livekit-plugins-azure 0.3.0.dev7__tar.gz → 0.3.2__tar.gz

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: livekit-plugins-azure
-Version: 0.3.0.dev7
+Version: 0.3.2
 Summary: Agent Framework plugin for services from Azure
 Home-page: https://github.com/livekit/agents
 License: Apache-2.0

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/livekit/plugins/azure/__init__.py

@@ -18,13 +18,12 @@ __all__ = ["STT", "SpeechStream", "TTS", "__version__"]
 
 from livekit.agents import Plugin
 
+from .log import logger
+
 
 class AzurePlugin(Plugin):
     def __init__(self):
-        super().__init__(__name__, __version__, __package__)
-
-    def download_files(self):
-        pass
+        super().__init__(__name__, __version__, __package__, logger)
 
 
 Plugin.register_plugin(AzurePlugin())

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/livekit/plugins/azure/stt.py

@@ -16,6 +16,7 @@ import asyncio
 import os
 from dataclasses import dataclass
 
+from livekit import rtc
 from livekit.agents import stt, utils
 
 import azure.cognitiveservices.speech as speechsdk  # type: ignore
@@ -44,6 +45,13 @@ class STT(stt.STT):
         num_channels: int = 1,
         languages: list[str] = [],  # when empty, auto-detect the language
     ):
+        """
+        Create a new instance of Azure STT.
+
+        ``speech_key`` and ``speech_region`` must be set, either using arguments or by setting the
+        ``AZURE_SPEECH_KEY`` and ``AZURE_SPEECH_REGION`` environmental variables, respectively.
+        """
+
         super().__init__(
             capabilities=stt.STTCapabilities(streaming=True, interim_results=True)
         )
@@ -102,7 +110,8 @@ class SpeechStream(stt.SpeechStream):
     async def _main_task(self) -> None:
         try:
             async for input in self._input_ch:
-                self._stream.write(input.data.tobytes())
+                if isinstance(input, rtc.AudioFrame):
+                    self._stream.write(input.data.tobytes())
 
             self._stream.close()
             await self._done_event.wait()

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/livekit/plugins/azure/tts.py

@@ -16,7 +16,6 @@ import asyncio
 import os
 from dataclasses import dataclass
 
-from livekit import rtc
 from livekit.agents import tts, utils
 
 import azure.cognitiveservices.speech as speechsdk  # type: ignore
@@ -42,6 +41,13 @@ class TTS(tts.TTS):
         speech_region: str | None = None,
         voice: str | None = None,
     ) -> None:
+        """
+        Create a new instance of Azure TTS.
+
+        ``speech_key`` and ``speech_region`` must be set, either using arguments or by setting the
+        ``AZURE_SPEECH_KEY`` and ``AZURE_SPEECH_REGION`` environmental variables, respectively.
+        """
+
         super().__init__(
             capabilities=tts.TTSCapabilities(
                 streaming=False,
@@ -73,17 +79,18 @@ class ChunkedStream(tts.ChunkedStream):
 
     @utils.log_exceptions()
     async def _main_task(self):
-        stream_callback = _PushAudioOutputStreamCallback(
-            asyncio.get_running_loop(), self._event_ch
+        stream_callback = speechsdk.audio.PushAudioOutputStream(
+            _PushAudioOutputStreamCallback(asyncio.get_running_loop(), self._event_ch)
         )
         synthesizer = _create_speech_synthesizer(
             config=self._opts,
-            stream=speechsdk.audio.PushAudioOutputStream(stream_callback),
+            stream=stream_callback,
         )
 
         def _synthesize() -> speechsdk.SpeechSynthesisResult:
             return synthesizer.speak_text_async(self._text).get()  # type: ignore
 
+        result = None
         try:
             result = await asyncio.to_thread(_synthesize)
             if result.reason != speechsdk.ResultReason.SynthesizingAudioCompleted:
@@ -93,8 +100,11 @@ class ChunkedStream(tts.ChunkedStream):
         finally:
 
             def _cleanup() -> None:
-                nonlocal synthesizer, result
+                # cleanup resources inside an Executor
+                # to avoid blocking the event loop
+                nonlocal synthesizer, stream_callback, result
                 del synthesizer
+                del stream_callback
                 del result
 
             await asyncio.to_thread(_cleanup)
@@ -112,20 +122,30 @@ class _PushAudioOutputStreamCallback(speechsdk.audio.PushAudioOutputStreamCallba
         self._request_id = utils.shortuuid()
         self._segment_id = utils.shortuuid()
 
-    def write(self, audio_buffer: memoryview) -> int:
-        audio = tts.SynthesizedAudio(
-            request_id=self._request_id,
-            segment_id=self._segment_id,
-            frame=rtc.AudioFrame(
-                data=audio_buffer,
-                sample_rate=AZURE_SAMPLE_RATE,
-                num_channels=AZURE_NUM_CHANNELS,
-                samples_per_channel=audio_buffer.nbytes // 2,
-            ),
+        self._bstream = utils.audio.AudioByteStream(
+            sample_rate=AZURE_SAMPLE_RATE, num_channels=AZURE_NUM_CHANNELS
         )
-        self._loop.call_soon_threadsafe(self._event_ch.send_nowait, audio)
+
+    def write(self, audio_buffer: memoryview) -> int:
+        for frame in self._bstream.write(audio_buffer.tobytes()):
+            audio = tts.SynthesizedAudio(
+                request_id=self._request_id,
+                segment_id=self._segment_id,
+                frame=frame,
+            )
+            self._loop.call_soon_threadsafe(self._event_ch.send_nowait, audio)
+
         return audio_buffer.nbytes
 
+    def close(self) -> None:
+        for frame in self._bstream.flush():
+            audio = tts.SynthesizedAudio(
+                request_id=self._request_id,
+                segment_id=self._segment_id,
+                frame=frame,
+            )
+            self._loop.call_soon_threadsafe(self._event_ch.send_nowait, audio)
+
 
 def _create_speech_synthesizer(
     *, config: _TTSOptions, stream: speechsdk.audio.AudioOutputStream

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/livekit/plugins/azure/version.py

@@ -12,4 +12,4 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__version__ = "0.3.0-dev.7"
+__version__ = "0.3.2"

{livekit_plugins_azure-0.3.0.dev7 → livekit_plugins_azure-0.3.2}/livekit_plugins_azure.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: livekit-plugins-azure
-Version: 0.3.0.dev7
+Version: 0.3.2
 Summary: Agent Framework plugin for services from Azure
 Home-page: https://github.com/livekit/agents
 License: Apache-2.0