#!/usr/bin/env python3
"""
Crawler for 云梦县生态环境局 - 环境评价
CMS: Hanweb/JPAAS, static .jhtml pages
Site: www.yunmeng.gov.cn/c/ymxsthjj/hjpj1.jhtml
Note: Requires --resolve + HTTPS to bypass JS challenge
"""
import os, re, json, sys, subprocess, tempfile
from datetime import datetime, timedelta
from bs4 import BeautifulSoup

BASE_URL = "https://www.yunmeng.gov.cn"
CHANNEL = "ymxsthjj"
LIST_PATH = "/c/{}/hjpj1".format(CHANNEL)
RESOLVE_IP = "58.51.146.198"
SITE_NAME = "云梦县生态环境局-环境评价"
INCREMENTAL_DAYS = 7
MAX_PAGES = 5
PER_PAGE = 15

BASE_CURL = [
    "curl", "-s", "-k", "--connect-timeout", "10", "--max-time", "20",
    "--resolve", "www.yunmeng.gov.cn:443:" + RESOLVE_IP,
    "-A", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36",
    "--compressed",
]


def fetch(url):
    """Fetch URL via curl with --resolve."""
    cmd = BASE_CURL + [url]
    r = subprocess.run(cmd, capture_output=True, text=True, timeout=25)
    if r.returncode != 0:
        raise Exception("curl failed: {} (stderr: {})".format(r.returncode, r.stderr[:200]))
    html = r.stdout
    # Verify we got real content, not JS challenge
    if "<script" in html and ("news-list" not in html and "news-title" not in html and "htmledit_views" not in html):
        if "P54" in html or "Yf1" in html or "function" in html[:200]:
            raise Exception("Got JS challenge page instead of real content")
    return html


def parse_list(html):
    """Parse list page, return list of (title, url, date)."""
    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"):
        a = li.find("a", class_="news-title", href=True)
        if not a:
            continue
        href = a["href"].strip()
        title = a.get("title") or a.get_text(strip=True)
        span = li.find("span", class_="news-time")
        date = span.get_text(strip=True) if span else ""

        if not href.startswith("http"):
            if href.startswith("/"):
                href = BASE_URL + href
            else:
                href = BASE_URL + "/" + href

        if title and href:
            items.append((title.strip(), href, date))
    return items


def parse_detail(html, url):
    """Parse detail page content."""
    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 span
    pub_date = ""
    news_inf = soup.find("div", class_="news-inf")
    if news_inf:
        spans = news_inf.find_all("span")
        for s in spans:
            text = s.get_text(strip=True)
            m = re.search(r'(\d{4}[-/]\d{1,2}[-/]\d{1,2})', text)
            if m:
                pub_date = m.group(1)
                break

    # Meta PubDate fallback
    if not pub_date:
        meta = soup.find("meta", attrs={"name": "PubDate"})
        if meta and meta.get("content"):
            pub_date = meta["content"][:10]

    # Content from article.htmledit_views
    content_html = ""
    article = soup.find("article", class_="htmledit_views")
    if article:
        content_html = article.decode_contents()
    else:
        # Fallback: news-box
        news_box = soup.find("div", class_="news-box")
        if news_box:
            content_html = news_box.decode_contents()

    # Attachments: external file links in content
    attachments = []
    if content_html:
        csoup = BeautifulSoup(content_html, "html.parser")
        for a_tag in csoup.find_all("a", href=True):
            ahref = a_tag["href"].strip()
            if any(ext in ahref.lower() for ext in [".pdf", ".doc", ".docx", ".xls", ".xlsx", ".zip", ".rar", ".jpg", ".png"]):
                if not ahref.startswith("http"):
                    ahref = BASE_URL + ahref if ahref.startswith("/") else BASE_URL + "/" + ahref
                attachments.append({
                    "title": a_tag.get_text(strip=True) or "附件",
                    "url": ahref
                })

    return {
        "title": title,
        "date": pub_date,
        "content": content_html,
        "attachments": attachments,
        "source_url": url,
    }


def render_content(content_html, title, url):
    """Convert HTML to readable text with paragraphs and tables."""
    if not content_html or len(content_html.strip()) < 50:
        return "[{}]({})".format(title, url)

    soup = BeautifulSoup(content_html, "html.parser")
    parts = []

    for el in soup.children:
        if el.name is None:
            text = el.strip()
            if text:
                parts.append(text)
        elif el.name == "p":
            text = el.get_text(" ", strip=True)
            if text:
                parts.append(text)
        elif el.name == "table":
            # Preserve as HTML table
            table_html = str(el)
            parts.append("<table>" + table_html[table_html.index(">"):].rsplit("</table>", 1)[0] + "</table>")
        elif el.name == "div":
            text = el.get_text(" ", strip=True)
            if text:
                parts.append(text)
        elif el.name == "br":
            pass
        elif el.name == "a":
            ahref = el.get("href", "")
            atext = el.get_text(strip=True)
            if atext and ahref and not ahref.startswith("javascript"):
                # Embed attachment links
                if any(ext in ahref.lower() for ext in [".pdf", ".doc", ".docx", ".xls", ".xlsx", ".zip", ".rar"]):
                    parts.append("[{}]({})".format(atext, ahref))
                elif ahref.startswith("http"):
                    parts.append("[{}]({})".format(atext, ahref))
                else:
                    parts.append(atext)

    result = "\n\n".join(p for p in parts if p).strip()
    result = re.sub(r"\n{3,}", "\n\n", result)

    if len(result.strip()) < 20:
        return "[{}]({})".format(title, url)
    return result


def is_incremental():
    return len(sys.argv) > 1 and sys.argv[1] == "incremental"


def main():
    incremental = is_incremental()
    print("[{}] Starting crawl (incremental={}, max_pages={})".format(SITE_NAME, incremental, MAX_PAGES))

    # Fetch list pages
    all_items = []
    for page in range(1, MAX_PAGES + 1):
        if page == 1:
            page_url = BASE_URL + LIST_PATH + ".jhtml"
        else:
            page_url = "{}_{}.jhtml".format(BASE_URL + LIST_PATH, page)

        print("  Fetching page {}: {}".format(page, page_url))
        try:
            html = fetch(page_url)
        except Exception as e:
            print("    ERROR: {}".format(e))
            continue
        items = parse_list(html)
        print("    Found {} items".format(len(items)))
        if not items:
            break
        all_items.extend(items)

    print("\n  Total items from list: {}".format(len(all_items)))
    if not all_items:
        print("  No items found, exiting")
        return

    # Incremental filter
    if incremental:
        cutoff_date = datetime.now() - timedelta(days=INCREMENTAL_DAYS)
        print("  Incremental mode: cutoff = {}".format(cutoff_date.date()))
        filtered = [(t, u, d) for t, u, d in all_items if d and _date_filter(d, cutoff_date)]
        all_items = filtered
        print("  After incremental filter: {} items".format(len(all_items)))

    # Fetch details
    results = []
    for idx, (title, url, date_from_list) in enumerate(all_items):
        print("  [{}/{}] Fetching: {}...".format(idx + 1, len(all_items), title[:40]))
        try:
            html = fetch(url)
        except Exception as e:
            print("    ERROR: {}".format(e))
            results.append({
                "title": title, "date": date_from_list,
                "content": "[{}]({})".format(title, url),
                "attachments": [], "source_url": url,
            })
            continue

        detail = parse_detail(html, url)
        if not detail["title"]:
            detail["title"] = title
        if not detail["date"]:
            detail["date"] = date_from_list

        detail["content"] = render_content(detail["content"], detail["title"], url)
        results.append(detail)

    print("\n  Total valid items: {}".format(len(results)))
    if not results:
        print("  No items to import, exiting")
        return

    # Write JSONL
    tmp = tempfile.NamedTemporaryFile(mode="w", suffix=".jsonl", delete=False, dir="/tmp")
    label = "{} ({}条)".format(SITE_NAME, len(results))
    for r in results:
        line = {
            "title": r["title"],
            "page_url": r["source_url"],
            "content": r["content"],
            "publish_date": r["date"],
            "site_name": SITE_NAME,
            "source_url": r["source_url"],
            "attachments": r.get("attachments", []),
        }
        tmp.write(json.dumps(line, ensure_ascii=False) + "\n")
    tmp.close()
    print("  JSONL written: {}".format(tmp.name))

    # Import
    ret = os.system("python3 /root/gov_crawler/import_jsonl.py '{}' '{}'".format(tmp.name, label))
    if ret == 0:
        print("  OK Import complete")
    else:
        print("  FAIL Import failed (exit={})".format(ret))
        sys.exit(1)

    try:
        os.unlink(tmp.name)
    except OSError:
        pass  # import_jsonl.py 已消费并删除


def _date_filter(date_str, cutoff):
    try:
        d = datetime.strptime(date_str, "%Y-%m-%d")
        return d >= cutoff
    except ValueError:
        return True


if __name__ == "__main__":
    main()