拆分抓取与交易服务

把需要账号登录态的链路从抓取服务里拆出成独立进程。分界线不是「要不要登录」，而是抓取无状态、幂等、可多开实例，而交易的写操作不可逆、登录态全局唯一、订单监控是常驻轮询——同进程时抓取一扩容就会复制出 N 份登录态与 N 个轮询，同一账号会被并发操作。 - app/shared：配置、错误码、日志、ApiResponse 信封 + Bearer 鉴权 + 异常处理器、导航请求头构造器 - app/scraping：站点常量、会话、解析器与 10 个抓取接口，:31107，可多开 - app/trading：登录态查询/重载与健康检查，:31108，只能单实例 - 依赖方向锁为 scraping→shared、trading→shared，两侧互不 import； tests/test_architecture.py 用 AST 检查 import 并校验两个 app 的路径不串 - 登录态 UA 在 trading 独立持有：与抓取 UA 值相同但变更理由不同，抓取 UA 为绕反爬可随时调整，登录 UA 一改可能触发设备校验使已落盘 cookie 失效 - scripts/login.py 与 AuthSession 共用 auth_site.PROFILES 与 is_logged_in，判据只写一遍 - 同一镜像两个启动命令，交易容器覆盖 command 并设 RAKUTEN_HEALTH_PORT 同时带上此前未提交的ラクマ分类接口与登录态基础设施。验证：239 个离线用例全绿；两个入口真实启动，/health 与鉴权正常。未验证：真实探测登录态（当前开发机无外网，对站点的连接全部超时）。 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-07-27 15:05:01 +08:00
parent 4250388762
commit 104d7fef6b
80 changed files with 2330 additions and 402 deletions
@@ -0,0 +1,157 @@
+"""交易服务 API 测试：健康检查、登录态查询与重载、鉴权
+
+登录态会话被替换为桩，不触达真实站点、不需要真实账号。
+AuthSession 自身的行为（cookie 加载、探测判据）在 tests/test_auth_session.py。
+"""
+from __future__ import annotations
+
+import pytest
+from fastapi.testclient import TestClient
+
+from app.shared.config import get_settings
+from app.shared.errors import NotLoggedInError
+from app.trading.main import create_app
+from app.trading.services.auth_session import AuthStatus
+
+TOKEN = get_settings().bearer_token
+AUTH = {"Authorization": f"Bearer {TOKEN}"}
+
+
+class StubAuthSession:
+    """记录调用并返回固定登录态的桩"""
+
+    def __init__(self) -> None:
+        self.checked: list[str] = []
+        self.reloaded: list[str] = []
+        self.logged_in = True
+
+    @property
+    def sites(self) -> tuple[str, ...]:
+        return ("rakuten", "rakuma")
+
+    def _status(self, site: str) -> AuthStatus:
+        return AuthStatus(
+            site=site,
+            state_file_exists=True,
+            logged_in=self.logged_in,
+            checked_at=None,
+            detail="stub",
+        )
+
+    async def check(self, site: str) -> AuthStatus:
+        self.checked.append(site)
+        return self._status(site)
+
+    def status(self, site: str) -> AuthStatus:
+        return self._status(site)
+
+    def status_all(self) -> dict[str, dict]:
+        return {site: self._status(site).to_dict() for site in self.sites}
+
+    def reload(self, site: str) -> int:
+        self.reloaded.append(site)
+        return 3
+
+    async def require_logged_in(self, site: str) -> None:
+        if not self.logged_in:
+            raise NotLoggedInError(site=site, detail="stub")
+
+    async def close(self) -> None:
+        """lifespan 收尾会调用；桩没有真实客户端要关"""
+
+
+@pytest.fixture
+def client_and_stub():
+    app = create_app()
+    with TestClient(app) as client:
+        stub = StubAuthSession()
+        app.state.container.auth_session = stub
+        yield client, stub
+
+
+@pytest.fixture
+def client(client_and_stub):
+    return client_and_stub[0]
+
+
+@pytest.fixture
+def stub(client_and_stub):
+    return client_and_stub[1]
+
+
+# ---- 健康检查 ----
+
+
+def test_health_needs_no_token(client):
+    response = client.get("/health")
+    assert response.status_code == 200
+    body = response.json()
+    assert body["data"]["status"] == "ok"
+    assert set(body["data"]["auth"]) == {"rakuten", "rakuma"}
+
+
+def test_health_does_not_probe_the_site(client, stub):
+    """健康检查只读缓存：它会被高频轮询，不能每次都去打站点"""
+    client.get("/health")
+    assert stub.checked == []
+
+
+# ---- 鉴权 ----
+
+
+@pytest.mark.parametrize("path", ["/api/auth/status", "/api/auth/reload"])
+def test_auth_endpoints_reject_missing_token(client, path):
+    response = client.post(path, json={})
+    assert response.status_code == 401
+    assert response.json()["code"] == 1001
+
+
+def test_auth_endpoints_reject_wrong_token(client):
+    response = client.post(
+        "/api/auth/status", json={}, headers={"Authorization": "Bearer wrong-token"}
+    )
+    assert response.status_code == 401
+
+
+# ---- 登录态查询 ----
+
+
+def test_status_probes_both_sites_by_default(client, stub):
+    response = client.post("/api/auth/status", json={}, headers=AUTH)
+    assert response.status_code == 200
+    body = response.json()
+    assert [item["site"] for item in body["data"]["sites"]] == ["rakuten", "rakuma"]
+    assert stub.checked == ["rakuten", "rakuma"]
+
+
+def test_status_can_skip_the_probe(client, stub):
+    """refresh=false 时读缓存，不打站点"""
+    response = client.post("/api/auth/status", json={"refresh": False}, headers=AUTH)
+    assert response.status_code == 200
+    assert stub.checked == []
+
+
+def test_status_accepts_a_single_site(client, stub):
+    response = client.post("/api/auth/status", json={"site": "rakuma"}, headers=AUTH)
+    assert response.status_code == 200
+    assert stub.checked == ["rakuma"]
+
+
+def test_status_rejects_unknown_site(client):
+    """站点名是枚举，未知值应在校验层就被挡下"""
+    response = client.post("/api/auth/status", json={"site": "mercari"}, headers=AUTH)
+    assert response.status_code == 422
+    assert response.json()["code"] == 1002
+
+
+# ---- 登录态重载 ----
+
+
+def test_reload_reloads_then_probes(client, stub):
+    response = client.post("/api/auth/reload", json={"site": "rakuten"}, headers=AUTH)
+    assert response.status_code == 200
+    body = response.json()
+    assert body["data"]["reloaded"] == {"rakuten": 3}
+    # 重载后必须立刻探测一次，否则调用方拿不到「这次登录到底成没成」
+    assert stub.reloaded == ["rakuten"]
+    assert stub.checked == ["rakuten"]