dashscope 1.8.0__py3-none-any.whl → 1.25.6__py3-none-any.whl

dashscope/audio/asr/transcription.py

@@ -1,3 +1,5 @@
+# Copyright (c) Alibaba, Inc. and its affiliates.
+
 import asyncio
 import time
 from typing import List, Union
@@ -24,23 +26,54 @@ class Transcription(BaseAsyncApi):
         paraformer_mtl_v1 = 'paraformer-mtl-v1'
 
     @classmethod
-    def call(cls, model: str, file_urls: List[str],
+    def call(cls,
+             model: str,
+             file_urls: List[str],
+             phrase_id: str = None,
+             api_key: str = None,
+             workspace: str = None,
              **kwargs) -> TranscriptionResponse:
         """Transcribe the given files synchronously.
 
         Args:
             model (str): The requested model_id.
             file_urls (List[str]): List of stored URLs.
-            channel_id (List[int], optional): The selected channel_id of audio file. # noqa: E501
+            phrase_id (str, `optional`): The ID of phrase.
+            workspace (str): The dashscope workspace id.
+
+            **kwargs:
+                channel_id (List[int], optional):
+                    The selected channel_id of audio file.
+                disfluency_removal_enabled(bool, `optional`):
+                    Filter mood words, turned off by default.
+                diarization_enabled (bool, `optional`):
+                    Speech auto diarization, turned off by default.
+                speaker_count (int, `optional`): The number of speakers.
+                timestamp_alignment_enabled (bool, `optional`):
+                    Timestamp-alignment calibration, turned off by default.
+                special_word_filter(str, `optional`): Sensitive word filter.
+                audio_event_detection_enabled(bool, `optional`):
+                    Audio event detection, turned off by default.
 
         Returns:
             TranscriptionResponse: The result of batch transcription.
         """
-        response = super().call(model, file_urls, **kwargs)
+        kwargs.update(cls._fill_resource_id(phrase_id, **kwargs))
+        kwargs = cls._tidy_kwargs(**kwargs)
+        response = super().call(model,
+                                file_urls,
+                                api_key=api_key,
+                                workspace=workspace,
+                                **kwargs)
         return TranscriptionResponse.from_api_response(response)
 
     @classmethod
-    def async_call(cls, model: str, file_urls: List[str],
+    def async_call(cls,
+                   model: str,
+                   file_urls: List[str],
+                   phrase_id: str = None,
+                   api_key: str = None,
+                   workspace: str = None,
                    **kwargs) -> TranscriptionResponse:
         """Transcribe the given files asynchronously,
         return the status of task submission for querying results subsequently.
@@ -48,23 +81,47 @@ class Transcription(BaseAsyncApi):
         Args:
             model (str): The requested model, such as paraformer-16k-1
             file_urls (List[str]): List of stored URLs.
-            channel_id (List[int], optional): The selected channel_id of audio file. # noqa: E501
+            phrase_id (str, `optional`): The ID of phrase.
+            workspace (str): The dashscope workspace id.
+
+        **kwargs:
+            channel_id (List[int], optional):
+                The selected channel_id of audio file.
+            disfluency_removal_enabled(bool, `optional`):
+                Filter mood words, turned off by default.
+            diarization_enabled (bool, `optional`):
+                Speech auto diarization, turned off by default.
+            speaker_count (int, `optional`): The number of speakers.
+            timestamp_alignment_enabled (bool, `optional`):
+                Timestamp-alignment calibration, turned off by default.
+            special_word_filter(str, `optional`): Sensitive word filter.
+            audio_event_detection_enabled(bool, `optional`):
+                Audio event detection, turned off by default.
 
         Returns:
             TranscriptionResponse: The response including task_id.
         """
-        response = cls._launch_request(model, file_urls, **kwargs)
+        kwargs.update(cls._fill_resource_id(phrase_id, **kwargs))
+        kwargs = cls._tidy_kwargs(**kwargs)
+        response = cls._launch_request(model,
+                                       file_urls,
+                                       api_key=api_key,
+                                       workspace=workspace,
+                                       **kwargs)
         return TranscriptionResponse.from_api_response(response)
 
     @classmethod
-    def fetch(
-            cls, task: Union[str,
-                             TranscriptionResponse]) -> TranscriptionResponse:
+    def fetch(cls,
+              task: Union[str, TranscriptionResponse],
+              api_key: str = None,
+              workspace: str = None,
+              **kwargs) -> TranscriptionResponse:
         """Fetch the status of task, including results of batch transcription when task_status is SUCCEEDED.  # noqa: E501
 
         Args:
             task (Union[str, TranscriptionResponse]): The task_id or
                 response including task_id returned from async_call().
+            workspace (str): The dashscope workspace id.
 
         Returns:
             TranscriptionResponse: The status of task_id,
@@ -73,7 +130,10 @@ class Transcription(BaseAsyncApi):
         try_count: int = 0
         while True:
             try:
-                response = super().fetch(task)
+                response = super().fetch(task,
+                                         api_key=api_key,
+                                         workspace=workspace,
+                                         **kwargs)
             except (asyncio.TimeoutError, aiohttp.ClientConnectorError) as e:
                 logger.error(e)
                 try_count += 1
@@ -87,28 +147,40 @@ class Transcription(BaseAsyncApi):
         return TranscriptionResponse.from_api_response(response)
 
     @classmethod
-    def wait(cls, task: Union[str,
-                              TranscriptionResponse]) -> TranscriptionResponse:
+    def wait(cls,
+             task: Union[str, TranscriptionResponse],
+             api_key: str = None,
+             workspace: str = None,
+             **kwargs) -> TranscriptionResponse:
         """Poll task until the final results of transcription is obtained.
 
         Args:
             task (Union[str, TranscriptionResponse]): The task_id or
                 response including task_id returned from async_call().
+            workspace (str): The dashscope workspace id.
 
         Returns:
             TranscriptionResponse: The result of batch transcription.
         """
-        response = super().wait(task)
+        response = super().wait(task,
+                                api_key=api_key,
+                                workspace=workspace,
+                                **kwargs)
         return TranscriptionResponse.from_api_response(response)
 
     @classmethod
-    def _launch_request(cls, model: str, files: List[str],
+    def _launch_request(cls,
+                        model: str,
+                        files: List[str],
+                        api_key: str = None,
+                        workspace: str = None,
                         **kwargs) -> DashScopeAPIResponse:
         """Submit transcribe request.
 
         Args:
             model (str): The requested model, such as paraformer-16k-1
             files (List[str]): List of stored URLs.
+            workspace (str): The dashscope workspace id.
 
         Returns:
             DashScopeAPIResponse: The result of task submission.
@@ -125,6 +197,8 @@ class Transcription(BaseAsyncApi):
                                               input={'file_urls': files},
                                               api_protocol=ApiProtocol.HTTP,
                                               http_method=HTTPMethod.POST,
+                                              api_key=api_key,
+                                              workspace=workspace,
                                               **kwargs)
             except (asyncio.TimeoutError, aiohttp.ClientConnectorError) as e:
                 logger.error(e)
@@ -136,3 +210,22 @@ class Transcription(BaseAsyncApi):
             break
 
         return response
+
+    @classmethod
+    def _fill_resource_id(cls, phrase_id: str, **kwargs):
+        resources_list: list = []
+        if phrase_id is not None and len(phrase_id) > 0:
+            item = {'resource_id': phrase_id, 'resource_type': 'asr_phrase'}
+            resources_list.append(item)
+
+            if len(resources_list) > 0:
+                kwargs['resources'] = resources_list
+
+        return kwargs
+
+    @classmethod
+    def _tidy_kwargs(cls, **kwargs):
+        for k in kwargs.copy():
+            if kwargs[k] is None:
+                kwargs.pop(k, None)
+        return kwargs