#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# 织金县 - 生态环境
# CMS: 多彩博虹/TRS
# List: index.html + index_{N}.html (17 pages, 21/page, ~322 total)
# Content: div.TRS_Editor or div.trs_editor_view

import requests
import re
import sqlite3
from bs4 import BeautifulSoup
from datetime import datetime

BASE_URL = "https://www.gzzhijin.gov.cn/zwgk/zdlyxx/sthj"
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36"
}
DB_PATH = "/root/search.db"
MAX_PAGES = 5  # Only top 5 pages for daily run

def get_pages():
    """Get all article URLs from list pages (first MAX_PAGES pages)"""
    articles = []
    # createPageHTML(17, 0, "index", "html", 322)
    # page 0 = index.html, page 1 = index_1.html, ..., page 16 = index_16.html
    for page_num in range(MAX_PAGES):
        if page_num == 0:
            url = f"{BASE_URL}/index.html"
        else:
            url = f"{BASE_URL}/index_{page_num}.html"
        
        try:
            r = requests.get(url, headers=HEADERS, verify=False, timeout=30)
            r.encoding = "utf-8"
            if r.status_code != 200:
                print(f"  [SKIP] Page {page_num+1}: HTTP {r.status_code}")
                continue
            
            soup = BeautifulSoup(r.text, "html.parser")
            ul = soup.find("ul", class_="NewsList")
            if not ul:
                print(f"  [SKIP] Page {page_num+1}: No NewsList found")
                continue
            
            count = 0
            for li in ul.find_all("li"):
                a = li.find("a")
                span = li.find("span")
                if a and span:
                    href = a.get("href", "")
                    title = a.get_text(strip=True)
                    date_str = span.get_text(strip=True)[:10]  # "2026-06-26"
                    
                    if href and title and date_str:
                        articles.append({
                            "url": href,
                            "title": title,
                            "date": date_str
                        })
                        count += 1
            
            print(f"  [OK] Page {page_num+1}: {count} items")
            
        except Exception as e:
            print(f"  [ERR] Page {page_num+1}: {e}")
    
    return articles

def get_detail(url):
    """Fetch detail page and extract content and date"""
    try:
        r = requests.get(url, headers=HEADERS, verify=False, timeout=30)
        r.encoding = "utf-8"
        if r.status_code != 200 or len(r.text) < 1000:
            return None, None
        
        soup = BeautifulSoup(r.text, "html.parser")
        
        # Content: try multiple selectors
        content = None
        
        # 1. TRS_Editor
        div = soup.find("div", class_="TRS_Editor")
        if div:
            content = "".join(str(tag) for tag in div.contents).strip()
        
        # 2. trs_editor_view
        if not content:
            div = soup.find("div", class_=re.compile(r"trs_editor_view"))
            if div:
                content = "".join(str(tag) for tag in div.contents).strip()
        
        # 3. Article_Con
        if not content:
            div = soup.find("div", class_="Article_Con")
            if div:
                content = "".join(str(tag) for tag in div.contents).strip()
        
        if not content:
            return None, None
        
        text_only = re.sub(r'<[^>]+>', '', content).strip()
        if len(text_only) < 50:
            return None, None
        
        # Date from PubDate meta
        pub_date = ""
        meta = soup.find("meta", attrs={"name": "PubDate"})
        if meta and meta.get("content"):
            pub_date = meta["content"].strip()[:10]
        
        return content, pub_date
    
    except Exception as e:
        print(f"  [ERR] {url[:80]}: {e}")
        return None, None

def main():
    print("=" * 60)
    print("织金县 - 生态环境 爬虫")
    print(f"启动时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
    print("=" * 60)
    
    # Step 1: Get list
    print("\n[1/3] 抓取列表页 (前5页)...")
    articles = get_pages()
    print(f"  共找到 {len(articles)} 条记录")
    
    if not articles:
        print("  未找到任何记录，退出")
        return
    
    # Step 2: Check DB for existing URLs
    print("\n[2/3] 检查数据库去重...")
    conn = sqlite3.connect(DB_PATH, timeout=60)
    c = conn.cursor()
    
    c.execute("SELECT page_url FROM gov_raw")
    existing = set(row[0] for row in c.fetchall())
    
    new_articles = [a for a in articles if a["url"] not in existing]
    print(f"  新增: {len(new_articles)}/{len(articles)} (已存在: {len(existing)})")
    
    if not new_articles:
        print("  没有新记录，跳过详情抓取")
        conn.close()
        return
    
    # Step 3: Fetch details
    print("\n[3/3] 抓取详情页...")
    imported = 0
    skipped = 0
    
    for i, article in enumerate(new_articles, 1):
        title_short = article["title"][:45].replace("\n", " ")
        print(f"  [{i}/{len(new_articles)}] {title_short}...", end=" ", flush=True)
        
        content, pub_date = get_detail(article["url"])
        
        if not content:
            print("✗ 无正文")
            skipped += 1
            continue
        
        date = pub_date or article["date"]
        
        text_only = re.sub(r'<[^>]+>', '', content).strip()
        summary = text_only[:200]
        
        try:
            c.execute("""INSERT OR IGNORE INTO gov_raw 
                (title, content, page_url, summary, site_name, publish_date, category)
                VALUES (?, ?, ?, ?, ?, ?, ?)""",
                (article["title"], content, article["url"], summary,
                 "织金县", date, "环评公示"))
            if c.rowcount > 0:
                conn.commit()
                imported += 1
                print(f"✓ ({len(text_only)}字)")
            else:
                print(f"• 已存在")
        except Exception as e:
            print(f"✗ DB错误: {e}")
            conn.rollback()
    
    conn.close()
    print(f"\n{'=' * 60}")
    print(f"导入完成: {imported} 条 (跳过 {skipped} 条)")
    print(f"{'=' * 60}")

if __name__ == "__main__":
    main()