#!/usr/bin/env python3
import os
"""Crawl gzhgyjy.com - 贵州省化工研究院 信息公示"""
import re, sys, os, json, time
import urllib.request, urllib.error
from bs4 import BeautifulSoup
import sqlite3

BASE_URL = "http://www.gzhgyjy.com"
SITE_NAME = "贵州省化工研究院-信息公示"
GROUP_NAME = "企业环评"
DB = os.environ.get("SEARCH_DB", "/root/search.db")
HEADERS = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}

_MAX_PG = None
for i, a in enumerate(sys.argv):
    if a == "--pages" and i + 1 < len(sys.argv):
        _MAX_PG = int(sys.argv[i + 1])
        break


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"""
    import urllib.parse
    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 fetch(url):
    req = urllib.request.Request(url, headers=HEADERS)
    try:
        return urllib.request.urlopen(req, timeout=30).read().decode("utf-8", errors="replace")
    except Exception as e:
        print(f"[WARN] fetch failed: {url} - {e}", file=sys.stderr)
        return ""


def extract_items(html):
    """Extract items from list page"""
    items = []
    for m in re.finditer(
        r'<a[^>]*href="(NewsShow\.asp\?[^"]+)"[^>]*>([^<]+)</a>',
        html
    ):
        href = m.group(1)
        title = m.group(2).strip()
        if not title or len(title) < 5:
            continue
        if not href.startswith("http"):
            href = BASE_URL + "/" + href if not href.startswith("/") else BASE_URL + href

        # Find date near this link
        date = ""
        date_m = re.search(r"(\d{4}-\d{2}-\d{2})", html[m.end():m.end()+300])
        if date_m:
            date = date_m.group(1)

        items.append({"title": title, "url": href, "date": date})
    return items


def parse_detail(html, url):
    """Parse detail page"""
    soup = BeautifulSoup(html, "html.parser")

    # Title - from h1 or title area
    title = ""
    h1 = soup.select_one("h1")
    if h1:
        title = h1.get_text(strip=True)
    if not title:
        # Try breadcrumb or page title
        for sel in [".page-title", ".title", ".article-title", ".ArticleTitle"]:
            el = soup.select_one(sel)
            if el:
                title = el.get_text(strip=True)
                break

    # Date
    date_text = ""
    for m in re.finditer(r"(\d{4}-\d{2}-\d{2})", html):
        date_text = m.group(1)
        break

    # Content
    content = ""
    attachments = []

    content_div = soup.select_one(".section-sub")
    if not content_div:
        for sel in ["[class*=content]", "[class*=article]", ".text", "#content", ".main-content"]:
            el = soup.select_one(sel)
            if el and len(el.get_text(" ", strip=True)) > 100:
                content_div = el
                break

    if not content_div:
        return {"title": title, "date": date_text, "content": "", "attachments": []}

    parts = []
    for el in content_div.find_all(["p", "table", "img"], recursive=True):
        if el.name == "p" and el.find_parent("table"):
            continue
        if el.name == "table" and el.find_parent("table"):
            continue

        if el.name == "img":
            src = el.get("src", "")
            if "icon_" in src.lower():
                continue
            if src.startswith("//"):
                src = "http:" + src
            elif src.startswith("/"):
                src = BASE_URL + src
            elif not src.startswith("http"):
                src = url.rsplit("/", 1)[0] + "/" + src
            alt = el.get("alt", "") or ""
            parts.append(f"![{alt}]({src})")
            continue

        if el.name == 'table':
            tbl_html = html_table_to_html(el, url)
            if tbl_html:
                parts.append(tbl_html)
                continue  # 表格已作富文本入正文, 避免再拍平
        text = el.get_text(" ", strip=True)
        if text:
            for a in el.find_all("a", href=True):
                ah = a["href"]
                if re.search(r'\.(docx?|pdf|xlsx?|rar|zip)$', ah, re.I):
                    atitle = a.get_text(strip=True) or "附件"
                    if ah.startswith("/"):
                        ah = BASE_URL + ah
                    attachments.append({"title": atitle, "url": ah})
            parts.append(text)

    content = "\n\n".join(p for p in parts if p.strip())
    content = content.replace("\u00a0", " ").replace("&nbsp;", " ")

    seen = set()
    unique_att = []
    for att in attachments:
        if not isinstance(att, dict):
            unique_att.append(att)
            continue
        if att["url"] not in seen:
            seen.add(att["url"])
            unique_att.append(att)

    return {"title": title, "date": date_text, "content": content, "attachments": unique_att}


def main():
    conn = sqlite3.connect(DB, timeout=60)
    conn.execute("PRAGMA busy_timeout=60000")
    c = conn.cursor()

# QC20260925 去掉整站清空再重灌(抢锁+中途死掉会清空整站; page_url 有 UNIQUE 索引，插入本就幂等)     c.execute("DELETE FROM gov_search WHERE site_name=?", (SITE_NAME,))
# QC20260925 去掉整站清空再重灌(抢锁+中途死掉会清空整站; page_url 有 UNIQUE 索引，插入本就幂等)     c.execute("DELETE FROM gov_raw WHERE site_name=?", (SITE_NAME,))
    conn.commit()
    print("[清理] 已清理该站旧数据", flush=True)

    all_items = []
    total_pages = 6
    if _MAX_PG:
        total_pages = min(total_pages, _MAX_PG)

    for pg in range(1, total_pages + 1):
        url = f"{BASE_URL}/News.asp?clsid=0&id=3"
        if pg > 1:
            url += f"&page={pg}"
        html = fetch(url)
        if not html:
            break
        items = extract_items(html)
        if not items:
            break
        print(f"[INFO] Page {pg}: {len(items)} items", flush=True)
        all_items.extend(items)

    # Remove duplicates by URL
    seen_set = set()
    deduped = []
    for it in all_items:
        if it["url"] not in seen_set:
            seen_set.add(it["url"])
            deduped.append(it)
    all_items = deduped

    print(f"[INFO] Total items from list: {len(all_items)}", flush=True)

    new_count = 0
    skip_count = 0
    empty_count = 0
    total_att = 0

    for idx, item in enumerate(all_items):
        c.execute("SELECT id FROM gov_raw WHERE page_url=?", (item["url"],))
        if c.fetchone():
            skip_count += 1
            continue

        if idx % 5 == 0:
            print(f"[INFO] [{idx}/{len(all_items)}] {item['title'][:40]}...", file=sys.stderr, end=" ", flush=True)

        html = fetch(item["url"])
        if not html:
            print("FETCH FAILED", file=sys.stderr)
            continue

        detail = parse_detail(html, item["url"])
        title = detail["title"] or item["title"]
        date_text = detail["date"] or item["date"]
        content = detail["content"]
        attachments_list = detail["attachments"]

        if not content.strip():
            empty_count += 1
            if idx % 5 == 0:
                print("EMPTY", file=sys.stderr)
            continue

        plain = re.sub(r'\s+', ' ', content).strip()
        summary = plain[:200] if plain else title[:200]
        att_json = json.dumps(attachments_list, ensure_ascii=False) if attachments_list else "[]"

        date_rank = 0
        if date_text:
            m2 = re.search(r"(\d{4})-(\d{2})-(\d{2})", date_text)
            if m2:
                date_rank = int(m2.group(1) + m2.group(2) + m2.group(3))

        try:
            c.execute(
                """INSERT OR IGNORE INTO gov_raw 
                (title, site_name, group_name, page_url, publish_date, content, summary, attachments, date_rank)
                VALUES (?,?,?,?,?,?,?,?,?)""",
                (title, SITE_NAME, GROUP_NAME, item["url"], date_text, content, summary, att_json, date_rank)
            )
            if c.rowcount > 0:
                new_count += 1
                total_att += len(attachments_list)
        except Exception as e:
            if idx % 5 == 0:
                print(f"DB ERROR: {e}", file=sys.stderr)

        time.sleep(0.3)

    conn.commit()
    conn.close()

    print(f"\n[RESULT] {SITE_NAME}: {new_count} new, {skip_count} skip, {empty_count} empty, {total_att} attachments", flush=True)


if __name__ == "__main__":
    main()