#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
镇江经济技术开发区 - 通告/公示 (tztg) 爬虫
列表: https://www.zjna.gov.cn/zjna/tztg/xxgk_list.shtml
  注意: 必须 https://www.zjna.gov.cn (带www+https), 裸域名 502
  静态分页: xxgk_list_N.shtml (createPageHTML, 34页 666条)
详情: /zjna/xxx/YYYYMM/UUID.shtml
  标题: <meta name="ArticleTitle"> (完整)
  正文: div#zoom 容器
用法: python3 crawl_zjna_gggs.py [--pages=N] [--limit=N]
"""
import os, sys, re, time, json, html as html_mod
import requests
from urllib.parse import urljoin

sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
from crawler_lib import push_to_searchdb

BASE_URL = "https://www.zjna.gov.cn"
LIST_URL = BASE_URL + "/zjna/tztg/xxgk_list.shtml"
SITE_NAME = "镇江经济技术开发区-通告公示"
CATEGORY = "通告公示"
MAX_PAGES = 34   # totalRecord=666, 约20条/页

UA = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/125.0.0.0 Safari/537.36"
HEADERS = {"User-Agent": UA, "Referer": LIST_URL}


def parse_args():
    pages, limit = 0, 0
    for a in sys.argv[1:]:
        if a.startswith("--pages="):
            pages = int(a.split("=")[1])
        elif a.startswith("--limit="):
            limit = int(a.split("=")[1])
    return pages, limit


def clean_title(t):
    t = html_mod.unescape(t or "")
    t = re.sub(r"[\u200b\u200e\u200f\ufeff\xa0]", "", t)
    t = re.sub(r"\s+", " ", t)
    t = re.sub(r"^[•·\-—]\s*", "", t)
    return t.strip()


def fetch_list(session, page):
    """静态分页: xxgk_list.html (p1) / xxgk_list_N.html"""
    if page == 1:
        url = LIST_URL
    else:
        url = BASE_URL + f"/zjna/tztg/xxgk_list_{page}.shtml"
    try:
        r = session.get(url, timeout=30)
        r.encoding = "utf-8"
        html = r.text
        items = []
        for m in re.finditer(r'<a href="([^"]+)"[^>]*title="([^"]*)"[^>]*>(?:<[^>]*>)*([^<]{2,})</a>\s*<span class="time">\s*([\d-]+)', html, re.S):
            href, title, linktext, date = m.group(1), m.group(2).strip(), m.group(3).strip(), m.group(4).strip()
            href = urljoin(BASE_URL, href) if not href.startswith("http") else href
            # 只保留站内 shtml 详情 (排除外部链接)
            if "zjna.gov.cn/zjna/" in href and href.endswith(".shtml"):
                t = clean_title(title if title else linktext)
                if len(t) >= 4:
                    items.append({"title": t, "href": href, "date": date[:10]})
        return items
    except Exception as e:
        print(f"    [ERR] list page {page}: {e}", flush=True)
        return []


def fetch_detail(session, url):
    """详情: meta ArticleTitle + div#zoom"""
    try:
        r = session.get(url, headers={"User-Agent": UA, "Referer": LIST_URL}, timeout=30)
        r.encoding = "utf-8"
        html = r.text
        title = ""
        m = re.search(r'<meta name="ArticleTitle" content="([^"]*)"', html)
        if m:
            title = clean_title(m.group(1))
        if not title:
            m = re.search(r"<h1[^>]*>([^<]+)</h1>", html)
            if m:
                title = clean_title(m.group(1))
        pub_date = ""
        m = re.search(r'<meta name="PubDate" content="([^"]*)"', html)
        if m:
            pub_date = m.group(1).strip()[:10]
        if not pub_date:
            m = re.search(r"(\d{4}-\d{1,2}-\d{1,2})", html)
            if m:
                pub_date = m.group(1)
        # 正文: div#zoomcon (或 div#zoom)
        body = ""
        i = html.find('id="zoomcon"')
        if i < 0:
            i = html.find('id="zoom"')
        if i >= 0:
            start = html.find(">", i) + 1
            depth = 0
            j = start
            while j < len(html):
                if html[j:j+4] == "<div":
                    depth += 1
                    j += 4
                elif html[j:j+6] == "</div>":
                    depth -= 1
                    j += 6
                    if depth == 0:
                        break
                else:
                    j += 1
            raw = html[start:j-6]
            raw = re.sub(r"<script[\s\S]*?</script>", "", raw)
            raw = re.sub(r"<style[\s\S]*?</style>", "", raw)
            raw = re.sub(r'href="([^"]*)"', lambda m: 'href="%s"' % (urljoin(url, m.group(1)) if not m.group(1).startswith(("http", "#", "javascript")) else m.group(1)), raw)
            raw = re.sub(r'src="([^"]*)"', lambda m: 'src="%s"' % (urljoin(url, m.group(1)) if not m.group(1).startswith(("http", "data:", "javascript")) else m.group(1)), raw)
            raw = re.sub(r'\sstyle="[^"]*"', "", raw)
            raw = re.sub(r"<span[^>]*>|</span>|<strong[^>]*>|</strong>|<b[^>]*>|</b>|<font[^>]*>|</font>", "", raw)
            body = raw.strip()
        if not body or len(body) < 20:
            body = ""
        return title, pub_date, body
    except Exception as e:
        print(f"    [ERR] detail {url}: {e}", flush=True)
        return "", "", ""


def main():
    max_pages, limit = parse_args()
    if not max_pages:
        max_pages = MAX_PAGES
    print(f"[ZJNA] SITE={SITE_NAME} max_pages={max_pages}", flush=True)

    session = requests.Session()
    session.headers.update({"User-Agent": UA})

    all_items = []
    seen = set()
    for pg in range(1, max_pages + 1):
        items = fetch_list(session, pg)
        new_count = 0
        for it in items:
            if it["href"] in seen:
                continue
            seen.add(it["href"])
            all_items.append(it)
            new_count += 1
        print(f"  Page {pg}: {len(items)}条(新{new_count}) 累计{len(all_items)}", flush=True)
        if not items:
            print(f"  [END] page {pg} empty", flush=True)
            break
        time.sleep(0.4)

    print(f"列表完成: {len(all_items)} 条, 抓详情...", flush=True)
    if limit and len(all_items) > limit:
        all_items = all_items[:limit]

    results = []
    for i, it in enumerate(all_items):
        d_title, d_date, body = fetch_detail(session, it["href"])
        title = d_title or it["title"]
        pub_date = d_date or it.get("date", "")
        summary = re.sub(r"<[^>]+>", " ", body) if body else title
        summary = re.sub(r"\s+", " ", summary).strip()[:300]
        atts = []
        if body:
            for m in re.finditer(r'<a[^>]*href="([^"]*\.(?:pdf|doc|docx|xls|xlsx|zip|rar|ofd|wps))"[^>]*>([^<]*)</a>', body, re.I):
                atts.append({"name": m.group(2).strip(), "url": m.group(1)})
        results.append({
            "site_name": SITE_NAME, "source_url": it["href"], "url": it["href"],
            "title": title, "pub_date": pub_date, "content": body,
            "summary": summary, "category": CATEGORY,
            "attachments": json.dumps(atts, ensure_ascii=False) if atts else "",
        })
        if (i + 1) % 10 == 0:
            print(f"  detail {i+1}/{len(all_items)}", flush=True)
        time.sleep(0.3)

    print(f"  pushing {len(results)} items to searchdb...", flush=True)
    push_to_searchdb(results, batch_label=SITE_NAME)
    empty = sum(1 for r in results if not (r.get("content") or "").strip())
    print(f"  DONE. pushed={len(results)} 空正文={empty}", flush=True)


if __name__ == "__main__":
    main()