#!/usr/bin/env python3
"""
天门市人民政府-环评公众参与 (www.tianmen.gov.cn)
===========================================
列表: index.shtml, index_1.shtml, ... 每页20条
详情: /zwgk/zdlyxxgk/hjbh/hpgzcy/202606/t20260608_5954048.shtml
      <div class="article_content"> → <div class="view TRS_UEDITOR"> 含正文

用法:
    python3 crawl_tianmen_hjbh.py          # 全量
    python3 crawl_tianmen_hjbh.py --test   # 测试 5 条
"""

import re, sys, os, time
from datetime import datetime, timedelta, timezone
import requests, urllib3
import os
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

SITE_NAME  = "天门市-环评公众参与"
LIST_BASE  = "https://www.tianmen.gov.cn/zwgk/zdlyxxgk/hjbh/hpgzcy/"
SEARCH_DB = os.getenv("SEARCH_DB", "/root/search.db")
MAX_PAGES = 5  # 3yr cutoff at page 6
CUTOFF     = (datetime.now(timezone.utc) - timedelta(days=365*3)).strftime("%Y-%m-%d")
HEADERS    = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}

def fetch_list(page):
    if page == 1:
        url = LIST_BASE + "index.shtml"
    else:
        url = LIST_BASE + f"index_{page-1}.shtml"
    try:
        r = requests.get(url, headers=HEADERS, timeout=20, verify=False)
        r.encoding = "utf-8"
        return r.text
    except Exception as e:
        print(f"  ! list page {page} fail: {str(e)[:60]}")
        return None

def parse_list(html):
    items = []
    for m in re.finditer(
        r'<li><span>(\d{4}-\d{2}-\d{2})</span>\s*<a[^>]*href="([^"]*)"[^>]*title="([^"]*)"',
        html, re.DOTALL
    ):
        pub_date = m.group(1)
        url = m.group(2)
        if not url.startswith("http"):
            if url.startswith("./"):
                url = LIST_BASE + url[2:]
            else:
                url = "https://www.tianmen.gov.cn" + url if url.startswith("/") else LIST_BASE + url
        title = m.group(3).strip()
        items.append({"title": title, "url": url, "pub_date": pub_date})
    return items

def fetch_detail(url):
    try:
        r = requests.get(url, headers=HEADERS, timeout=20, verify=False)
        r.encoding = "utf-8"
        html = r.text
        # Content in <div class="view TRS_UEDITOR ..."> inside article_content
        m = re.search(r'<div class="view[^>]*>(.*?)</div>\s*</div>\s*</div>', html, re.DOTALL)
        if m:
            content = m.group(1)
            content = re.sub(r'<script[^>]*>.*?</script>', '', content, flags=re.DOTALL|re.I)
            content = re.sub(r'<style[^>]*>.*?</style>', '', content, flags=re.DOTALL|re.I)
            content = content.strip()
            return content
        return ""
    except Exception as e:
        print(f"  ! detail fail: {str(e)[:60]}")
        return ""

def to_db(items):
    if not items:
        return 0, 0
    import sqlite3
    db = sqlite3.connect(SEARCH_DB, timeout=60)
    db.execute("PRAGMA journal_mode=WAL")
    db.execute("PRAGMA synchronous=NORMAL")
    ok, fail = 0, 0
    for it in items:
        try:
            db.execute(
                "INSERT OR REPLACE INTO gov_raw "
                "(site_name, title, page_url, content, publish_date, summary, tags) "
                "VALUES (?,?,?,?,?,?,?)",
                (
                    SITE_NAME,
                    (it.get("title") or "")[:500],
                    it.get("url", ""),
                    it.get("content", ""),
                    (it.get("pub_date") or "")[:10],
                    "",
                    "环评公众参与",
                )
            )
            if db.total_changes > 0:
                ok += 1
            else:
                fail += 1
        except Exception as e:
            fail += 1
    db.execute(
        "INSERT INTO gov_search(rowid, title, site_name, summary) "
        "SELECT r.id, r.title, r.site_name, r.summary "
        "FROM gov_raw r WHERE r.id NOT IN (SELECT rowid FROM gov_search) AND r.site_name=?",
        (SITE_NAME,)
    )
    db.commit()
    db.close()
    return ok, fail

def crawl(test=False):
    all_items = []
    seen_urls = set()
    for page in range(1, MAX_PAGES + 1):
        html = fetch_list(page)
        if not html:
            break
        items = parse_list(html)
        if not items:
            print(f"  [Page {page}] empty, reached end")
            break
        new = 0
        for item in items:
            if item["url"] in seen_urls:
                continue
            seen_urls.add(item["url"])
            if item["pub_date"] and item["pub_date"] < CUTOFF:
                continue
            all_items.append(item)
            new += 1
        f = items[0]["pub_date"] if items else "?"
        l = items[-1]["pub_date"] if items else "?"
        print(f"  [Page {page}] {len(items)} items ({f} ~ {l}), new: {new}")
        if new == 0:
            break
    print(f"  Total: {len(all_items)} items within 3yr")
    if test:
        all_items = all_items[:5]
        print(f"  TEST mode: {len(all_items)} items")
    if not all_items:
        return 0, 0
    for i, item in enumerate(all_items):
        print(f"  [{i+1}/{len(all_items)}] {item['title'][:40]}... ", end="", flush=True)
        content = fetch_detail(item["url"])
        item["content"] = content
        print(f"{len(content)}B")
    ok, fail = to_db(all_items)
    return ok, fail

if __name__ == "__main__":
    test = "--test" in sys.argv
    mode = "TEST" if test else "FULL"
    print()
    print(f"[{SITE_NAME}] {mode}")
    t0 = time.time()
    ok, fail = crawl(test=test)
    print(f"  Time: {round(time.time()-t0, 1)}s")
    print(f"  New: {ok} Skip: {fail}")