fix(media): recognize MP3 and M4A as voice-compatible audio (#15438)

* fix(media): recognize MP3 and M4A as voice-compatible audio

Telegram sendVoice supports OGG/Opus, MP3, and M4A, but
isVoiceCompatibleAudio only recognized OGG/Opus formats.

- Add MP3 and M4A extensions and MIME types
- Use explicit MIME set instead of substring matching
- Handle MIME parameters (e.g. 'audio/ogg; codecs=opus')
- Add test coverage for all supported and unsupported formats

* fix: narrow MIME allowlist per review feedback

Remove audio/mp4 and audio/aac from voice MIME types — too broad.
Keep only M4A-specific types (audio/x-m4a, audio/m4a).
Add audio/mp4 and audio/aac as negative test cases.

* fix: align voice compatibility and channel coverage (#15438) (thanks @azade-c)

---------

Co-authored-by: Peter Steinberger <steipete@gmail.com>

extensions/matrix/src/matrix/send.test.ts

@@ -24,6 +24,8 @@ const loadWebMediaMock = vi.fn().mockResolvedValue({
   contentType: "image/png",
   kind: "image",
 });
+const mediaKindFromMimeMock = vi.fn(() => "image");
+const isVoiceCompatibleAudioMock = vi.fn(() => false);
 const getImageMetadataMock = vi.fn().mockResolvedValue(null);
 const resizeToJpegMock = vi.fn();
 
@@ -33,8 +35,8 @@ const runtimeStub = {
   },
   media: {
     loadWebMedia: (...args: unknown[]) => loadWebMediaMock(...args),
-    mediaKindFromMime: () => "image",
-    isVoiceCompatibleAudio: () => false,
+    mediaKindFromMime: (...args: unknown[]) => mediaKindFromMimeMock(...args),
+    isVoiceCompatibleAudio: (...args: unknown[]) => isVoiceCompatibleAudioMock(...args),
     getImageMetadata: (...args: unknown[]) => getImageMetadataMock(...args),
     resizeToJpeg: (...args: unknown[]) => resizeToJpegMock(...args),
   },
@@ -71,6 +73,8 @@ describe("sendMessageMatrix media", () => {
 
   beforeEach(() => {
     vi.clearAllMocks();
+    mediaKindFromMimeMock.mockReturnValue("image");
+    isVoiceCompatibleAudioMock.mockReturnValue(false);
     setMatrixRuntime(runtimeStub);
   });
 
@@ -133,6 +137,66 @@ describe("sendMessageMatrix media", () => {
     expect(content.url).toBeUndefined();
     expect(content.file?.url).toBe("mxc://example/file");
   });
+
+  it("marks voice metadata and sends caption follow-up when audioAsVoice is compatible", async () => {
+    const { client, sendMessage } = makeClient();
+    mediaKindFromMimeMock.mockReturnValue("audio");
+    isVoiceCompatibleAudioMock.mockReturnValue(true);
+    loadWebMediaMock.mockResolvedValueOnce({
+      buffer: Buffer.from("audio"),
+      fileName: "clip.mp3",
+      contentType: "audio/mpeg",
+      kind: "audio",
+    });
+
+    await sendMessageMatrix("room:!room:example", "voice caption", {
+      client,
+      mediaUrl: "file:///tmp/clip.mp3",
+      audioAsVoice: true,
+    });
+
+    expect(isVoiceCompatibleAudioMock).toHaveBeenCalledWith({
+      contentType: "audio/mpeg",
+      fileName: "clip.mp3",
+    });
+    expect(sendMessage).toHaveBeenCalledTimes(2);
+    const mediaContent = sendMessage.mock.calls[0]?.[1] as {
+      msgtype?: string;
+      body?: string;
+      "org.matrix.msc3245.voice"?: Record<string, never>;
+    };
+    expect(mediaContent.msgtype).toBe("m.audio");
+    expect(mediaContent.body).toBe("Voice message");
+    expect(mediaContent["org.matrix.msc3245.voice"]).toEqual({});
+  });
+
+  it("keeps regular audio payload when audioAsVoice media is incompatible", async () => {
+    const { client, sendMessage } = makeClient();
+    mediaKindFromMimeMock.mockReturnValue("audio");
+    isVoiceCompatibleAudioMock.mockReturnValue(false);
+    loadWebMediaMock.mockResolvedValueOnce({
+      buffer: Buffer.from("audio"),
+      fileName: "clip.wav",
+      contentType: "audio/wav",
+      kind: "audio",
+    });
+
+    await sendMessageMatrix("room:!room:example", "voice caption", {
+      client,
+      mediaUrl: "file:///tmp/clip.wav",
+      audioAsVoice: true,
+    });
+
+    expect(sendMessage).toHaveBeenCalledTimes(1);
+    const mediaContent = sendMessage.mock.calls[0]?.[1] as {
+      msgtype?: string;
+      body?: string;
+      "org.matrix.msc3245.voice"?: Record<string, never>;
+    };
+    expect(mediaContent.msgtype).toBe("m.audio");
+    expect(mediaContent.body).toBe("voice caption");
+    expect(mediaContent["org.matrix.msc3245.voice"]).toBeUndefined();
+  });
 });
 
 describe("sendMessageMatrix threads", () => {