sarvamai 0.1.5a12__tar.gz → 0.1.5a13__tar.gz

{sarvamai-0.1.5a12 → sarvamai-0.1.5a13}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sarvamai
-Version: 0.1.5a12
+Version: 0.1.5a13
 Summary: 
 Requires-Python: >=3.8,<4.0
 Classifier: Intended Audience :: Developers

{sarvamai-0.1.5a12 → sarvamai-0.1.5a13}/pyproject.toml

@@ -3,7 +3,7 @@ name = "sarvamai"
 
 [tool.poetry]
 name = "sarvamai"
-version = "0.1.5a12"
+version = "0.1.5a13"
 description = ""
 readme = "README.md"
 authors = []

{sarvamai-0.1.5a12 → sarvamai-0.1.5a13}/src/sarvamai/core/client_wrapper.py

@@ -17,10 +17,10 @@ class BaseClientWrapper:
 
     def get_headers(self) -> typing.Dict[str, str]:
         headers: typing.Dict[str, str] = {
-            "User-Agent": "sarvamai/0.1.5a12",
+            "User-Agent": "sarvamai/0.1.5a13",
             "X-Fern-Language": "Python",
             "X-Fern-SDK-Name": "sarvamai",
-            "X-Fern-SDK-Version": "0.1.5a12",
+            "X-Fern-SDK-Version": "0.1.5a13",
         }
         headers["api-subscription-key"] = self.api_subscription_key
         return headers

{sarvamai-0.1.5a12 → sarvamai-0.1.5a13}/src/sarvamai/speech_to_text_streaming/socket_client.py

@@ -7,6 +7,7 @@ import websockets
 import websockets.sync.connection as websockets_sync_connection
 from ..core.events import EventEmitterMixin, EventType
 from ..core.pydantic_utilities import parse_obj_as
+from ..types.audio_data import AudioData
 from ..types.audio_message import AudioMessage
 from ..types.speech_to_text_streaming_response import SpeechToTextStreamingResponse
 
@@ -20,6 +21,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
 
     async def __aiter__(self):
         async for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextStreamingSocketClientResponse, message)  # type: ignore
 
     async def start_listening(self):
@@ -35,6 +37,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             async for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -42,7 +45,19 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+    async def transcribe(self, audio: str, encoding="audio/wav", sample_rate=16000):
+        """
+        Sends transcription request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding: Audio encoding format (default is "audio/wav")
+        :param sample_rate: Audio sample rate in Hz (default is 16000)
+        """
+
+        return await self._send_speech_to_text_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+        )
+
+    async def _send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
         """
         Send a message to the websocket connection.
         The message will be sent as a AudioMessage.
@@ -54,6 +69,7 @@ class AsyncSpeechToTextStreamingSocketClient(EventEmitterMixin):
         Receive a message from the websocket connection.
         """
         data = await self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextStreamingSocketClientResponse, data)  # type: ignore
 
     async def _send(self, data: typing.Any) -> None:
@@ -78,6 +94,7 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
 
     def __iter__(self):
         for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextStreamingSocketClientResponse, message)  # type: ignore
 
     def start_listening(self):
@@ -93,6 +110,7 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -100,20 +118,32 @@ class SpeechToTextStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+    def transcribe(self, audio: str, encoding="audio/wav", sample_rate=16000) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends transcription request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding (Optional): Audio encoding format (default is "audio/wav")
+        :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
-        self._send_model(message)
+        return self._send_speech_to_text_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, sample_rate=sample_rate, encoding=encoding))
+        )
 
     def recv(self) -> SpeechToTextStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextStreamingSocketClientResponse, data)  # type: ignore
 
+    def _send_speech_to_text_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        self._send_model(message)
+
     def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.

{sarvamai-0.1.5a12 → sarvamai-0.1.5a13}/src/sarvamai/speech_to_text_translate_streaming/socket_client.py

@@ -7,6 +7,7 @@ import websockets
 import websockets.sync.connection as websockets_sync_connection
 from ..core.events import EventEmitterMixin, EventType
 from ..core.pydantic_utilities import parse_obj_as
+from ..types.audio_data import AudioData
 from ..types.audio_message import AudioMessage
 from ..types.config_message import ConfigMessage
 from ..types.speech_to_text_translate_streaming_response import SpeechToTextTranslateStreamingResponse
@@ -21,6 +22,7 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
 
     async def __aiter__(self):
         async for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, message)  # type: ignore
 
     async def start_listening(self):
@@ -36,6 +38,7 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             async for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -43,27 +46,48 @@ class AsyncSpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    async def send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    async def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends audio translation request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding (Optional): Audio encoding format (default is "audio/wav")
+        :param sample_rate (Optional): Audio sample rate in Hz (default is 16000)
         """
-        await self._send_model(message)
+        return await self._send_speech_to_text_translate_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+        )
 
-    async def send_config_message(self, message: ConfigMessage) -> None:
+    async def set_prompt(self, prompt: str) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a ConfigMessage.
+        Set the prompt for the translation.
+        The prompt is useful for detecting hotwords in the audio stream.
+        This will send a ConfigMessage to the websocket connection.
         """
-        await self._send_model(message)
+        message = ConfigMessage(prompt=prompt)
+        await self._send_config_message(message)
 
     async def recv(self) -> SpeechToTextTranslateStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = await self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, data)  # type: ignore
 
+    async def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        await self._send_model(message)
+
+    async def _send_config_message(self, message: ConfigMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a ConfigMessage.
+        """
+        await self._send_model(message)
+
     async def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.
@@ -86,6 +110,7 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
 
     def __iter__(self):
         for message in self._websocket:
+            message = json.loads(message) if isinstance(message, str) else message
             yield parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, message)  # type: ignore
 
     def start_listening(self):
@@ -101,6 +126,7 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         self._emit(EventType.OPEN, None)
         try:
             for raw_message in self._websocket:
+                raw_message = json.loads(raw_message) if isinstance(raw_message, str) else raw_message
                 parsed = parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, raw_message)  # type: ignore
                 self._emit(EventType.MESSAGE, parsed)
         except websockets.WebSocketException as exc:
@@ -108,27 +134,48 @@ class SpeechToTextTranslateStreamingSocketClient(EventEmitterMixin):
         finally:
             self._emit(EventType.CLOSE, None)
 
-    def send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+    def translate(self, audio: str, encoding: str = "audio/wav", sample_rate: int = 16000):
         """
-        Send a message to the websocket connection.
-        The message will be sent as a AudioMessage.
+        Sends audio translation request to the server.
+        :param audio: Base64 encoded audio data
+        :param encoding: Audio encoding format (default is "audio/wav")
+        :param sample_rate: Audio sample rate in Hz (default is 16000)
         """
-        self._send_model(message)
+        return self._send_speech_to_text_translate_streaming_audio_message(
+            message=AudioMessage(audio=AudioData(data=audio, encoding=encoding, sample_rate=sample_rate))
+        )
 
-    def send_config_message(self, message: ConfigMessage) -> None:
+    def set_prompt(self, prompt: str) -> None:
         """
-        Send a message to the websocket connection.
-        The message will be sent as a ConfigMessage.
+        Set the prompt for the translation.
+        The prompt is useful for detecting hotwords in the audio stream.
+        This will send a ConfigMessage to the websocket connection.
         """
-        self._send_model(message)
+        message = ConfigMessage(prompt=prompt)
+        self._send_config_message(message)
 
     def recv(self) -> SpeechToTextTranslateStreamingSocketClientResponse:
         """
         Receive a message from the websocket connection.
         """
         data = self._websocket.recv()
+        data = json.loads(data) if isinstance(data, str) else data
         return parse_obj_as(SpeechToTextTranslateStreamingSocketClientResponse, data)  # type: ignore
 
+    def _send_config_message(self, message: ConfigMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a ConfigMessage.
+        """
+        self._send_model(message)
+
+    def _send_speech_to_text_translate_streaming_audio_message(self, message: AudioMessage) -> None:
+        """
+        Send a message to the websocket connection.
+        The message will be sent as a AudioMessage.
+        """
+        self._send_model(message)
+
     def _send(self, data: typing.Any) -> None:
         """
         Send a message to the websocket connection.