#!/usr/bin/env python3
import requests, re, json

url = "https://www.jinjiang.gov.cn/ztzl/lhzt/lyxx/sthj/"
headers = {"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36"}
r = requests.get(url, timeout=30, headers=headers)
html = r.text

# Look for SSP configuration / API endpoints
patterns = ['ssp', 'api', 'search', 'pageSize', 'pageSize', 'siteId', 'channelId', 'docType', 'category', 'terton']
for p in patterns:
    indices = [m.start() for m in re.finditer(p, html, re.I)]
    for idx in indices[:5]:
        start = max(0, idx-50)
        end = min(len(html), idx+200)
        print(f"=== '{p}' at pos {idx} ===")
        print(html[start:end])
        print()

# Look for JSON data blocks
print("=== JSON-like data blocks ===")
for m in re.finditer(r'\{[^}]{100,500}\}', html):
    try:
        data = json.loads(m.group())
        print(json.dumps(data, ensure_ascii=False)[:300])
        print()
    except:
        pass

# Look for ms-controller and ms-avalon data
print("=== avalon data ===")
for m in re.finditer(r'avalon\.(?:define|scan|vmodels)', html):
    start = max(0, m.start()-100)
    end = min(len(html), m.end()+200)
    print(html[start:end])
    print("---")

# Look for the data embedded in HTML attributes
print("=== ms- data ===")
for m in re.finditer(r'ms-[\w]+=["\'][^"\']+["\']', html):
    print(m.group()[:150])