#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
crawl_guanyun_ghjh.py - 灌云县-规划计划 (ghjh 栏目)
URL: http://www.guanyun.gov.cn/gyxzf/ghjh/ghjh.html
CMS: 灌云 gkml/Hanweb (同站群), 列表单页含全量(无翻页)
列表: div#initData ul.xxgk-list-con > li(style=line-height:40px) > img + a[title] + span日期
详情: /gyxzf/ghjh/content/{uuid}.html
标题: meta ArticleTitle
日期: meta PubDate
正文: div.content.pd25
"""
import sys, os, re, time, html as html_lib
from datetime import datetime, timezone, timedelta

sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
from crawler_lib import push_to_searchdb
import requests
from bs4 import BeautifulSoup
from urllib.parse import urljoin

_MAX_PG = 5
for i, a in enumerate(sys.argv):
    if a == '--pages' and i + 1 < len(sys.argv):
        _MAX_PG = int(sys.argv[i + 1])
        break

SITE_NAME = "灌云县-规划计划"
BASE = "http://www.guanyun.gov.cn"
LIST_PATH = "/gyxzf/ghjh"
CUTOFF = (datetime.now(timezone.utc) - timedelta(days=365 * 3)).strftime("%Y-%m-%d")
HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) Chrome/120.0 Safari/537.36"}

LIST_RE = re.compile(r'<li[^>]*style="line-height: 40px[^"]*">.*?<a href="([^"]+)"[^>]*title="([^"]+)"[^>]*>.*?</a><span[^>]*>\s*(\d{4}-\d{2}-\d{2})\s*</span>', re.S)


def fetch(url, timeout=30):
    for _ in range(3):
        try:
            r = requests.get(url, headers=HEADERS, timeout=timeout, verify=False)
            r.encoding = "utf-8"
            return r.text
        except Exception:
            time.sleep(2)
    return None


def parse_list(html):
    items = []
    for m in LIST_RE.finditer(html):
        href, title, date = m.group(1).strip(), html_lib.unescape(m.group(2)).strip(), m.group(3).strip()
        if not href.startswith("http"):
            href = urljoin(BASE, href)
        if title and date:
            items.append((title, href, date))
    return items


def table_abs(tbl, base_url):
    tb = BeautifulSoup(str(tbl), "html.parser")
    for a in tb.find_all("a", href=True):
        h = a["href"]
        if h and not h.startswith(("http", "javascript", "#")):
            a["href"] = urljoin(base_url, h)
    for im in tb.find_all("img", src=True):
        s = im["src"]
        if s and not s.startswith(("http", "data:")):
            im["src"] = urljoin(base_url, s)
    return str(tb)


def clean_span(text):
    text = re.sub(r'(?<=[\u4e00-\u9fff])\s+(?=[\u4e00-\u9fff])', '', text)
    text = re.sub(r'(?<=[\u4e00-\u9fff])\s+(?=\d)', '', text)
    text = re.sub(r'(?<=\d)\s+(?=[\u4e00-\u9fff])', '', text)
    text = re.sub(r'\s+([，。、；：？！)】」》])', r'\1', text)
    text = re.sub(r'([（【「《])\s+', r'\1', text)
    return re.sub(r'\s+', ' ', text).strip()


def extract_content(div, page_url):
    # 去掉页眉工具条(标题/时间/来源/分享)
    parts = []
    body = None
    for c in div.find_all('div', recursive=True):
        if c.find_parent('div') is not div:
            continue
        cls = ' '.join(c.get('class', []))
        if c is div:
            continue
    for child in div.find_all(['p', 'div', 'table', 'ul', 'ol'], recursive=True):
        if child.find_parent('table'):
            continue
        if child.name == 'div':
            cl = ' '.join(child.get('class', []))
            if re.search(r'(share|ewm|qrcode|print|fx|toolbar|title|date)', cl, re.I) and len(child.get_text('', strip=True)) < 200:
                continue
            if child.find(['p', 'table', 'ul', 'ol'], recursive=False):
                continue
        if child.name == 'table':
            parts.append(table_abs(child, page_url))
            continue
        txt = clean_span(child.get_text('', strip=True))
        txt = re.sub(r'&#\d+;', '', txt)
        if child.name in ('p', 'div', 'ul', 'ol') and txt and len(txt) > 2:
            parts.append(f'<p>{txt}</p>')
    uniq = []
    for p in parts:
        if p not in uniq:
            uniq.append(p)
    return '\n'.join(uniq)


def parse_detail(url):
    html = fetch(url)
    if not html:
        return None, None, None
    soup = BeautifulSoup(html, "html.parser")
    title = ""
    mt = soup.find("meta", attrs={"name": re.compile("ArticleTitle", re.I)})
    if mt and mt.get("content"):
        title = html_lib.unescape(mt["content"]).strip()
    if not title:
        h1 = soup.find("h1")
        if h1:
            title = h1.get_text(strip=True)
    date = ""
    pd = soup.find("meta", attrs={"name": re.compile("PubDate", re.I)})
    if pd and pd.get("content"):
        m = re.search(r'(\d{4}-\d{2}-\d{2})', pd["content"])
        if m:
            date = m.group(1)
    content = ""
    d = soup.select_one("div.content.pd25") or soup.select_one("div#zoom") or soup.select_one("div.xxgk_contnet")
    if d:
        content = extract_content(d, url)
    return title, content, date


def main():
    print(f"=== {SITE_NAME} ===")
    all_items = []
    html = fetch(BASE + LIST_PATH + "/ghjh.html")
    if html:
        items = parse_list(html)
        all_items.extend(items)
        print(f"  page1: {len(items)}")
    seen, filtered = set(), []
    for title, url, date in all_items:
        if url in seen:
            continue
        seen.add(url)
        if date and date < CUTOFF:
            continue
        filtered.append((title, url, date))
    print(f"  去重/3yr后: {len(filtered)}")
    results = []
    for idx, (title, url, date) in enumerate(filtered):
        dt, content, dd = parse_detail(url)
        ft = dt or title
        fd = date or dd
        if not content or len(content.strip()) < 10:
            print(f"  [{idx+1}] ⚠空 {title[:30]}")
            continue
        summary = re.sub(r'<[^>]+>', '', content)[:200].strip() or ft[:200]
        results.append({"site_name": SITE_NAME, "title": ft, "url": url,
                        "content": content, "summary": summary, "pub_date": fd})
        print(f"  [{idx+1}] ✅ {ft[:35]} ({len(content)}B)")
        time.sleep(0.3)
    if results:
        push_to_searchdb(results, "guanyun_ghjh")
        print(f"入库 {len(results)}")
    print("完成")


if __name__ == "__main__":
    main()