#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
爬虫：广西百色田东县 - 建设项目环境影响评价审批
站点：www.gxtd.gov.cn
CMS：广西政府网站统一平台（static pagination: createPageHTML）
"""

import sys, os, re, requests, time
from bs4 import BeautifulSoup

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

SITE_NAME = "田东县-建设项目环评审批"

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

BASE_URL = 'http://www.gxtd.gov.cn'
LIST_BASE = 'http://www.gxtd.gov.cn/xxgk/fdzdgknr/zdlyxxgk/shgysy/hjbhly/jsxm'

session = requests.Session()
session.headers.update(HEADERS)


# ─── 正文取文本（2026-09-11）：行内节点直接拼接，只在块级边界 / <br> 处换行 ───
# ⚠️ 不要用 el.get_text("\n") 取正文 —— 它是「每个**文本节点**之间插 \n」，Word 粘贴的
#    公文把一行拆成 <span>提取码：</span>pwaj<span>。查阅…</span>，这些行内节点于是各自
#    成行（福泉 id=2095080103703914437 实例：`提取码：`/`pwaj`/`。查阅…` 各占一行）。
_BLOCK_TAGS = {'address', 'article', 'aside', 'blockquote', 'details', 'dialog', 'dd', 'div',
               'dl', 'dt', 'fieldset', 'figcaption', 'figure', 'footer', 'form', 'h1', 'h2',
               'h3', 'h4', 'h5', 'h6', 'header', 'hgroup', 'hr', 'li', 'main', 'nav', 'ol',
               'p', 'pre', 'section', 'table', 'tbody', 'thead', 'tfoot', 'tr', 'td', 'th',
               'ul', 'center', 'caption'}


def body_text(el):
    """块级边界出换行、行内节点直接拼接、<br> 出换行（≈ 浏览器看到的换行结构）。"""
    if el is None:
        return ''
    import re as _re
    from bs4 import NavigableString
    out = []

    def walk(node):
        for ch in node.children:
            if isinstance(ch, NavigableString):
                out.append(str(ch))
            elif getattr(ch, 'name', None) == 'br':
                out.append('\n')
            elif getattr(ch, 'name', None) in _BLOCK_TAGS:
                out.append('\n')
                walk(ch)
                out.append('\n')
            else:
                walk(ch)
    walk(el)
    t = ''.join(out)
    t = _re.sub(r'[ \t\r\f\v]*\n[ \t\r\f\v]*', '\n', t)
    t = _re.sub(r'\n{3,}', '\n\n', t)
    return t.strip()


def safe_get(url, max_retries=3):
    for attempt in range(max_retries):
        try:
            r = session.get(url, timeout=30)
            r.encoding = 'utf-8'
            return r
        except (requests.ConnectionError, requests.Timeout) as e:
            if attempt < max_retries - 1:
                time.sleep(2)
                continue
            print(f"    ⚠️ 请求失败({url}): {e}")
            return None
    return None


def fetch_list_page(page):
    """获取列表页"""
    if page == 1:
        url = f"{LIST_BASE}/index.shtml"
    else:
        url = f"{LIST_BASE}/index_{page}.shtml"

    r = safe_get(url)
    if r is None:
        return [], 0

    soup = BeautifulSoup(r.text, 'html.parser')

    # 解析总页数
    total_pages = 0
    script_tag = soup.find('script', text=re.compile(r'createPageHTML'))
    if script_tag:
        m = re.search(r'createPageHTML\((\d+)', script_tag.string or '')
        if m:
            total_pages = int(m.group(1))

    items = []
    morelist = soup.find('div', id='morelist')
    if not morelist:
        morelist = soup.find('div', class_='more')
    if not morelist:
        return [], total_pages

    for ul in morelist.find_all('ul', class_='more-list'):
        for li in ul.find_all('li', recursive=False):
            a_tag = li.find('a', href=True)
            if not a_tag:
                continue
            href = a_tag.get('href', '')
            title = a_tag.get('title', '') or a_tag.get_text(strip=True)
            if not title or not href:
                continue

            # 相对路径转绝对
            if href.startswith('./'):
                href = LIST_BASE + '/' + href[2:]
            elif not href.startswith('http'):
                if href.startswith('/'):
                    href = BASE_URL + href
                else:
                    href = LIST_BASE + '/' + href

            date_span = li.find('span')
            date_str = date_span.get_text(strip=True) if date_span else ''

            items.append({'title': title, 'url': href, 'date': date_str})

    return items, total_pages


def fetch_detail(url):
    """获取详情页正文"""
    r = safe_get(url)
    if r is None:
        return ''

    soup = BeautifulSoup(r.text, 'html.parser')

    # 正文容器
    content_div = soup.find('div', class_='article-con')
    if not content_div:
        content_div = soup.find('div', class_=re.compile(r'content|nrzw|detail|zoom', re.I))
    if not content_div:
        return ''

    # 找trs_editor_view或直接取p/table
    editor = content_div.find('div', class_=re.compile(r'trs_editor|TRS_UEDITOR|trs_paper', re.I))
    if editor:
        target = editor
    else:
        target = content_div

    parts = []
    for child in target.children:
        if child.name == 'p':
            text = child.get_text(' ', strip=True)
            if text:
                parts.append(text)
        elif child.name == 'table':
            md = html_table_to_html(str(child))
            if md:
                parts.append(md)
        elif child.name in ('div', 'section', 'article'):
            _extract_nodes(child, parts)

    if not parts:
        text = body_text(target)
        if text:
            parts = [text]

    body = '\n\n'.join(parts)
    # 清除中文字符间多余空格
    body = re.sub(r'(?<=[\u4e00-\u9fff])[ \t\u3000]+(?=[\u4e00-\u9fff])', '', body)

    # 获取附件
    attachment_links = []
    download_div = content_div.find('div', id='downloadfile')
    if download_div:
        file_div = download_div.find('div', id='article-file')
        if file_div:
            for a in file_div.find_all('a', href=True):
                href = a.get('href', '')
                text = a.get_text(strip=True) or os.path.basename(href)
                if href.startswith('./'):
                    href = url.rsplit('/', 1)[0] + '/' + href[2:]
                elif not href.startswith('http'):
                    if href.startswith('/'):
                        href = BASE_URL + href
                    else:
                        href = url.rsplit('/', 1)[0] + '/' + href
                attachment_links.append(f"[{text}]({href})")

    if attachment_links:
        body += '\n\n**附件：**\n' + '\n'.join(attachment_links)

    return body


def _extract_nodes(node, parts):
    for child in node.children:
        if child.name == 'p':
            text = child.get_text(' ', strip=True)
            if text:
                parts.append(text)
        elif child.name == 'table':
            md = html_table_to_html(str(child))
            if md:
                parts.append(md)
        elif child.name in ('div', 'section', 'article'):
            _extract_nodes(child, parts)


def html_table_to_html(table, base_url=""):
    """保留 HTML 表格结构，仅将相对链接/图片转绝对 URL"""
    from bs4 import BeautifulSoup
    tbl = BeautifulSoup(str(table), 'html.parser')
    for a in tbl.find_all('a'):
        href = a.get('href', '')
        if href and not href.startswith(('http', 'javascript', '#')):
            a['href'] = urllib.parse.urljoin(base_url, href) if base_url else href
    for img in tbl.find_all('img'):
        src = img.get('src', '')
        if src and not src.startswith(('http', '//', 'data:')):
            img['src'] = urllib.parse.urljoin(base_url, src) if base_url else src
    return str(tbl)


def main():
    import argparse
    parser = argparse.ArgumentParser()
    parser.add_argument('--pages', type=int, default=0, help='仅爬前N页（0=全部）')
    args = parser.parse_args()
    max_pages = args.pages if args.pages else 0

    print(f"[{SITE_NAME}] 开始")

    all_items = []
    total_pages_found = 0

    for page in range(1, 100):  # 最多100页
        items, pagecount = fetch_list_page(page)
        if not items:
            if page == 1:
                print("  ⚠️ 第1页为空，退出")
                return
            print(f"  第{page}页: 空，结束")
            break
        if page == 1:
            total_pages_found = pagecount
            if total_pages_found:
                print(f"  总计 {total_pages_found} 页")
        print(f"  第{page}页: {len(items)} 条")
        all_items.extend(items)
        if total_pages_found and page >= total_pages_found:
            break
        if max_pages and page >= max_pages:
            print(f"  --pages={max_pages} 限制，停止")
            break

    print(f"\n列表合计: {len(all_items)} 条")

    if not all_items:
        print("  列表为空，退出")
        return

    db_items = []
    success = 0
    failed = 0
    for i, item in enumerate(all_items):
        url = item['url']
        title = item['title']
        date_str = item['date']
        print(f"  [{i+1}/{len(all_items)}] {title[:40]}...", end=' ')
        body = fetch_detail(url)
        if not body:
            print("⚠️ 空正文")
            failed += 1
        else:
            print(f"✅ {len(body)}字")
            success += 1

        db_items.append({
            'site_name': SITE_NAME,
            'source_url': url,
            'url': url,
            'title': title,
            'pub_date': date_str,
            'summary': body[:500] if body else '',
            'content': body,
        })

    push_to_searchdb(db_items, batch_label=SITE_NAME)
    print(f"\n[{SITE_NAME}] 完成: 共{len(all_items)}条, 正文成功{success}, 空{failed}")


if __name__ == '__main__':
    main()