From a3df2e00dcc46c95350e0cb6a7174fabeb27e6a2 Mon Sep 17 00:00:00 2001
From: "2569718930@qq.com" <2569718930@qq.com>
Date: Sun, 31 May 2026 21:33:35 +0800
Subject: [PATCH] Clamp SSE replay window on backend

---
 tests/test_sse_replay.py  | 49 +++++++++++++++++++++++++++++++++++++--
 web/routers/sse_router.py | 17 ++++++++++----
 2 files changed, 60 insertions(+), 6 deletions(-)

diff --git a/tests/test_sse_replay.py b/tests/test_sse_replay.py
index d3e0d40d..bca3d1d0 100644
--- a/tests/test_sse_replay.py
+++ b/tests/test_sse_replay.py
@@ -129,8 +129,53 @@ def test_events_endpoint_emits_resync_when_replay_is_incomplete(monkeypatch):
 
 def test_replay_limit_is_bounded():
     assert sse_router._bounded_replay_limit(0) == 1
-    assert sse_router._bounded_replay_limit(500) == 500
-    assert sse_router._bounded_replay_limit(5000) == 2000
+    assert sse_router._bounded_replay_limit(500) == 60
+    assert sse_router._bounded_replay_limit(5000) == 60
+    assert sse_router._bounded_replay_limit(500, city_count=5) == 120
+    assert sse_router._bounded_replay_limit(500, city_count=20) == 240
+    assert sse_router._bounded_replay_limit(25, city_count=5) == 25
+
+
+def test_legacy_high_replay_limit_is_clamped_by_city_count(monkeypatch):
+    captured = {}
+
+    class FakeStore:
+        def latest_revision(self):
+            return 44
+
+        def replay_events(self, *, cities, since_revision, limit):
+            captured["cities"] = cities
+            captured["limit"] = limit
+            return []
+
+        def replay_requires_resync(self, *, cities, since_revision, replay_count, limit):
+            captured["resync_limit"] = limit
+            return False
+
+    async def finite_stream(
+        user_id,
+        *,
+        cities=None,
+        replay_events=None,
+        connected_revision=0,
+        resync_event=None,
+    ):
+        yield sse_router.sse_manager._format_event(
+            {"type": "connected", "revision": connected_revision}
+        )
+
+    monkeypatch.setattr(sse_router, "event_store", FakeStore())
+    monkeypatch.setattr(sse_router.sse_manager, "event_stream", finite_stream)
+
+    response = TestClient(app).get(
+        "/api/events?cities=ankara,buenos%20aires,istanbul,jeddah,seoul"
+        "&since_revision=42&replay_limit=500"
+    )
+
+    assert response.status_code == 200
+    assert captured["cities"] == {"ankara", "buenos aires", "istanbul", "jeddah", "seoul"}
+    assert captured["limit"] == 120
+    assert captured["resync_limit"] == 120
 
 
 def test_ingest_patch_uses_external_fanout_without_direct_broadcast(monkeypatch):
diff --git a/web/routers/sse_router.py b/web/routers/sse_router.py
index 3234a1e4..2001dadc 100644
--- a/web/routers/sse_router.py
+++ b/web/routers/sse_router.py
@@ -19,6 +19,9 @@ router = APIRouter(tags=["events"])
 event_store = create_realtime_event_store()
 _live_subscription_lock = threading.Lock()
 _live_subscription_started = False
+SSE_REPLAY_BASE_LIMIT = 60
+SSE_REPLAY_EVENTS_PER_CITY = 24
+SSE_REPLAY_MAX_LIMIT = 240
 
 
 def _parse_cities_param(cities: str) -> Set[str]:
@@ -29,12 +32,18 @@ def _parse_cities_param(cities: str) -> Set[str]:
     }
 
 
-def _bounded_replay_limit(value: int) -> int:
+def _recommended_replay_limit(city_count: int) -> int:
+    requested = max(1, int(city_count or 0)) * SSE_REPLAY_EVENTS_PER_CITY
+    return max(SSE_REPLAY_BASE_LIMIT, min(SSE_REPLAY_MAX_LIMIT, requested))
+
+
+def _bounded_replay_limit(value: int, *, city_count: int = 0) -> int:
     try:
         limit = int(value)
     except (TypeError, ValueError):
-        limit = 500
-    return max(1, min(MAX_REPLAY_LIMIT, limit))
+        limit = _recommended_replay_limit(city_count)
+    route_limit = _recommended_replay_limit(city_count)
+    return max(1, min(MAX_REPLAY_LIMIT, route_limit, limit))
 
 
 def _ensure_live_subscription() -> None:
@@ -65,7 +74,7 @@ async def sse_events(
     origin = request.headers.get("origin", "")
     allowed = origin in {"https://polyweather.top", "https://www.polyweather.top", "http://localhost:3000"}
     city_set = _parse_cities_param(cities)
-    limit = _bounded_replay_limit(replay_limit)
+    limit = _bounded_replay_limit(replay_limit, city_count=len(city_set))
     _ensure_live_subscription()
     latest_revision = event_store.latest_revision()
     replay_events = []