#!/usr/bin/env python3
"""
Crawler for 云梦县生态环境局 - 业务工作
CMS: Custom CMS, UTF-8, ul.news-list列表 + article.htmledit_views正文
Site: www.yunmeng.gov.cn/c/ymxsthjj/ywgz.jhtml
"""
import requests, re, json, sys, os
from datetime import datetime, timedelta
from bs4 import BeautifulSoup
from urllib.parse import urljoin

BASE_URL = "http://www.yunmeng.gov.cn"
LIST_PATH = "/c/ymxsthjj/ywgz"
LIST_URL = BASE_URL + LIST_PATH + ".jhtml"
SITE_NAME = "云梦县生态环境局-业务工作"
INCREMENTAL_DAYS = 7
MAX_PAGES = 5

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
    "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8",
    "Accept-Language": "zh-CN,zh;q=0.9,en;q=0.8",
}
TIMEOUT = 20
session = requests.Session()
session.headers.update(HEADERS)


def fetch(url):
    resp = session.get(url, timeout=TIMEOUT)
    resp.encoding = "utf-8"
    return resp.text


def parse_list(html):
    """Parse list page — ul.news-list > li > a.news-title[title] + span.news-time"""
    soup = BeautifulSoup(html, "html.parser")
    items = []
    ul = soup.find("ul", class_="news-list")
    if not ul:
        return items
    for li in ul.find_all("li", recursive=False):
        a = li.find("a", class_="news-title", href=True)
        if not a:
            continue
        title = a.get("title") or a.get_text(strip=True)
        href = a["href"].strip()
        if not title or not href:
            continue
        date_span = li.find("span", class_="news-time")
        date = date_span.get_text(strip=True) if date_span else ""
        # Resolve URL
        if href.startswith("http"):
            pass
        elif href.startswith("/"):
            href = BASE_URL + href
        else:
            href = BASE_URL + LIST_PATH + "/" + href.lstrip("./")
        items.append((title.strip(), href, date[:10]))
    return items


def parse_detail(html, url):
    """Parse detail page — div.news-title > h2 + div.news-box#js_contentBox > article.htmledit_views."""
    soup = BeautifulSoup(html, "html.parser")

    # Title from h2 in news-title
    title = ""
    news_title = soup.find("div", class_="news-title")
    if news_title:
        h2 = news_title.find("h2")
        if h2:
            title = h2.get_text(strip=True)

    # Date from news-inf
    pub_date = ""
    news_inf = soup.find("div", class_="news-inf")
    if news_inf:
        text = news_inf.get_text()
        m = re.search(r"(\d{4}-\d{2}-\d{2})", text)
        if m:
            pub_date = m.group(1)

    # Content from article.htmledit_views inside news-box#js_contentBox
    content_html = ""
    news_box = soup.find("div", class_="news-box", id="js_contentBox")
    if news_box:
        article = news_box.find("article", class_="htmledit_views")
        if article:
            parts = []
            processed = set()
            for child in article.find_all(["p", "div", "table", "img"], recursive=True):
                if id(child) in processed:
                    continue
                processed.add(id(child))
                tag = child.name.lower()

                if tag == "div" and not child.find(["p", "table", "img", "a"]):
                    continue

                if tag == "p":
                    p_parts = []
                    for elem in child.contents:
                        if isinstance(elem, str):
                            txt = elem.strip()
                            if txt:
                                p_parts.append(txt)
                        elif elem.name == "a":
                            processed.add(id(elem))
                            ahref = elem.get("href", "").strip()
                            atxt = elem.get_text(strip=True)
                            if ahref and atxt:
                                full_href = urljoin(url, ahref)
                                p_parts.append(f"[{atxt}]({full_href})")
                            elif atxt:
                                p_parts.append(atxt)
                        elif elem.name == "img":
                            processed.add(id(elem))
                            src = elem.get("src", "")
                            if src:
                                alt = elem.get("alt", "")
                                full_src = urljoin(url, src)
                                p_parts.append(f"![{alt}]({full_src})" if alt else f"![]({full_src})")
                        elif hasattr(elem, "get_text"):
                            txt = elem.get_text(" ", strip=True)
                            if txt:
                                p_parts.append(txt)
                    txt = " ".join(p_parts)
                    if txt:
                        parts.append(txt)
                elif tag == "div":
                    div_parts = []
                    for a in child.find_all("a", href=True):
                        processed.add(id(a))
                        ahref = a.get("href", "").strip()
                        atxt = a.get_text(strip=True)
                        if ahref and atxt:
                            full_href = urljoin(url, ahref)
                            div_parts.append(f"[{atxt}]({full_href})")
                    txt = child.get_text(" ", strip=True)
                    if txt and div_parts:
                        text_only = txt
                        for p in div_parts:
                            link_text = re.search(r'\[(.+?)\]\(', p)
                            if link_text:
                                text_only = text_only.replace(link_text.group(1), "", 1).strip()
                        if text_only:
                            parts.append(text_only)
                        parts.extend(div_parts)
                    elif txt:
                        parts.append(txt)
                elif tag == "table":
                    parts.append(str(child))
                elif tag == "img":
                    src = child.get("src", "")
                    if src:
                        alt = child.get("alt", "")
                        full_src = urljoin(url, src)
                        parts.append(f'<img src="{full_src}" alt="{alt}">')
            content_html = "\n\n".join(parts)

    # Attachments
    scope = news_box or soup
    attachments = []
    for a in scope.find_all("a", href=True):
        ahref = a["href"].strip().lower()
        if re.search(r"\.(pdf|doc|docx|xls|xlsx|zip|rar|txt)$", ahref):
            full_url = urljoin(url, a["href"].strip())
            attachments.append({"title": a.get_text(strip=True) or full_url.split("/")[-1], "url": full_url})

    if len(content_html.strip()) < 20 and attachments:
        content_html = f'<p><a href="{url}">{title}</a></p>\n\n附件列表：\n'
        for att in attachments:
            content_html += f"  - [{att['title']}]({att['url']})\n"

    return title, pub_date, content_html, attachments


def incremental_filter(items):
    cutoff = datetime.now() - timedelta(days=INCREMENTAL_DAYS)
    filtered = []
    for title, url, date_str in items:
        try:
            item_date = datetime.strptime(date_str, "%Y-%m-%d")
            if item_date >= cutoff:
                filtered.append((title, url, date_str))
        except (ValueError, IndexError):
            filtered.append((title, url, date_str))
    return filtered


def main():
    is_incremental = any(arg in sys.argv for arg in ["--incremental", "incremental", "inc"])
    page_urls = [LIST_URL]
    for i in range(2, MAX_PAGES + 1):
        page_urls.append(BASE_URL + LIST_PATH + f"_{i}.jhtml")

    all_items = []
    for idx, url in enumerate(page_urls):
        try:
            html = fetch(url)
            items = parse_list(html)
            print(f"Page {idx+1}: {len(items)} items", file=sys.stderr)
            all_items.extend(items)
        except Exception as e:
            print(f"Page {idx+1} error ({url}): {e}", file=sys.stderr)

    print(f"Total: {len(all_items)}", file=sys.stderr)
    if is_incremental:
        all_items = incremental_filter(all_items)
        print(f"Incremental: {len(all_items)}", file=sys.stderr)

    seen = set()
    unique_items = []
    for item in all_items:
        if item[1] not in seen:
            seen.add(item[1])
            unique_items.append(item)

    print(f"Unique: {len(unique_items)}", file=sys.stderr)

    results = []
    for title, url, list_date in unique_items:
        try:
            html = fetch(url)
            det_title, det_date, content, attachments = parse_detail(html, url)
            final_title = det_title or title
            final_date = det_date or list_date
            results.append({
                "title": final_title,
                "page_url": url,
                "publish_date": final_date,
                "content": content,
                "attachments": json.dumps(attachments, ensure_ascii=False) if attachments else "",
                "site_name": SITE_NAME,
            })
            print(f"  OK: {final_title[:50]}", file=sys.stderr)
        except Exception as e:
            print(f"  ERR {url}: {e}", file=sys.stderr)

    for r in results:
        print(json.dumps(r, ensure_ascii=False))


if __name__ == "__main__":
    main()