ws

.gitignore

@@ -8,3 +8,4 @@ dist/
 .env_private
 scripts/
 .cache/
+output.wav

cozepy/__init__.py

@@ -1,6 +1,6 @@
 from .audio.rooms import CreateRoomResp
 from .audio.speech import AudioFormat
-from .audio.translations import CreateTranslationResp
+from .audio.transcriptions import CreateTranslationResp
 from .audio.voices import Voice
 from .auth import (
     AsyncDeviceOAuthApp,
@@ -110,7 +110,7 @@
     # audio.voices
     "Voice",
     "AudioFormat",
-    # audio.translations
+    # audio.transcriptions
     "CreateTranslationResp",
     # auth
     "AsyncDeviceOAuthApp",

cozepy/audio/__init__.py

@@ -7,7 +7,7 @@
 if TYPE_CHECKING:
     from .rooms import AsyncRoomsClient, RoomsClient
     from .speech import AsyncSpeechClient, SpeechClient
-    from .translations import AsyncTranslationsClient, TranslationsClient
+    from .transcriptions import AsyncTranscriptionsClient, TranscriptionsClient
     from .voices import AsyncVoicesClient, VoicesClient
 
 
@@ -20,7 +20,7 @@ def __init__(self, base_url: str, auth: Auth, requester: Requester):
         self._rooms: Optional[RoomsClient] = None
         self._voices: Optional[VoicesClient] = None
         self._speech: Optional[SpeechClient] = None
-        self._translations: Optional[TranslationsClient] = None
+        self._transcriptions: Optional[TranscriptionsClient] = None
 
     @property
     def rooms(self) -> "RoomsClient":
@@ -39,12 +39,14 @@ def speech(self) -> "SpeechClient":
         return self._speech
 
     @property
-    def translations(self) -> "TranslationsClient":
-        if self._translations is None:
-            from .translations import TranslationsClient
+    def transcriptions(self) -> "TranscriptionsClient":
+        if self._transcriptions is None:
+            from .transcriptions import TranscriptionsClient
 
-            self._translations = TranslationsClient(base_url=self._base_url, auth=self._auth, requester=self._requester)
-        return self._translations
+            self._transcriptions = TranscriptionsClient(
+                base_url=self._base_url, auth=self._auth, requester=self._requester
+            )
+        return self._transcriptions
 
     @property
     def voices(self) -> "VoicesClient":
@@ -64,7 +66,7 @@ def __init__(self, base_url: str, auth: Auth, requester: Requester):
         self._rooms: Optional[AsyncRoomsClient] = None
         self._voices: Optional[AsyncVoicesClient] = None
         self._speech: Optional[AsyncSpeechClient] = None
-        self._translations: Optional[AsyncTranslationsClient] = None
+        self._transcriptions: Optional[AsyncTranscriptionsClient] = None
 
     @property
     def rooms(self) -> "AsyncRoomsClient":
@@ -91,11 +93,11 @@ def voices(self) -> "AsyncVoicesClient":
         return self._voices
 
     @property
-    def translations(self) -> "AsyncTranslationsClient":
-        if self._translations is None:
-            from .translations import AsyncTranslationsClient
+    def transcriptions(self) -> "AsyncTranscriptionsClient":
+        if self._transcriptions is None:
+            from .transcriptions import AsyncTranscriptionsClient
 
-            self._translations = AsyncTranslationsClient(
+            self._transcriptions = AsyncTranscriptionsClient(
                 base_url=self._base_url, auth=self._auth, requester=self._requester
             )
-        return self._translations
+        return self._transcriptions

cozepy/audio/translations/__init__.py → cozepy/audio/transcriptions/__init__.py

@@ -12,7 +12,7 @@ class CreateTranslationResp(CozeModel):
     text: str
 
 
-class TranslationsClient(object):
+class TranscriptionsClient(object):
     def __init__(self, base_url: str, auth: Auth, requester: Requester):
         self._base_url = remove_url_trailing_slash(base_url)
         self._auth = auth
@@ -30,15 +30,15 @@ def create(
         :param file: The file to be translated.
         :return: create translation result
         """
-        url = f"{self._base_url}/v1/audio/translations"
+        url = f"{self._base_url}/v1/audio/transcriptions"
         headers: Optional[dict] = kwargs.get("headers")
         files = {"file": _try_fix_file(file)}
         return self._requester.request(
             "post", url, stream=False, cast=CreateTranslationResp, headers=headers, files=files
         )
 
 
-class AsyncTranslationsClient(object):
+class AsyncTranscriptionsClient(object):
     """
     Room service async client.
     """
@@ -60,7 +60,7 @@ async def create(
         :param file: The file to be translated.
         :return: create translation result
         """
-        url = f"{self._base_url}/v1/audio/translations"
+        url = f"{self._base_url}/v1/audio/transcriptions"
         files = {"file": _try_fix_file(file)}
         headers: Optional[dict] = kwargs.get("headers")
         return await self._requester.arequest(

cozepy/audio/voices/__init__.py

@@ -210,11 +210,7 @@ async def clone(
         return await self._requester.arequest("post", url, False, Voice, headers=headers, body=body, files=files)
 
     async def list(
-        self,
-        *,
-        filter_system_voice: bool = False,
-        page_num: int = 1,
-        page_size: int = 100,
+        self, *, filter_system_voice: bool = False, page_num: int = 1, page_size: int = 100, **kwargs
     ) -> AsyncNumberPaged[Voice]:
         """
         Get available voices, including system voices + user cloned voices
@@ -227,6 +223,7 @@ async def list(
         :return: list of Voice
         """
         url = f"{self._base_url}/v1/audio/voices"
+        headers: Optional[dict] = kwargs.get("headers")
 
         def request_maker(i_page_num: int, i_page_size: int) -> HTTPRequest:
             return self._requester.make_request(
@@ -237,6 +234,7 @@ def request_maker(i_page_num: int, i_page_size: int) -> HTTPRequest:
                     "page_num": i_page_num,
                     "page_size": i_page_size,
                 },
+                headers=headers,
                 cast=_PrivateListVoiceData,
                 is_async=False,
                 stream=False,

cozepy/chat/__init__.py

@@ -1,3 +1,4 @@
+import base64
 import json
 import time
 from enum import Enum
@@ -24,6 +25,8 @@ class MessageRole(str, Enum):
 
 
 class MessageType(str, Enum):
+    UNKNOWN = ""
+
     # User input content.
     # 用户输入内容。
     QUESTION = "question"
@@ -52,8 +55,6 @@ class MessageType(str, Enum):
     # 多 answer 场景下，服务端会返回一个 verbose 包，对应的 content 为 JSON 格式，content.msg_type =generate_answer_finish 代表全部 answer 回复完成。不支持在请求中作为入参。
     VERBOSE = "verbose"
 
-    UNKNOWN = ""
-
 
 class MessageContentType(str, Enum):
     # Text.
@@ -187,12 +188,19 @@ def build_assistant_answer(content: str, meta_data: Optional[Dict[str, str]] = N
             meta_data=meta_data,
         )
 
+    def get_audio(self) -> Optional[bytes]:
+        if self.content_type == MessageContentType.AUDIO:
+            return base64.b64decode(self.content)
+        return None
+
 
 class ChatStatus(str, Enum):
     """
     The running status of the session
     """
 
+    UNKNOWN = ""
+
     # The session has been created.
     CREATED = "created"
 
@@ -214,9 +222,9 @@ class ChatStatus(str, Enum):
 
 class ChatError(CozeModel):
     # The error code. An integer type. 0 indicates success, other values indicate failure.
-    code: int
+    code: int = 0
     # The error message. A string type.
-    msg: str
+    msg: str = ""
 
 
 class ChatRequiredActionType(str, Enum):
@@ -266,13 +274,13 @@ class ChatRequiredAction(CozeModel):
 class ChatUsage(CozeModel):
     # The total number of Tokens consumed in this chat, including the consumption for both the input
     # and output parts.
-    token_count: int
+    token_count: int = 0
 
     # The total number of Tokens consumed for the output part.
-    output_count: int
+    output_count: int = 0
 
     # The total number of Tokens consumed for the input part.
-    input_count: int
+    input_count: int = 0
 
 
 class Chat(CozeModel):
@@ -301,7 +309,7 @@ class Chat(CozeModel):
     # completed: The Bot has finished processing, and the session has ended.
     # failed: The session has failed.
     # requires_action: The session is interrupted and requires further processing.
-    status: ChatStatus
+    status: ChatStatus = ChatStatus.UNKNOWN
 
     # Details of the information needed for execution.
     required_action: Optional[ChatRequiredAction] = None

cozepy/coze.py

@@ -15,6 +15,7 @@
     from .files import AsyncFilesClient, FilesClient
     from .knowledge import AsyncKnowledgeClient, KnowledgeClient  # deprecated
     from .templates import AsyncTemplatesClient, TemplatesClient
+    from .websockets import AsyncWebsocketsClient
     from .workflows import AsyncWorkflowsClient, WorkflowsClient
     from .workspaces import AsyncWorkspacesClient, WorkspacesClient
 
@@ -151,6 +152,7 @@ def __init__(
         self._workspaces: Optional[AsyncWorkspacesClient] = None
         self._audio: Optional[AsyncAudioClient] = None
         self._templates: Optional[AsyncTemplatesClient] = None
+        self._websockets: Optional[AsyncWebsocketsClient] = None
 
     @property
     def bots(self) -> "AsyncBotsClient":
@@ -237,3 +239,11 @@ def templates(self) -> "AsyncTemplatesClient":
 
             self._templates = AsyncTemplatesClient(self._base_url, self._auth, self._requester)
         return self._templates
+
+    @property
+    def websockets(self) -> "AsyncWebsocketsClient":
+        if not self._websockets:
+            from .websockets import AsyncWebsocketsClient
+
+            self._websockets = AsyncWebsocketsClient(self._base_url, self._auth, self._requester)
+        return self._websockets

cozepy/util.py

@@ -44,6 +44,18 @@ def remove_url_trailing_slash(base_url: str) -> str:
     return base_url
 
 
+def http_base_url_to_ws(base_url: str) -> str:
+    if not base_url:
+        raise ValueError("base_url cannot be empty")
+    if not base_url.startswith("https://"):
+        raise ValueError("base_url must start with 'https://'")
+    base_url = base_url.replace("https://", "wss://")
+
+    if "api-" in base_url:
+        return base_url.replace("api-", "ws-")
+    return base_url.replace("api.", "ws.")
+
+
 def remove_none_values(d: dict) -> dict:
     return {k: v for k, v in d.items() if v is not None}
 
@@ -55,7 +67,7 @@ def write_pcm_to_wav_file(
     Save PCM binary data to WAV file
 
     :param pcm_data: PCM binary data (24kHz, 16-bit, 1 channel, little-endian)
-    :param output_filename: Output WAV filename
+    :param filepath: Output WAV filename
     """
 
     with wave.open(filepath, "wb") as wav_file:

cozepy/websockets/__init__.py

@@ -0,0 +1,29 @@
+from cozepy import Auth
+from cozepy.request import Requester
+from cozepy.util import http_base_url_to_ws, remove_url_trailing_slash
+
+from .audio import AsyncWebsocketsAudioClient
+from .chat import AsyncWebsocketsChatClient
+
+
+class AsyncWebsocketsClient(object):
+    def __init__(self, base_url: str, auth: Auth, requester: Requester):
+        self._base_url = http_base_url_to_ws(remove_url_trailing_slash(base_url))
+        self._auth = auth
+        self._requester = requester
+
+    @property
+    def audio(self) -> AsyncWebsocketsAudioClient:
+        return AsyncWebsocketsAudioClient(
+            base_url=self._base_url,
+            auth=self._auth,
+            requester=self._requester,
+        )
+
+    @property
+    def chat(self) -> AsyncWebsocketsChatClient:
+        return AsyncWebsocketsChatClient(
+            base_url=self._base_url,
+            auth=self._auth,
+            requester=self._requester,
+        )

cozepy/websockets/audio/__init__.py

@@ -0,0 +1,28 @@
+from cozepy.auth import Auth
+from cozepy.request import Requester
+
+from .speech import AsyncWebsocketsAudioSpeechClient
+from .transcriptions import AsyncWebsocketsAudioTranscriptionsClient
+
+
+class AsyncWebsocketsAudioClient(object):
+    def __init__(self, base_url: str, auth: Auth, requester: Requester):
+        self._base_url = base_url
+        self._auth = auth
+        self._requester = requester
+
+    @property
+    def transcriptions(self) -> "AsyncWebsocketsAudioTranscriptionsClient":
+        return AsyncWebsocketsAudioTranscriptionsClient(
+            base_url=self._base_url,
+            auth=self._auth,
+            requester=self._requester,
+        )
+
+    @property
+    def speech(self) -> "AsyncWebsocketsAudioSpeechClient":
+        return AsyncWebsocketsAudioSpeechClient(
+            base_url=self._base_url,
+            auth=self._auth,
+            requester=self._requester,
+        )