#!/usr/bin/env python3
"""crawl_huizhou_hpslgg.py - 惠州市-环评受理公告

站点: https://shj.huizhou.gov.cn/zwgk/jsxmhpglxxgk/hpslgg/
CMS: NFCMS (广东南方网系网站群, window.NFCMS_SITE_ID=752011)
列表: UL.ul_art_row > LI.li_art_title > A[title] 标题 + LI.li_art_date 日期(YYYY-MM-DD)
分页: 首页 index.html + index_N.html (N=2..8, 无index_1; 每页20条)
⚠️ WAF: 分页页须 Chrome UA + Referer, 否则 nginx 直接 404
详情: TD.artTitle.xltd_9 标题 + meta PubDate 日期 + TD.artContent.xltd_10 > DIV#divZoom 正文(平衡div, UEditor表格)
附件: TD.xltd_11 > div.m-attachment > div.f-APPENDIX > a.pdf (绝对URL) → 转 <p><a>附件名</a></p> 段落嵌入
"""
import sys, os, re, json, argparse
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
from datetime import datetime, timedelta

CUTOFF = (datetime.now() - timedelta(days=365 * 3)).strftime('%Y-%m-%d')
SITE_NAME = '惠州市-环评受理公告'
GROUP_NAME = '广东'
BASE = 'https://shj.huizhou.gov.cn'
LIST_DIR = '/zwgk/jsxmhpglxxgk/hpslgg/'
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',
    'Referer': BASE + LIST_DIR + 'index.html',
}


def fetch(url):
    try:
        r = requests.get(url, timeout=(5, 25), headers=HEADERS, verify=False)
        r.encoding = 'utf-8'
        return r.text
    except Exception as e:
        print(f"[WARN] fetch failed: {url} - {e}", file=sys.stderr)
        return ''


def extract_balanced_div(html, open_tag_re):
    """平衡 div 匹配 (大小写不敏感)"""
    m = re.search(open_tag_re, html, re.I)
    if not m:
        return ''
    gt = html.find('>', m.start())
    if gt == -1:
        return ''
    i = gt + 1
    depth = 1
    for mm in re.finditer(r'<div[\s>]|</div>', html[i:], re.I):
        if mm.group(0).startswith('<div') or mm.group(0).startswith('<DIV'):
            depth += 1
        else:
            depth -= 1
            if depth == 0:
                return html[i:i + mm.start()]
    return html[i:]


def fetch_list(page_idx):
    """列表页: 0=首页 index.html, N=index_N.html (N>=2, 无index_1)"""
    if page_idx == 0:
        url = BASE + LIST_DIR + 'index.html'
    else:
        url = BASE + LIST_DIR + f'index_{page_idx + 1}.html'
    html = fetch(url)
    if not html:
        return []
    items = []
    for m in re.finditer(
            r'<LI class="li_art_title">\s*<A[^>]*title="([^"]*)"[^>]*href="([^"]+)"[^>]*>(.*?)</A>.*?'
            r'<LI class="li_art_date">([^<]*)</LI>', html, re.DOTALL):
        title_attr, href, title_html, date_text = m.group(1), m.group(2), m.group(3), m.group(4).strip()
        title = (title_attr or re.sub(r'<[^>]+>', '', title_html)).strip()
        title = re.sub(r'\s+', ' ', title).strip()
        # strip &middot;&nbsp; 实体前缀
        title = re.sub(r'^\s*(?:&middot;|&nbsp;|·|\s)+', '', title)
        if len(title) < 4:
            continue
        # 只抓精准子栏目 (post_ 详情链接须在本栏目路径下)
        if '/content/post_' not in href or '/hpslgg/' not in href:
            continue
        url = urljoin(BASE + LIST_DIR, href)
        md = re.search(r'(\d{4}-\d{2}-\d{2})', date_text)
        pd = md.group(1) if md else ''
        items.append({'url': url, 'pub_date': pd, 'title': title})
    return items


def fetch_detail(url):
    """详情页: artTitle 标题 + PubDate 日期 + divZoom 正文 + f-APPENDIX 附件转段落"""
    html = fetch(url)
    if not html:
        return '', '', ''
    # 标题: TD.artTitle (移除 artSource 发布时间区)
    title = ''
    m = re.search(r'<TD[^>]*class="[^"]*artTitle[^"]*"[^>]*>(.*?)</TD>', html, re.DOTALL | re.I)
    if m:
        seg = re.sub(r'<DIV[^>]*artSource[^>]*>.*?</DIV>', '', m.group(1), flags=re.DOTALL | re.I)
        title = re.sub(r'<[^>]+>', '', seg).strip()
        title = re.sub(r'\s+', ' ', title).strip()
    if not title:
        m = re.search(r'<meta\s+name="ArticleTitle"\s+content="([^"]+)"', html)
        if m:
            title = m.group(1).strip()
    title = re.sub(r'^\s*(?:&middot;|&nbsp;|·|\s)+', '', title)
    title = re.sub(r'_(惠州市生态环境局|惠州市|惠州市生态环境局网站)$', '', title)
    # 日期: meta PubDate
    date_text = ''
    m = re.search(r'<meta\s+name="PubDate"\s+content="([^"]+)"', html)
    if m:
        md = re.search(r'(\d{4}-\d{2}-\d{2})', m.group(1))
        if md:
            date_text = md.group(1)
    if not date_text:
        m = re.search(r'发布时间：(\d{4}-\d{2}-\d{2})', html)
        if m:
            date_text = m.group(1)
    # 正文: DIV#divZoom (平衡 div)
    content = ''
    c = extract_balanced_div(html, r'<div[^>]*id="divZoom"[^>]*>')
    if c.strip():
        content = c.strip()
    if not content.strip():
        c = extract_balanced_div(html, r'<TD[^>]*class="[^"]*artContent[^"]*"[^>]*>')
        if c.strip():
            content = c.strip()
    content = re.sub(r'<script.*?</script>', '', content, flags=re.DOTALL | re.I)
    content = re.sub(r'<style.*?</style>', '', content, flags=re.DOTALL | re.I)
    # 剥掉正文开头残留的 divZoom 容器标签
    content = re.sub(r'^<div[^>]*id="divZoom"[^>]*>\s*', '', content, flags=re.I)
    # 图片/附件相对路径 → 详情页目录 urljoin
    content = re.sub(r'href="(?!https?://|//|#|javascript:)([^"]+)"',
                     lambda mm: f'href="{urljoin(url, mm.group(1))}"', content)
    content = re.sub(r'src="(?!https?://|//)([^"]+)"',
                     lambda mm: f'src="{urljoin(url, mm.group(1))}"', content)
    # 正文内附件链接保留 (含附件/表格的段落勿拍平)
    # 附件区 f-APPENDIX → 转 <p><a>附件名</a></p> 段落嵌入正文末尾
    attach_ps = []
    m = re.search(r'<div[^>]*class="[^"]*f-APPENDIX[^"]*"[^>]*>(.*?)</div>', html, re.DOTALL)
    if m:
        for am in re.finditer(r'<a[^>]*href="([^"]+)"[^>]*>(.*?)</a>', m.group(1), re.DOTALL):
            h, t = am.group(1).strip(), re.sub(r'<[^>]+>', '', am.group(2)).strip()
            if h:
                if h.startswith('/'):
                    h = BASE + h
                attach_ps.append(f'<p><a href="{h}">{t or "附件"}</a></p>')
    if attach_ps:
        content = content.rstrip() + '\n' + '\n'.join(attach_ps)
    # 清首尾空 p
    content = re.sub(r'^(<p[^>]*>(\s*(<br\s*/?>)?\s*)?</p>)+', '', content)
    content = re.sub(r'(<p[^>]*>(\s*(<br\s*/?>)?\s*)?</p>)+$', '', content)
    return title, date_text, content.strip()


def run(max_pages=5):
    records = []
    seen = set()
    for pg in range(max_pages):
        items = fetch_list(pg)
        if not items:
            print(f"  Page {pg + 1}: 0 items (stop)")
            break
        print(f"  Page {pg + 1}: {len(items)} items")
        for it in items:
            if it['pub_date'] and it['pub_date'] < CUTOFF:
                continue
            if it['url'] in seen:
                continue
            seen.add(it['url'])
            title, date_text, content = fetch_detail(it['url'])
            if not content:
                print(f"    SKIP empty content: {it['url']}")
                continue
            if not title:
                title = it['title']
            pd = date_text or it['pub_date']
            # 附件提取 (从正文段落)
            atts = []
            for am in re.finditer(r'<a[^>]*href="([^"]+)"[^>]*>(.*?)</a>', content, re.DOTALL):
                h, t = am.group(1), re.sub(r'<[^>]+>', '', am.group(2)).strip()
                if re.search(r'\.(pdf|doc|docx|xls|xlsx|zip|rar|caj|wps)$', h, re.I) or 'attachment' in h.lower():
                    atts.append({'title': t or '附件', 'url': h})
            records.append({
                'title': title, 'url': it['url'], 'source_url': it['url'],
                'pub_date': pd, 'site_name': SITE_NAME, 'content': content,
                'summary': '', 'attachments': json.dumps(atts, ensure_ascii=False) if atts else '[]',
                'group_name': GROUP_NAME,
            })
    print(f"  Total: {len(records)} items")
    valid = [r for r in records if r['content'].strip()]
    if valid:
        push_to_searchdb(valid, "huizhou_hpslgg")
    return len(valid)


if __name__ == '__main__':
    ap = argparse.ArgumentParser()
    ap.add_argument('--pages', type=int, default=5)
    args = ap.parse_args()
    n = run(args.pages)
    print(f"[RESULT] {SITE_NAME}: {n} items")