date.replace('-', ''))
            source_url = f"{BASE_URL}/xwzx/tzgg/"

            try:
                cur.execute(
                    "INSERT OR IGNORE INTO gov_raw (page_url, title, content, site_name, publish_date, summary, category, date_rank, source_url) VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?)",
                    (article['url'], title, content, SITE_NAME, pub_date, summary, CATEGORY, date_rank, source_url)
                )
                conn.commit()
                total_new += 1
                print(f"  [OK] {title[:40]}... | {pub_date}", flush=True)
            except Exception as e:
                print(f"  [ERROR] 入库失败: {title[:30]}... - {e}", flush=True)
                conn.rollback()
                total_error += 1

    conn.close()
    print(f"\n=== 完成: 新增 {total_new}, 跳过 {total_skip}, 错误 {total_error} ===", flush=True)


if __name__ == '__main__':
    main()