test: skip network/live crawl by default; fix settings e2e key

- pytest_collection_modifyitems: skip @pytest.mark.network unless PROXYPOOL_RUN_NETWORK_TESTS=1
- Document opt-in in tests/README.md
- e2e: replace removed crawl_timeout with validation_timeout

Made-with: Cursor
This commit is contained in:
祀梦
2026-04-05 14:33:29 +08:00
parent e582067316
commit ce667dba13
3 changed files with 34 additions and 12 deletions

View File

@@ -24,22 +24,24 @@ tests/
## 网络与真实调用
集成测试与 E2E **不再 mock** `PluginRunner` / `ValidatorService`:会发起真实 HTTP 爬取与代理验证(视设置而定)。运行全量 `pytest` 需要 **可用的出站网络**,且含 `network` / `slow` 标记的用例可能耗时数分钟
`@pytest.mark.network` 的用例会发起真实 HTTP 爬取,**默认不运行**(避免全量 `pytest` 卡住数十分种)
跳过需外网的用例(例如离线快速检查
需要跑外网验收时PowerShell 示例
```powershell
$env:PROXYPOOL_RUN_NETWORK_TESTS="1"; pytest tests/integration/test_plugins_live_crawl.py -v --tb=short
```
离线快速检查直接:
```bash
pytest -m "not network"
pytest
```
**插件爬取验收**`test_plugins_live_crawl.py`
- 核心 8 插件:必须至少 1 条代理且无 Runner 失败。
- `fpw_*`对照 [Free_Proxy_Website](https://github.com/cyubuchen/Free_Proxy_Website) 的公开源,允许 0 条(国际网络差异),使用更长超时
```bash
pytest tests/integration/test_plugins_live_crawl.py -v
```
- 核心 8 插件:必须至少 1 条代理且无 Runner 失败(依赖稳定外网)
- `fpw_*`允许部分源 0 条;单插件可能接近各插件 `crawl_timeout_seconds` 上限
## 运行测试

View File

@@ -11,6 +11,26 @@ if sys.platform == "win32":
asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())
import pytest
def _network_tests_enabled() -> bool:
v = os.environ.get("PROXYPOOL_RUN_NETWORK_TESTS", "").strip().lower()
return v in ("1", "true", "yes", "on")
def pytest_collection_modifyitems(config, items) -> None:
"""默认跳过 @pytest.mark.network避免全量 pytest 触发真实爬取卡住数十分种。"""
if _network_tests_enabled():
return
skip = pytest.mark.skip(
reason=(
"外网/真实爬取用例默认跳过。需要验收时设置环境变量 "
"PROXYPOOL_RUN_NETWORK_TESTS=1 后再运行对应文件或 -m network。"
)
)
for item in items:
if "network" in item.keywords:
item.add_marker(skip)
import pytest_asyncio
from typing import AsyncGenerator
from httpx import AsyncClient, ASGITransport

View File

@@ -155,9 +155,9 @@ class TestFullWorkflow:
assert response.status_code == 200
original_settings = response.json()["data"]
# 2. 修改设置
# 2. 修改设置(爬取限时已改为各插件 crawl_timeout_seconds不再使用全局 crawl_timeout
new_settings = original_settings.copy()
new_settings["crawl_timeout"] = 45
new_settings["validation_timeout"] = 8
new_settings["auto_validate"] = not original_settings["auto_validate"]
response = await client.post("/api/settings", json=new_settings)
@@ -166,7 +166,7 @@ class TestFullWorkflow:
# 3. 验证设置已保存
response = await client.get("/api/settings")
saved_settings = response.json()["data"]
assert saved_settings["crawl_timeout"] == 45
assert saved_settings["validation_timeout"] == 8
# 4. 恢复原始设置
response = await client.post("/api/settings", json=original_settings)