finish voice things, start voice-recv things

Author: DA-344

discord/opus.py

@@ -43,6 +43,7 @@
 
 if TYPE_CHECKING:
     T = TypeVar("T")
+    APPLICATION_CTL = Literal['audio', 'voip', 'lowdelay']
     BAND_CTL = Literal["narrow", "medium", "wide", "superwide", "full"]
     SIGNAL_CTL = Literal["auto", "voice", "music"]
 
@@ -61,6 +62,12 @@ class SignalCtl(TypedDict):
     music: int
 
 
+class ApplicationCtl(TypedDict):
+    audio: int
+    voip: int
+    lowdelay: int
+
+
 __all__ = (
     "Encoder",
     "Decoder",
@@ -74,6 +81,7 @@ class SignalCtl(TypedDict):
 c_int_ptr = ctypes.POINTER(ctypes.c_int)
 c_int16_ptr = ctypes.POINTER(ctypes.c_int16)
 c_float_ptr = ctypes.POINTER(ctypes.c_float)
+OPUS_SILENCE = b'\xf8\xff\xfe'
 
 _lib = None
 
@@ -96,9 +104,11 @@ class DecoderStruct(ctypes.Structure):
 BAD_ARG = -1
 
 # Encoder CTLs
-APPLICATION_AUDIO = 2049
-APPLICATION_VOIP = 2048
-APPLICATION_LOWDELAY = 2051
+application_ctl: ApplicationCtl = {
+    'audio': 2049,
+    'lowdelay': 2051,
+    'voip': 2048,
+}
 
 CTL_SET_BITRATE = 4002
 CTL_SET_BANDWIDTH = 4008
@@ -365,16 +375,37 @@ def get_opus_version() -> str:
 
 
 class Encoder(_OpusStruct):
-    def __init__(self, application: int = APPLICATION_AUDIO):
+    def __init__(
+        self,
+        *,
+        application: APPLICATION_CTL = 'audio',
+        bitrate: int = 128,
+        fec: bool = True,
+        expected_packet_loss: float = 0.15,
+        bandwidth: BAND_CTL = 'full',
+        signal_type: SIGNAL_TL = 'auto',
+    ) -> None:
+        if application not in application_ctl:
+            raise ValueError(
+                'invalid application ctl type provided'
+            )
+        if not 16 <= bitrate <= 512:
+            raise ValueError('bitrate must be between 16 and 512, both included')
+        if not 0 < expected_packet_loss <= 1:
+            raise ValueError(
+                'expected_packet_loss must be between 0 and 1, including 1',
+            )
+
         _OpusStruct.get_opus_version()
 
-        self.application: int = application
+        self.application: int = application_ctl[application]
         self._state: EncoderStruct = self._create_state()
-        self.set_bitrate(128)
-        self.set_fec(True)
-        self.set_expected_packet_loss_percent(0.15)
-        self.set_bandwidth("full")
-        self.set_signal_type("auto")
+
+        self.set_bitrate(bitrate)
+        self.set_fec(fec)
+        self.set_expected_packet_loss_percent(expected_packet_loss)
+        self.set_bandwidth(bandwidth)
+        self.set_signal_type(signal_type)
 
     def __del__(self) -> None:
         if hasattr(self, "_state"):

discord/sinks/core.py

@@ -37,7 +37,7 @@
 from .errors import SinkException
 
 if TYPE_CHECKING:
-    from ..voice_client import VoiceClient
+    from ..voice.client import VoiceClient
 
 __all__ = (
     "Filters",
@@ -103,9 +103,14 @@ class RawData:
     .. versionadded:: 2.0
     """
 
-    def __init__(self, data, client):
-        self.data = bytearray(data)
-        self.client = client
+    if TYPE_CHECKING:
+        sequence: int
+        timestamp: int
+        ssrc: int
+
+    def __init__(self, data: bytes, client: VoiceClient):
+        self.data: bytearray = bytearray(data)
+        self.client: VoiceClient = client
 
         unpacker = struct.Struct(">xxHII")
         self.sequence, self.timestamp, self.ssrc = unpacker.unpack_from(self.data[:12])
@@ -120,16 +125,16 @@ def __init__(self, data, client):
         else:
             cutoff = 12
 
-        self.header = data[:cutoff]
+        self.header: bytes = data[:cutoff]
         self.data = self.data[cutoff:]
 
-        self.decrypted_data = getattr(self.client, f"_decrypt_{self.client.mode}")(
+        self.decrypted_data: bytes = getattr(self.client, f"_decrypt_{self.client.mode}")(
             self.header, self.data
         )
-        self.decoded_data = None
+        self.decoded_data: bytes | None = None
 
-        self.user_id = None
-        self.receive_time = time.perf_counter()
+        self.user_id: int | None = None
+        self.receive_time: float = time.perf_counter()
 
 
 class AudioData:

discord/state.py

@@ -1853,7 +1853,8 @@ def parse_voice_state_update(self, data) -> None:
             if int(data["user_id"]) == self_id:
                 voice = self._get_voice_client(guild.id)
                 if voice is not None:
-                    coro = voice.on_voice_state_update(data)
+                    payload = RawVoiceStateUpdateEvent(data=data, state=self)
+                    coro = voice.on_voice_state_update(payload)
                     asyncio.create_task(
                         logging_coroutine(
                             coro, info="Voice Protocol voice state update handler"
@@ -1890,8 +1891,9 @@ def parse_voice_server_update(self, data) -> None:
             key_id = int(data["channel_id"])
 
         vc = self._get_voice_client(key_id)
+        payload = RawVoiceServerUpdateEvent(data=data, state=self)
         if vc is not None:
-            coro = vc.on_voice_server_update(data)
+            coro = vc.on_voice_server_update(payload)
             asyncio.create_task(
                 logging_coroutine(
                     coro, info="Voice Protocol voice server update handler"