Add MiniMax voice cloning operations

apps/models_provider/impl/minimax_model_provider/model/tts.py

@@ -43,6 +43,81 @@ def new_instance(model_type, model_name, model_credential: Dict[str, object], **
     def check_auth(self):
         self.text_to_speech(_('Hello'))
 
+    def _post_voice_operation(self, path, **request_kwargs):
+        api_base = self.api_base.rstrip('/')
+        headers = {'Authorization': f'Bearer {self.api_key}'}
+        if 'json' in request_kwargs:
+            headers['Content-Type'] = 'application/json'
+
+        response = requests.post(
+            f'{api_base}/{path.lstrip("/")}',
+            headers=headers,
+            timeout=60,
+            **request_kwargs,
+        )
+        response.raise_for_status()
+
+        result = response.json()
+        if result.get('base_resp', {}).get('status_code', 0) != 0:
+            error_msg = result.get('base_resp', {}).get('status_msg', 'Unknown error')
+            raise Exception(f'MiniMax voice API error: {error_msg}')
+        return result
+
+    def _upload_voice_audio(self, audio_file, purpose):
+        if purpose not in {'voice_clone', 'prompt_audio'}:
+            raise ValueError(f'Unsupported MiniMax voice upload purpose: {purpose}')
+
+        if hasattr(audio_file, 'read'):
+            result = self._post_voice_operation(
+                'files/upload',
+                data={'purpose': purpose},
+                files={'file': audio_file},
+            )
+        else:
+            with open(audio_file, 'rb') as file_handle:
+                result = self._post_voice_operation(
+                    'files/upload',
+                    data={'purpose': purpose},
+                    files={'file': file_handle},
+                )
+
+        file_id = result.get('file', {}).get('file_id') or result.get('file_id')
+        if not file_id:
+            raise Exception('MiniMax voice upload returned no file ID')
+        return file_id
+
+    def upload_clone_audio(self, audio_file):
+        return self._upload_voice_audio(audio_file, 'voice_clone')
+
+    def upload_prompt_audio(self, audio_file):
+        return self._upload_voice_audio(audio_file, 'prompt_audio')
+
+    def voice_clone(self, file_id, voice_id, model=None, **params):
+        result = self._post_voice_operation(
+            'voice_clone',
+            json={
+                'file_id': file_id,
+                'voice_id': voice_id,
+                'model': model or self.model,
+                **params,
+            },
+        )
+        return result.get('voice_id') or voice_id
+
+    def voice_design(self, prompt, voice_id, **params):
+        result = self._post_voice_operation(
+            'voice_design',
+            json={
+                'prompt': prompt,
+                'voice_id': voice_id,
+                **params,
+            },
+        )
+        designed_voice_id = result.get('voice_id')
+        if not designed_voice_id:
+            raise Exception('MiniMax voice design returned no voice ID')
+        return designed_voice_id
+
     def text_to_speech(self, text):
         text = _remove_empty_lines(text)
         api_base = self.api_base.rstrip('/')