#!/usr/bin/env python3
"""整站 GET 接口全量扫描：对接口清单逐一探测并留存响应。只读、限速。"""
import json, os, re, time, urllib.request, urllib.error

BASE = "https://zzc-serv.thinkerx.com/api/"
TOKEN = "eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJodHRwczpcL1wvenpjLXNlcnYudGhpbmtlcnguY29tXC9hcGlcL3dlY2hhdC1sb2dpbi13aXRoLXBob25lIiwiaWF0IjoxNzg5OTA5NTIwLCJleHAiOjE3ODk5OTU5MjAsIm5iZiI6MTc4OTkwOTUyMCwianRpIjoiTGVyV0J4VTZONTVOMlNwQSIsInN1YiI6MzU2Mjg1NiwicHJ2IjoiNTRjMjZlNjE2MjNhYTQzMzZkZTQ1YTcxNGIwYzdhYWIxZTJhYzA3NSJ9.MXBrpBh3gseq8I91EQklxsgLfHafUKwNxc24JdcUkcc"
OUT = r"D:\Code\huamenchuang\data\full-site\get"
os.makedirs(OUT, exist_ok=True)

def safe_name(ep):
    return re.sub(r'[^A-Za-z0-9_-]', '_', ep).strip('_') or 'root'

def fetch(ep, method="GET", body=None):
    req = urllib.request.Request(BASE + ep, method=method, headers={
        "Authorization": f"Bearer {TOKEN}",
        "X-Requested-With": "XMLHttpRequest",
        "Accept": "application/json",
        "Content-Type": "application/json",
    }, data=body)
    try:
        with urllib.request.urlopen(req, timeout=15) as r:
            return r.status, r.read().decode("utf-8", "replace")
    except urllib.error.HTTPError as e:
        return e.code, e.read().decode("utf-8", "replace")[:4000]
    except Exception as e:
        return -1, str(e)[:200]

endpoints = [l.strip() for l in open(r"D:\Code\huamenchuang\data\api-endpoint-inventory.txt", encoding="utf-8") if l.strip()]
manifest = []
ok200 = 0
for i, ep in enumerate(endpoints):
    fn = safe_name(ep)
    status, body = fetch(ep)
    entry = {"endpoint": ep, "method": "GET", "status": status, "size": len(body)}
    if status == 200:
        ok200 += 1
        try:
            j = json.loads(body)
            entry["code"] = j.get("code") if isinstance(j, dict) else None
            entry["message"] = (j.get("message") or "")[:40] if isinstance(j, dict) else ""
        except Exception:
            entry["code"] = "RAW"
    json.dump({"endpoint": ep, "status": status, "body": body[:100000]},
              open(f"{OUT}\\{fn}.json", "w", encoding="utf-8"), ensure_ascii=False)
    manifest.append(entry)
    if (i + 1) % 40 == 0:
        print(f"进度 {i+1}/{len(endpoints)}  200数={ok200}", flush=True)
    time.sleep(0.25)

json.dump(manifest, open(r"D:\Code\huamenchuang\data\full-site\get-manifest.json", "w", encoding="utf-8"), ensure_ascii=False, indent=1)
print(f"完成: {len(endpoints)} 个接口, HTTP200={ok200}")
ok_valid = sum(1 for m in manifest if m.get("code") == 1)
paywall = [m["endpoint"] for m in manifest if m.get("code") == "10086"]
print(f"其中 code=1 业务成功: {ok_valid}; 付费墙拦截: {len(paywall)}")