neverlib 0.2.4__py3-none-any.whl → 0.2.6__py3-none-any.whl

neverlib/utils/lazy_module.py

@@ -0,0 +1,81 @@
+'''
+Author: 凌逆战 | Never
+Date: 2025-09-04
+Description: 懒加载模块
+'''
+import sys
+import importlib
+from types import ModuleType
+from typing import Dict, Tuple, Union, Iterable
+
+class LazyModule(ModuleType):
+    """
+    一个懒加载模块，可以按需导入子模块或对象。
+    - 支持 name -> 'pkg.subpkg.module' 的懒导入(返回模块对象)
+    - 支持 name -> ('pkg.subpkg.module', 'attr') 的懒导入(返回模块内的对象)
+    - 透传 __spec__/__path__/__all__ 以兼容 import 系统和 IDE 补全
+    """
+    def __init__(self, name: str, import_structure: Dict[str, Union[str, Tuple[str, str]]], module_spec=None, module_path: Iterable[str] = None):
+        super().__init__(name)
+        # 提取代理模块配置: __all_from__ 可为 str 或 list[str]
+        proxy_modules = []
+        if import_structure and "__all_from__" in import_structure:
+            raw = import_structure.get("__all_from__")
+            if isinstance(raw, (list, tuple)):
+                proxy_modules = list(raw)
+            elif isinstance(raw, str):
+                proxy_modules = [raw]
+            # 不参与普通键映射
+            import_structure = {k: v for k, v in import_structure.items() if k != "__all_from__"}
+
+        self.__dict__["_import_structure"] = import_structure   # 导入结构
+        self.__dict__["_modules"] = {}                          # 模块字典
+        self.__dict__["_proxy_modules"] = proxy_modules         # 代理模块(其公开符号可直接在包下访问)
+        if module_spec is not None:
+            self.__dict__["__spec__"] = module_spec
+        if module_path is not None:
+            self.__dict__["__path__"] = list(module_path)
+        # 若存在代理模块，则不预设 __all__，避免阻断 import * 的 dir() 行为
+        if not proxy_modules:
+            self.__dict__["__all__"] = sorted(list(import_structure.keys()))
+
+    def __getattr__(self, name):
+        import_structure = self.__dict__["_import_structure"]   # 导入结构
+        if name in import_structure:
+            target = import_structure[name]
+            # 目标是模块名: 懒导入并返回模块
+            if isinstance(target, str):
+                module = importlib.import_module(target)
+                self.__dict__["_modules"][name] = module
+                setattr(self, name, module)
+                return module
+            # 目标是 (模块, 属性): 懒导入并返回对象
+            if isinstance(target, tuple) and len(target) == 2:
+                module_name, attr_name = target
+                module = importlib.import_module(module_name)
+                obj = getattr(module, attr_name)
+                setattr(self, name, obj)
+                return obj
+        # 若未在显式映射中，尝试到代理模块中查找公开符号
+        for proxy_name in self.__dict__.get("_proxy_modules", []):
+            mod_cache_key = f"__proxy__::{proxy_name}"
+            module = self.__dict__["_modules"].get(mod_cache_key)
+            if module is None:
+                module = importlib.import_module(proxy_name)
+                self.__dict__["_modules"][mod_cache_key] = module
+            if hasattr(module, name):
+                obj = getattr(module, name)
+                setattr(self, name, obj)
+                return obj
+        raise AttributeError(f"Module {self.__name__} has no attribute {name}")  # 如果模块没有属性，则抛出异常
+
+    def __dir__(self):
+        # 将懒加载键与现有属性合并，便于 IDE 补全
+        keys = set(self.__dict__.keys()) | set(self.__dict__.get("_import_structure", {}).keys())
+        # 若配置了代理模块，将其公开符号加入补全(会触发代理模块导入)
+        for proxy_name in self.__dict__.get("_proxy_modules", []):
+            module = importlib.import_module(proxy_name)
+            for n in dir(module):
+                if not n.startswith('_'):
+                    keys.add(n)
+        return sorted(keys)

neverlib/utils/message.py

@@ -1,14 +1,9 @@
 '''
 Author: 凌逆战 | Never
-Date: 2025-02-13 20:06:07
-Description: 请填写文件描述
+Date: 2025-02-13
+Description: 通过邮件发送通知
 '''
-# -*- coding:utf-8 -*-
-# Author:凌逆战 | Never
-# Date: 2024/2/1
-"""
-通过邮件发送通知
-"""
+
 import os
 import smtplib
 from email.header import Header

neverlib/utils/utils.py

@@ -28,6 +28,37 @@ def get_path_list(source_path, end="*.wav", shuffle=False):
     return wav_list
 
 
+def get_audio_segments(sentence_len, wav_path_list, sr=16000, insert_silence=None):
+    """
+    从音频列表中随机拼接指定长度音频
+    Args:
+        sentence_len: 需要返回的音频长度
+        wav_path_list: 音频路径列表
+        sr: 采样率
+        insert_silence: 插入静音的长度
+    Returns:返回指定长度的音频
+    """
+    merge_wav_len = 0
+    wav_list = []
+    while merge_wav_len < sentence_len:
+        wav_path = random.choice(wav_path_list)
+        wav, wav_sr = sf.read(wav_path, always_2d=True, dtype='float32')
+        assert wav_sr == sr, f"音频采样率是{wav_sr}, 期望{sr}"
+        merge_wav_len += len(wav)
+        wav_list.append(wav)
+        if insert_silence is not None:
+            silence_len = random.randint(0, insert_silence)
+            silence = np.zeros((silence_len, ), dtype=np.float32)
+            merge_wav_len += silence_len
+            wav_list.append(silence)
+    wav = np.concatenate(wav_list, axis=0)
+    if len(wav) > sentence_len:
+        # 随机截取clean_len
+        start = random.randint(0, merge_wav_len - sentence_len)
+        wav = wav[start:start + sentence_len, :]
+    return wav
+
+
 def rename_files_and_folders(directory, replace='_-', replacement='_'):
     # 将路径的指定字符替换为指定字符
     for root, dirs, files in os.walk(directory):
@@ -67,9 +98,7 @@ def TrainValSplit(dataset_dir, train_dir, val_dir, percentage=0.9):
     :param val_dir: 验证集地址
     :param percentage: 分割百分比
     """
-    wav_path_list = sorted(get_path_list(dataset_dir))
-    random.seed(10086)
-    random.shuffle(wav_path_list)  # 打乱列表的顺序
+    wav_path_list = get_path_list(dataset_dir, end="*.wav", shuffle=True)
     total_wav_num = len(wav_path_list)
     # 计算训练集和验证集的分割点
     split_idx = int(total_wav_num * percentage)

neverlib/vad/__init__.py

@@ -7,12 +7,19 @@ from neverlib.vad import EnergyVad_C
 如果没有用户必须完整路径
 from neverlib.vad.VAD_Energy import EnergyVad_C
 """
-from .PreProcess import *
-from .VAD_Energy import EnergyVad_C
-from .VAD_funasr import FunASR_VAD_C
-from .VAD_Silero import Silero_VAD_C
-from .VAD_statistics import Statistics_VAD
-from .VAD_vadlib import Vadlib_C
-from .VAD_WebRTC import WebRTC_VAD_C
-from .VAD_whisper import Whisper_VAD_C
-from .utils import from_vadArray_to_vadEndpoint, vad2nad
+from lazy_loader import attach
+
+__getattr__, __dir__, __all__ = attach(
+    __name__,
+    submodules=["PreProcess", "VAD_Energy", "VAD_funasr", "VAD_Silero", "VAD_statistics", "VAD_vadlib", "VAD_WebRTC", "VAD_whisper", "utils"],
+    submod_attrs={
+        "VAD_Energy": ["EnergyVad_C"],
+        "VAD_funasr": ["FunASR_VAD_C"],
+        "VAD_Silero": ["Silero_VAD_C"],
+        "VAD_statistics": ["Statistics_VAD"],
+        "VAD_vadlib": ["Vadlib_C"],
+        "VAD_WebRTC": ["WebRTC_VAD_C"],
+        "VAD_whisper": ["Whisper_VAD_C"],
+        "utils": ["from_vadArray_to_vadEndpoint", "vad2nad"],
+    },
+)

neverlib/vad/utils.py

@@ -1,9 +1,8 @@
-# -*- coding:utf-8 -*-
-# Author:凌逆战 | Never
-# Date: 2024/9/19
-"""
-
-"""
+'''
+Author: 凌逆战 | Never
+Date: 2025-08-22
+Description: 
+'''
 import numpy as np
 
 
@@ -65,3 +64,18 @@ def vad2nad(vad, total_length):
         nad.append({'start': last_end, 'end': total_length})
 
     return nad
+
+
+def vad_smooth(vad, sr=16000, min_speech_duration=0.2, min_silent_duration=0.2):
+    # 把极短的语音帧 置零
+    speech_point_list = from_vadArray_to_vadEndpoint(vad)
+    for endpoint in speech_point_list:
+        if endpoint["end"] - endpoint["start"] < min_speech_duration * sr:
+            vad[endpoint["start"]:endpoint["end"]] = 0
+
+    # 把极端的静音帧 置1
+    silent_point_list = from_vadArray_to_vadEndpoint(1 - vad)
+    for endpoint in silent_point_list:
+        if endpoint["end"] - endpoint["start"] < min_silent_duration * sr:
+            vad[endpoint["start"]:endpoint["end"]] = 1
+    return vad

{neverlib-0.2.4.dist-info → neverlib-0.2.6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: neverlib
-Version: 0.2.4
+Version: 0.2.6
 Summary: A successful sign for python setup
 Author-email: "Never.Ling" <1786088386@qq.com>
 License: MIT
@@ -9,19 +9,26 @@ Project-URL: Bug Tracker, https://github.com/yourusername/neverlib/issues
 Project-URL: Documentation, https://github.com/yourusername/neverlib
 Project-URL: Source Code, https://github.com/yourusername/neverlib
 Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Classifier: License :: OSI Approved :: MIT License
 Classifier: Operating System :: OS Independent
 Requires-Python: >=3.6
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: numpy
-Requires-Dist: noisereduce
-Requires-Dist: soundfile
-Requires-Dist: matplotlib
-Requires-Dist: scipy
-Requires-Dist: tqdm
-Requires-Dist: joblib
-Requires-Dist: pydub
+Requires-Dist: lazy-loader>=0.4
+Provides-Extra: analysis
+Requires-Dist: numpy; extra == "analysis"
+Requires-Dist: scipy; extra == "analysis"
+Requires-Dist: tqdm; extra == "analysis"
+Requires-Dist: joblib; extra == "analysis"
+Requires-Dist: noisereduce; extra == "analysis"
+Provides-Extra: io
+Requires-Dist: soundfile; extra == "io"
+Requires-Dist: pydub; extra == "io"
+Provides-Extra: plot
+Requires-Dist: matplotlib; extra == "plot"
 Provides-Extra: vad
 Requires-Dist: torch; extra == "vad"
 Requires-Dist: torchaudio; extra == "vad"
@@ -33,14 +40,11 @@ Requires-Dist: transformers; extra == "vad"
 Provides-Extra: gpu
 Requires-Dist: GPUtil; extra == "gpu"
 Provides-Extra: all
-Requires-Dist: torch; extra == "all"
-Requires-Dist: torchaudio; extra == "all"
-Requires-Dist: librosa; extra == "all"
-Requires-Dist: webrtcvad; extra == "all"
-Requires-Dist: funasr; extra == "all"
-Requires-Dist: openai-whisper; extra == "all"
-Requires-Dist: transformers; extra == "all"
-Requires-Dist: GPUtil; extra == "all"
+Requires-Dist: neverlib[analysis]; extra == "all"
+Requires-Dist: neverlib[io]; extra == "all"
+Requires-Dist: neverlib[plot]; extra == "all"
+Requires-Dist: neverlib[vad]; extra == "all"
+Requires-Dist: neverlib[gpu]; extra == "all"
 Dynamic: license-file
 
 # NeverLib