#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
crawl_ysia_zdxm.py - 榆林经济技术开发区-重大项目
CMS: TRS, index_N.html 静态分页
列表: /zwgk/fdzdgknr/zdxm/index.html (page1), index_1.html (page2+)
     20条/页, 64条, 4页
详情: /zwgk/fdzdgknr/zdxm/{YYYYMM}/t{YYYYMMDD}_{id}.html
标题: <meta ArticleTitle>
日期: <meta PubDate>
正文: div.TRS_UEDITOR > p
"""
import requests
import re
import sys
import os
import time
from bs4 import BeautifulSoup

DB_PATH = "/root/search.db"
BASE_URL = "https://ysia.yl.gov.cn"
LIST_PATH = "/zwgk/fdzdgknr/zdxm"
SITE_NAME = "榆林经济技术开发区-重大项目"
GROUP = "陕西"
INDUSTRY = "环境公示"
MAX_PAGES = 5

HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36",
}

SESSION = requests.Session()


def get(url):
    try:
        r = SESSION.get(url, headers=HEADERS, timeout=60)
        r.encoding = "utf-8"
        return r.text
    except Exception as e:
        print(f"ERR:{e}")
        return None


def parse_list(html, base_url):
    """Extract article links from list page"""
    items = []
    soup = BeautifulSoup(html, "html.parser")
    for a in soup.find_all("a", href=True):
        h = a["href"]
        # Match ./YYYYMM/tYYYYMMDD_id.html pattern
        if h.startswith("./") and h.endswith(".html") and "/t2" in h:
            full_url = base_url.rstrip("/") + "/" + h[2:]
            title = a.get_text(strip=True)
            if title:
                items.append((full_url, title))
    return items


def extract_detail(html, url):
    soup = BeautifulSoup(html, "html.parser")

    # Title
    title = ""
    meta_title = soup.find("meta", {"name": "ArticleTitle"})
    if meta_title:
        title = meta_title.get("content", "").strip()

    # Date
    date_str = ""
    pubdate_meta = soup.find("meta", {"name": "PubDate"})
    if pubdate_meta:
        content = pubdate_meta.get("content", "")
        m = re.search(r"(\d{4}-\d{2}-\d{2})", content)
        if m:
            date_str = m.group(1)

    # Content from TRS_UEDITOR
    content = ""
    ue = soup.find(class_=lambda c: c and isinstance(c, str) and "TRS_UEDITOR" in c)
    if ue:
        parts = []
        for p in ue.find_all(["p", "section"]):
            txt = p.get_text("", strip=True)
            if txt:
                parts.append(txt)
        for tbl in ue.find_all("table"):
            parts.append(str(tbl))
        if parts:
            content = "\n\n".join(parts)
        else:
            content = ue.get_text(separator="\n").strip()

    if content:
        content = re.sub(r"&nbsp;", "", content)

    summary = content[:200] if content else ""
    return title, date_str, content, summary


def save_to_db(items_data):
    import sqlite3
    conn = sqlite3.connect(DB_PATH, timeout=10)
    c = conn.cursor()
    inserted = 0
    fts_batch = []
    for title, date_str, content, summary, page_url in items_data:
        try:
            c.execute("SELECT id FROM gov_raw WHERE page_url = ?", (page_url,))
            if c.fetchone():
                continue
            c.execute(
                """INSERT INTO gov_raw 
                   (page_url, title, site_name, publish_date, content, summary,
                    source_url, category, industry, group_name, script_name)
                   VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)""",
                (page_url, title, SITE_NAME, date_str, content, summary,
                 page_url, "重大项目", INDUSTRY, GROUP, "crawl_ysia_zdxm.py"),
            )
            new_id = c.lastrowid
            if new_id:
                fts_batch.append((new_id, title, SITE_NAME, summary or ""))
            inserted += 1
        except Exception as e:
            print(f"  [DB ERROR] {title[:30]}: {e}")
    conn.commit()
    if fts_batch:
        for rid, title, site, summary in fts_batch:
            try:
                c.execute(
                    "INSERT OR IGNORE INTO gov_search(rowid, title, site_name, summary) VALUES (?, ?, ?, ?)",
                    (rid, title, site, summary),
                )
            except Exception as e:
                print(f"  [FTS ERROR] id={rid}: {e}")
        conn.commit()
        print(f"  FTS同步: {len(fts_batch)}条")
    conn.close()
    return inserted


def main():
    import argparse
    parser = argparse.ArgumentParser(description=f"{SITE_NAME}爬虫")
    parser.add_argument("--pages", type=int, default=MAX_PAGES, help="爬取页数")
    args = parser.parse_args()

    max_pages = args.pages
    print(f"[{time.strftime('%H:%M:%S')}] {SITE_NAME}...", flush=True)

    list_base = BASE_URL + LIST_PATH
    all_items = []

    for pg in range(1, max_pages + 1):
        if pg == 1:
            url = list_base + "/index.html"
        else:
            url = list_base + f"/index_{pg-1}.html"
        html = get(url)
        if not html:
            print(f"  第{pg}页: 请求失败", flush=True)
            continue
        items = parse_list(html, list_base)
        print(f"  第{pg}页: {len(items)}条", flush=True)
        if not items:
            print(f"  第{pg}页无数据，停止", flush=True)
            break
        all_items.extend(items)

    if not all_items:
        print("无数据")
        return

    print(f"\n共获取 {len(all_items)} 条链接", flush=True)

    batch = []
    for href, list_title in all_items:
        detail_html = get(href)
        if not detail_html:
            continue
        dt, dd, content, summary = extract_detail(detail_html, href)
        batch.append((dt or list_title, dd, content, summary, href))
        time.sleep(0.3)

    saved = save_to_db(batch)
    print(f"入库{saved}/{len(batch)}条", flush=True)

    print(f"\n===== 完成 =====")
    print(f"新增入库: {saved} 条")
    print(f"站点: {SITE_NAME}")


if __name__ == "__main__":
    main()