#!/usr/bin/env python3
"""昌图县人民政府 - 公示公告 爬虫
慧蓝/jQuery UI CMS, 静态HTML分页, 15条/页, 83页
列表: <li class="mod_page_content_li"><a href="...">title</a><span class="mod_page_content_time">date</span>
详情: <div class="article_content" id="easysiteText"> 正文 (含HTML p标签)
"""

import os
import re
import sys
import urllib.request
from datetime import datetime, timedelta

DB_PATH = os.getenv("SEARCH_DB", "/root/search.db")
SITE_NAME = "changtu.gov.cn-公示公告"
BASE_URL = "https://www.changtu.gov.cn"
LIST_PREFIX = "/changtu/ywdt/gsgg44/77133ae2"
PAGE_SIZE = 15

HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
CUTOFF_DATE = (datetime.now() - timedelta(days=3*365)).strftime("%Y-%m-%d")
print(f"[info] 日期过滤: >= {CUTOFF_DATE}")


def fetch_list_page(page_no):
    """获取列表页HTML"""
    url = f"{BASE_URL}{LIST_PREFIX}-{page_no}.html"
    try:
        req = urllib.request.Request(url, headers=HEADERS)
        resp = urllib.request.urlopen(req, timeout=30)
        return resp.read().decode("utf-8")
    except Exception as e:
        print(f"[error] 第{page_no}页请求失败: {e}")
        return None


def parse_list_items(html):
    """解析列表HTML，提取标题、URL、日期"""
    items = re.findall(
        r'<li class="mod_page_content_li">\s*<span class="mod_page_content_time fr">([^<]+)</span>\s*<p class="mod_page_content_title"><em[^>]*>.*?</em><a href="([^"]+)"[^>]*>([^<]+)</a></p>\s*</li>',
        html
    )
    results = []
    for date, url, title in items:
        if url.startswith("/"):
            url = BASE_URL + url
        results.append({"title": title.strip(), "url": url, "date": date.strip()})
    return results


def fetch_detail_content(url):
    """从详情页提取正文HTML和完整标题（列表页标题可能被截断）"""
    try:
        req = urllib.request.Request(url, headers=HEADERS)
        resp = urllib.request.urlopen(req, timeout=30)
        html = resp.read().decode("utf-8")
    except Exception as e:
        print(f"  [warn] 详情页请求失败: {url[-60:]}, {e}")
        return "", ""

    # Full title from meta or div (列表页标题可能被CSS截断带...)
    full_title = ""
    meta_t = re.search(r'<meta name="ArticleTitle" content="([^"]+)"', html)
    if meta_t:
        full_title = meta_t.group(1).strip()
    if not full_title:
        div_t = re.search(r'<div class="article_title">([^<]+)</div>', html)
        if div_t:
            full_title = div_t.group(1).strip()

    # Main content: <div class="article_content" id="easysiteText"> ... </div> <div class="print_close"
    m = re.search(
        r'<div class="article_content" id="easysiteText">([\s\S]*?)</div>\s*<div class="print_close',
        html
    )
    if m:
        content = m.group(1).strip()
        content = re.sub(r'\n\s*\n', '\n', content)
        return content, full_title

    return "", full_title


def main():
    import sqlite3

    conn = sqlite3.connect(DB_PATH, timeout=60)
    c = conn.cursor()
    c.execute("PRAGMA journal_mode=WAL")

    # Determine total pages by checking which pages exist
    # Known: max page 83 from analysis
    total_pages = 83
    print(f"[info] 总页数: {total_pages}, 每页: {PAGE_SIZE}")

    new_count = 0
    old_count = 0
    dup_count = 0
    err_count = 0

    for page in range(1, total_pages + 1):
        html = fetch_list_page(page)
        if not html:
            print(f"[warn] 第{page}页获取失败，跳过")
            continue

        items = parse_list_items(html)
        if not items:
            print(f"[warn] 第{page}页未解析到条目，跳过")
            continue

        page_new = 0
        for item in items:
            title = item["title"]
            url = item["url"]
            date = item["date"]

            # Date filter
            if date and date < CUTOFF_DATE:
                old_count += 1
                continue

            # Check duplicate
            c.execute("SELECT COUNT(*) FROM gov_raw WHERE page_url=?", (url,))
            if c.fetchone()[0] > 0:
                dup_count += 1
                continue

            content, full_title = fetch_detail_content(url)
            if not content:
                content = "[无正文] 附件型或图片型条目，请查看原文链接"

            # Use full title from detail page if available (list page may truncate with ...)
            final_title = full_title if full_title else title

            try:
                c.execute("""INSERT OR IGNORE INTO gov_raw
                    (page_url, title, content, site_name, publish_date, summary, category)
                    VALUES (?, ?, ?, ?, ?, ?, ?)""",
                    (url, final_title, content, SITE_NAME, date, final_title, "公示公告"))
                if c.rowcount > 0:
                    page_new += 1
            except Exception as e:
                print(f"  [error] 入库失败: {e}")
                err_count += 1

        conn.commit()
        new_count += page_new
        print(f"[page {page}/{total_pages}] +{page_new} new, cumulative: {new_count}")

    conn.close()
    print(f"\n[完成] {SITE_NAME}: 新增 {new_count} 条, 旧跳过 {old_count}, 重复 {dup_count}, 错误 {err_count}")


if __name__ == "__main__":
    main()