🎨 refactor(generate_voice): update speaker attributes and API usage

- Renamed 'sid' to 'voice' and updated default seed in `Speaker`
- Updated `VoiceSpeakerV1` enum to use 'voice' attribute
- Changed `voice_engine` to `speaker` in `build` method
- Adjusted API request method from `get` to `post`
- Added support for additional audio content types in response handling

🔖 chore: bump version to 0.7.3 in pyproject.toml

playground/generate_voice.py

@@ -23,7 +23,7 @@ async def generate_voice(text: str):
     try:
         voice_gen = VoiceGenerate.build(
             text=text,
-            voice_engine=VoiceSpeakerV1.Crina,  # VoiceSpeakerV2.Ligeia,
+            speaker=VoiceSpeakerV2.Ligeia,  # VoiceSpeakerV2.Ligeia,
         )
         result = await voice_gen.request(
             session=credential

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "novelai-python"
-version = "0.7.2"
+version = "0.7.3"
 description = "NovelAI Python Binding With Pydantic"
 authors = [
     { name = "sudoskys", email = "[email protected]" },

src/novelai_python/sdk/ai/generate_voice/__init__.py

@@ -1,5 +1,6 @@
 # -*- coding: utf-8 -*-
 # @Author  : sudoskys
+from enum import Enum
 from typing import Optional, Union, Literal
 from urllib.parse import urlparse
 
@@ -11,7 +12,7 @@
 from pydantic import ConfigDict, PrivateAttr, Field, model_validator
 from tenacity import wait_random, retry, stop_after_attempt, retry_if_exception
 
-from ._enum import VoiceSpeakerV1, VoiceSpeakerV2
+from ._enum import VoiceSpeakerV1, VoiceSpeakerV2, Speaker
 from ...schema import ApiBaseModel
 from ...._exceptions import APIError, SessionHttpError
 from ...._response.ai.generate_voice import VoiceResponse
@@ -80,43 +81,36 @@ async def necessary_headers(self, request_data) -> dict:
     @classmethod
     def build(cls,
               text: str,
-              voice_engine: Union[VoiceSpeakerV1, VoiceSpeakerV2, str],
+              speaker: Union[VoiceSpeakerV2, VoiceSpeakerV1, Speaker, str],
               *,
-              opus: bool = False
+              opus: bool = True
               ) -> "VoiceGenerate":
         """
         生成图片
         :param opus: unknown
         :param text: str
-        :param voice_engine: VoiceSpeakerV1 or VoiceSpeakerV2 or str
+        :param speaker: Speaker import from novelai_python.sdk.ai.generate_voice._enum
         :return: VoiceGenerate instance
         :raises: ValueError
         """
-        if isinstance(voice_engine, str):
+        if isinstance(speaker, Enum):
+            speaker = speaker.value
+
+        if isinstance(speaker, str):
             return cls(
                 text=text,
                 voice=-1,
-                seed=voice_engine,
-                opus=opus,
-                version="v2"
-            )
-        if isinstance(voice_engine, VoiceSpeakerV2):
-            return cls(
-                text=text,
-                seed=voice_engine.value.seed,
-                voice=voice_engine.value.sid,
+                seed=speaker,
                 opus=opus,
                 version="v2"
             )
-        if isinstance(voice_engine, VoiceSpeakerV1):
-            return cls(
-                text=text,
-                seed=voice_engine.value.seed,
-                voice=voice_engine.value.sid,
-                opus=opus,
-                version="v1"
-            )
-        raise ValueError("Invalid voice engine")
+        return cls(
+            text=text,
+            seed=speaker.seed,
+            voice=speaker.voice,
+            opus=opus,
+            version=speaker.version
+        )
 
     @retry(
         wait=wait_random(min=1, max=3),
@@ -147,13 +141,13 @@ async def request(self,
             logger.debug(f"Voice request data: {request_data}")
             # Request
             try:
-                assert hasattr(sess, "get"), "session must have get method."
-                response = await sess.get(
-                    self.base_url,
-                    params=request_data
+                assert hasattr(sess, "post"), "session must have get method."
+                response = await sess.post(
+                    url=self.base_url,
+                    json=request_data
                 )
                 header_type = response.headers.get('Content-Type')
-                if header_type not in ['audio/mpeg', 'audio/ogg', 'audio/opus']:
+                if header_type not in ['audio/mpeg', 'audio/ogg', 'audio/opus', 'audio/wav', 'audio/webm']:
                     logger.warning(
                         f"Error with content type: {header_type} and code: {response.status_code}"
                     )

src/novelai_python/sdk/ai/generate_voice/_enum.py

@@ -1,37 +1,36 @@
-from typing import Optional
+from enum import Enum
 
 from pydantic import BaseModel
-from enum import Enum
 
 
 class Speaker(BaseModel):
     """
     Speaker for /ai/generated_voice
     """
-    sid: int = -1
-    seed: Optional[str] = None
+    voice: int = -1
+    seed: str = "kurumuz12"
     name: str
     category: str
 
     @property
     def version(self):
-        return "v2" if self.sid is None else "v1"
+        return "v2" if self.voice == -1 else "v1"
 
 
 class VoiceSpeakerV1(Enum):
     """
     Speaker for /ai/generated_voice
     """
-    Cyllene = Speaker(sid=17, name="Cyllene", category="female")
-    Leucosia = Speaker(sid=95, name="Leucosia", category="female")
-    Crina = Speaker(sid=44, name="Crina", category="female")
-    Hespe = Speaker(sid=80, name="Hespe", category="female")
-    Ida = Speaker(sid=106, name="Ida", category="female")
-    Alseid = Speaker(sid=6, name="Alseid", category="male")
-    Daphnis = Speaker(sid=10, name="Daphnis", category="male")
-    Echo = Speaker(sid=16, name="Echo", category="male")
-    Thel = Speaker(sid=41, name="Thel", category="male")
-    Nomios = Speaker(sid=77, name="Nomios", category="male")
+    Cyllene = Speaker(voice=17, name="Cyllene", category="female")
+    Leucosia = Speaker(voice=95, name="Leucosia", category="female")
+    Crina = Speaker(voice=44, name="Crina", category="female")
+    Hespe = Speaker(voice=80, name="Hespe", category="female")
+    Ida = Speaker(voice=106, name="Ida", category="female")
+    Alseid = Speaker(voice=6, name="Alseid", category="male")
+    Daphnis = Speaker(voice=10, name="Daphnis", category="male")
+    Echo = Speaker(voice=16, name="Echo", category="male")
+    Thel = Speaker(voice=41, name="Thel", category="male")
+    Nomios = Speaker(voice=77, name="Nomios", category="male")
     # SeedInput = Speaker(sid=-1, name="Seed Input", category="custom")