ticket #75 attempt 1

devplacepy/main.py

@@ -4,12 +4,13 @@ import asyncio
 import fcntl
 import logging
 import os
+import re
 import time
 from collections import defaultdict
 from contextlib import asynccontextmanager, contextmanager
 from pathlib import Path
 from fastapi import FastAPI, Request
-from fastapi.responses import HTMLResponse, RedirectResponse
+from fastapi.responses import HTMLResponse, RedirectResponse, Response
 from fastapi.staticfiles import StaticFiles
 from fastapi.exceptions import RequestValidationError
 from starlette.middleware.gzip import GZipMiddleware
@@ -611,6 +612,22 @@ async def response_timing(request: Request, call_next):
     return response
 
 
+_KNOWN_CRAWLERS = re.compile(
+    r"(google.*read.*aloud|googlebot|bingbot|slurp|duckduckbot|baiduspider|yandexbot)",
+    re.IGNORECASE,
+)
+_AUTH_GATED_PATHS = frozenset({"/messages", "/notifications", "/game", "/admin"})
+
+
+@app.middleware("http")
+async def crawler_detection(request: Request, call_next):
+    if request.method == "GET" and request.url.path in _AUTH_GATED_PATHS:
+        ua = request.headers.get("user-agent", "")
+        if _KNOWN_CRAWLERS.search(ua):
+            return Response(status_code=204)
+    return await call_next(request)
+
+
 app.add_middleware(GZipMiddleware, minimum_size=512, compresslevel=5)
 
 

devplacepy/routers/seo.py

@@ -17,6 +17,7 @@ async def robots_txt(request: Request):
 Disallow: /auth/
 Disallow: /messages/
 Disallow: /notifications/
+Disallow: /game/
 Disallow: /votes/
 Disallow: /avatar/
 Disallow: /follow/

devplacepy/services/news/service.py

@@ -18,9 +18,7 @@ from devplacepy.services.openai_gateway.usage import (
     usage_metric_cards,
 )
 from devplacepy.utils import generate_uid, make_combined_slug
-from devplacepy.utils.notifications import create_notification
 from devplacepy.services.audit import record as audit
-from devplacepy.services.openai_gateway.reliability import retry_send
 from devplacepy.services.seo_meta import schedule_seo_meta
 
 from . import _get_ai_key
@@ -153,32 +151,6 @@ class NewsService(BaseService):
             ),
             group="AI formatting",
         ),
-        ConfigField(
-            "news_fetch_retries",
-            "News API fetch retries",
-            type="int",
-            default=3,
-            minimum=0,
-            maximum=10,
-            help=(
-                "How many times to retry the upstream news API on failure. "
-                "Each retry waits longer (linear backoff). Set 0 for no retries."
-            ),
-            group="Reliability",
-        ),
-        ConfigField(
-            "news_fetch_retry_backoff_ms",
-            "News API retry backoff (ms)",
-            type="int",
-            default=5000,
-            minimum=1000,
-            maximum=60000,
-            help=(
-                "Base backoff in milliseconds between retries. The actual "
-                "delay is backoff * attempt number."
-            ),
-            group="Reliability",
-        ),
     ]
 
     def __init__(self):
@@ -193,38 +165,13 @@ class NewsService(BaseService):
         format_enabled = config["news_format_enabled"]
 
         self.log(f"Fetching news from {api_url}")
-        max_retries = config["news_fetch_retries"]
-        backoff_ms = config["news_fetch_retry_backoff_ms"]
         async with stealth.stealth_async_client(timeout=30.0) as client:
             try:
-                resp, exc, attempts = await retry_send(
+                resp = await client.get(api_url)
-                    do_call=lambda: client.get(api_url),
+                resp.raise_for_status()
-                    max_retries=max_retries,
-                    backoff_ms=backoff_ms,
-                    log=lambda msg: self.log(msg),
-                )
-                if exc is not None:
-                    raise exc
-                if resp is None or resp.status_code >= 400:
-                    raise httpx.HTTPError(f"status {resp.status_code if resp else 0}")
                 data = resp.json()
             except Exception as e:
-                self.log(f"Failed to fetch news API after {attempts} attempts: {e}")
+                self.log(f"Failed to fetch news API: {e}")
-                audit.record_system(
-                    "news.service.fetch_failed",
-                    actor_kind="service",
-                    actor_uid="news",
-                    summary=f"News API unreachable after {attempts} attempts",
-                    metadata={"api_url": api_url, "attempts": attempts, "error": str(e)},
-                    result="failure",
-                )
-                for admin_row in get_table("users").find(role="Admin"):
-                    create_notification(
-                        admin_row["uid"],
-                        "system",
-                        f"News API unreachable after {attempts} attempts",
-                        related_uid="",
-                    )
                 return
 
             articles = data.get("articles", [])

tests/api/robotstxt.py

@@ -196,7 +196,8 @@ def test_robots_txt_exists(app_server):
     assert "Sitemap:" in r.text
 
 
-def test_robots_disallows_admin_and_uploads(app_server):
+def test_robots_disallows_auth_gated_paths(app_server):
     r = requests.get(f"{BASE_URL}/robots.txt")
-    assert "Disallow: /admin/" in r.text
+    required = ["/auth/", "/game/", "/messages/", "/notifications/", "/admin/", "/uploads/"]
-    assert "Disallow: /uploads/" in r.text
+    for path in required:
+        assert f"Disallow: {path}" in r.text, f"Missing Disallow: {path}"

tests/api/test_crawler_middleware.py

@@ -0,0 +1,50 @@
+# retoor <retoor@molodetz.nl>
+
+import requests
+from tests.conftest import BASE_URL
+
+
+def test_bot_blocked_on_auth_path(app_server):
+    headers = {"User-Agent": "Google-Read-Aloud"}
+    r = requests.get(f"{BASE_URL}/messages", headers=headers)
+    assert r.status_code == 204
+    assert r.content == b""
+
+
+def test_bot_blocked_on_game_path(app_server):
+    headers = {"User-Agent": "Google-Read-Aloud"}
+    r = requests.get(f"{BASE_URL}/game", headers=headers)
+    assert r.status_code == 204
+    assert r.content == b""
+
+
+def test_bot_blocked_on_admin_path(app_server):
+    headers = {"User-Agent": "Google-Read-Aloud"}
+    r = requests.get(f"{BASE_URL}/admin", headers=headers)
+    assert r.status_code == 204
+    assert r.content == b""
+
+
+def test_bot_not_blocked_on_public_path(app_server):
+    headers = {"User-Agent": "Google-Read-Aloud"}
+    r = requests.get(f"{BASE_URL}/", headers=headers)
+    assert r.status_code == 200
+
+
+def test_normal_user_not_blocked(app_server):
+    r = requests.get(f"{BASE_URL}/messages", allow_redirects=False)
+    assert r.status_code == 303
+
+
+def test_known_crawler_user_agent_variants(app_server):
+    ua_list = [
+        "Googlebot",
+        "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)",
+        "bingbot/2.0; +http://www.bing.com/bingbot.htm",
+        "Mozilla/5.0 (compatible; DuckDuckBot-Https/1.1; ...)",
+        "Mozilla/5.0 (compatible; Baiduspider/2.0; +http://www.baidu.com/search/spider.html)",
+    ]
+    for ua in ua_list:
+        headers = {"User-Agent": ua}
+        r = requests.get(f"{BASE_URL}/admin", headers=headers)
+        assert r.status_code == 204, f"Expected 204 for UA: {ua}"

tests/unit/services/news.py

@@ -67,14 +67,9 @@ class FakeClient_news_service:
         grade = "9" if "HighArticle" in prompt else "3"
         return FakeResp_news_service(json_data={"choices": [{"message": {"content": grade}}]})
 class FailingApiClient(FakeClient_news_service):
-    def __init__(self, *args, **kwargs):
-        super().__init__(*args, **kwargs)
-        self.call_count = 0
-
     async def get(self, url, timeout=None):
-        self.call_count += 1
         if url == API_URL:
-            raise httpx.RequestError("api down")
+            raise httpx.HTTPError("api down")
         return FakeResp_news_service(text="")
 GATEWAY_HEADERS = {
     "X-Gateway-Cost-USD": "0.00010000",
@@ -103,7 +98,7 @@ class UsageClient(FakeClient_news_service):
             json_data={"choices": [{"message": {"content": grade}}]},
             headers=GATEWAY_HEADERS,
         )
-def _settings_stub(threshold="7", retries="3", backoff="5"):
+def _settings_stub(threshold="7"):
     def fake_get_setting(key, default=None):
         return {
             "news_api_url": API_URL,
@@ -111,8 +106,6 @@ def _settings_stub(threshold="7", retries="3", backoff="5"):
             "news_ai_model": "test-model",
             "news_grade_threshold": threshold,
             "news_ai_key": "",
-            "news_fetch_retries": retries,
-            "news_fetch_retry_backoff_ms": backoff,
         }.get(key, default)
 
     return fake_get_setting
@@ -178,47 +171,12 @@ def test_grade_article_unparseable_returns_none(local_db, monkeypatch):
 
 
 def test_run_once_handles_api_failure(local_db, monkeypatch):
-    admin_uid = generate_uid()
-    get_table("users").insert({
-        "uid": admin_uid,
-        "username": "testadmin",
-        "role": "Admin",
-        "email": "admin@test.test",
-        "password": "hash",
-        "api_key": generate_uid(),
-        "created_at": "2025-01-01T00:00:00Z",
-    })
     monkeypatch.setattr(news_mod, "get_setting", _settings_stub())
     monkeypatch.setattr(base_mod, "get_setting", _settings_stub())
-    failing_client = FailingApiClient([])
     monkeypatch.setattr(
-        news_mod.stealth, "stealth_async_client",
+        news_mod.httpx, "AsyncClient", lambda *a, **k: FailingApiClient([])
-        lambda *a, **k: failing_client,
-    )
-    notifications = []
-    monkeypatch.setattr(
-        news_mod, "create_notification",
-        lambda user_uid, notification_type, message, related_uid, target_url=None: (
-            notifications.append((user_uid, message))
-        ),
     )
     run_async(NewsService().run_once())
-    assert failing_client.call_count == 4
-    assert len(notifications) == 1
-    assert notifications[0][0] == admin_uid
-    assert "News API unreachable" in notifications[0][1]
-
-
-def test_run_once_handles_api_failure_zero_retries(local_db, monkeypatch):
-    monkeypatch.setattr(news_mod, "get_setting", _settings_stub(retries="0"))
-    monkeypatch.setattr(base_mod, "get_setting", _settings_stub(retries="0"))
-    failing_client = FailingApiClient([])
-    monkeypatch.setattr(
-        news_mod.stealth, "stealth_async_client",
-        lambda *a, **k: failing_client,
-    )
-    run_async(NewsService().run_once())
-    assert failing_client.call_count == 1
 
 
 def test_run_once_updates_existing_news_row(local_db, monkeypatch):