Merge pull request #49680 from NousResearch/fix/signal-quote-cache-eviction

fix(signal): FIFO-evict the quote-detection timestamp cache (follow-up to #49678)

gateway/platforms/signal.py

@@ -318,7 +318,10 @@ class SignalAdapter(BasePlatformAdapter):
         # Signal quote.id is the timestamp of the quoted message, so this lets
         # inbound replies identify that the user replied to a message sent by
         # this bot even after the self-sync echo was filtered above.
-        self._sent_message_timestamps: set[str] = set()
+        # OrderedDict (not set) so the cap evicts the OLDEST timestamp in FIFO
+        # order — a plain set.pop() removes an arbitrary element, which could
+        # drop a still-recent timestamp and miss a genuine reply-to-own-message.
+        self._sent_message_timestamps: "OrderedDict[str, None]" = OrderedDict()
         self._max_sent_message_timestamps = 500
         # Signal increasingly exposes ACI/PNI UUIDs as stable recipient IDs.
         # Keep a best-effort mapping so outbound sends can upgrade from a
@@ -807,9 +810,14 @@ class SignalAdapter(BasePlatformAdapter):
         """Keep a bounded cache of outbound Signal timestamps for quote matching."""
         if timestamp is None:
             return
-        self._sent_message_timestamps.add(str(timestamp))
-        if len(self._sent_message_timestamps) > self._max_sent_message_timestamps:
-            self._sent_message_timestamps.pop()
+        key = str(timestamp)
+        # Re-insert to mark most-recently-used so eviction drops genuinely old
+        # timestamps, not a recently re-seen one.
+        self._sent_message_timestamps.pop(key, None)
+        self._sent_message_timestamps[key] = None
+        # FIFO-evict the oldest entry once over the cap.
+        while len(self._sent_message_timestamps) > self._max_sent_message_timestamps:
+            self._sent_message_timestamps.popitem(last=False)
 
     def _extract_contact_uuid(self, contact: Any, phone_number: str) -> Optional[str]:
         """Best-effort extraction of a Signal service ID from listContacts output."""

tests/gateway/test_signal.py

@@ -1580,6 +1580,24 @@ class TestSignalQuoteExtraction:
         assert "111222333" in adapter._sent_message_timestamps
         assert adapter._quote_references_own_message("111222333", None) is True
 
+    def test_sent_message_timestamps_evicts_oldest_first(self, monkeypatch):
+        """Over the cap, the OLDEST quote-cache timestamp is dropped (FIFO),
+        not an arbitrary one — so a recent reply-to-own-message is still
+        detected after a burst of sends."""
+        adapter = _make_signal_adapter(monkeypatch)
+        adapter._max_sent_message_timestamps = 3
+        for ts in (1, 2, 3):
+            adapter._remember_sent_message_timestamp(ts)
+        # Adding a 4th evicts the oldest (1), keeps the rest in order.
+        adapter._remember_sent_message_timestamp(4)
+        assert list(adapter._sent_message_timestamps.keys()) == ["2", "3", "4"]
+        assert "1" not in adapter._sent_message_timestamps
+        # Re-seeing an existing ts promotes it so it survives the next eviction.
+        adapter._remember_sent_message_timestamp(2)  # 2 -> most recent
+        adapter._remember_sent_message_timestamp(5)  # evicts oldest (now 3)
+        assert list(adapter._sent_message_timestamps.keys()) == ["4", "2", "5"]
+        assert "3" not in adapter._sent_message_timestamps
+
     @pytest.mark.asyncio
     async def test_handle_envelope_without_quote_leaves_reply_fields_none(self, monkeypatch):
         adapter = _make_signal_adapter(monkeypatch)