fix(signal): detect RIFF/WAVE attachments as .wav so they route to STT

gateway/platforms/signal.py

@@ -97,6 +97,12 @@ def _guess_extension(data: bytes) -> str:
         return ".gif"
     if len(data) >= 12 and data[:4] == b"RIFF" and data[8:12] == b"WEBP":
         return ".webp"
+    # RIFF/WAVE shares the ``RIFF`` chunk header with WebP; the form-type at
+    # bytes 8-11 disambiguates. Without this, an inbound WAV voice note fell
+    # through to ``.bin`` (octet-stream) and was cached as a document, so STT
+    # never saw it — even though ``.wav`` is in the audio allowlist + MIME map.
+    if len(data) >= 12 and data[:4] == b"RIFF" and data[8:12] == b"WAVE":
+        return ".wav"
     if data[:4] == b"%PDF":
         return ".pdf"
     if len(data) >= 8 and data[4:8] == b"ftyp":

tests/gateway/test_signal.py

@@ -164,6 +164,43 @@ class TestSignalHelpers:
         from gateway.platforms.signal import _guess_extension
         assert _guess_extension(b"\x00\x00\x00\x18ftypisom" + b"\x00" * 100) == ".mp4"
 
+    def test_guess_extension_wav(self):
+        """RIFF/WAVE must be detected as ``.wav`` (regression for the WAV gap).
+
+        WAV shares the ``RIFF`` chunk header with WebP; only the form-type at
+        bytes 8-11 (``WAVE`` vs ``WEBP``) distinguishes them. Before the fix the
+        sniffer only handled ``WEBP`` and a WAV file fell through to ``.bin``.
+        """
+        from gateway.platforms.signal import _guess_extension
+        # Canonical WAV header: 'RIFF' <size> 'WAVE' 'fmt '...
+        wav = b"RIFF\x24\x08\x00\x00WAVEfmt " + b"\x00" * 100
+        assert _guess_extension(wav) == ".wav"
+
+    def test_guess_extension_wav_not_misread_as_webp(self):
+        """A RIFF container that is NOT WebP must not be returned as ``.webp``."""
+        from gateway.platforms.signal import _guess_extension
+        wav = b"RIFF\x24\x08\x00\x00WAVEfmt " + b"\x00" * 100
+        assert _guess_extension(wav) != ".webp"
+
+    def test_guess_extension_wav_routes_to_audio_cache(self):
+        """A detected WAV must route to the audio cache, not the document cache.
+
+        ``.wav`` is already in ``_is_audio_ext``; the bug was purely that
+        ``_guess_extension`` never produced ``.wav`` for raw bytes, so the
+        attachment was treated as a document and STT never received it.
+        """
+        from gateway.platforms.signal import _is_audio_ext, _guess_extension
+        wav = b"RIFF\x24\x08\x00\x00WAVEfmt " + b"\x00" * 100
+        ext = _guess_extension(wav)
+        assert ext == ".wav"
+        assert _is_audio_ext(ext) is True
+
+    def test_guess_extension_webp_still_detected(self):
+        """Guard that adding WAVE detection didn't break the WebP branch."""
+        from gateway.platforms.signal import _guess_extension
+        webp = b"RIFF\x24\x08\x00\x00WEBPVP8 " + b"\x00" * 100
+        assert _guess_extension(webp) == ".webp"
+
     def test_guess_extension_aac_adts_unprotected(self):
         """ADTS AAC, MPEG-4, no CRC (the canonical Android Signal voice note).