Minor code cleanup

Author: Bebra777228

audio_separator/separator/uvr_lib_v5/demucs/apply.py

@@ -10,7 +10,6 @@
 from concurrent.futures import ThreadPoolExecutor
 import random
 import typing as tp
-from multiprocessing import Process, Queue, Pipe
 
 import torch as th
 from torch import nn

audio_separator/separator/uvr_lib_v5/playsound.py

@@ -31,7 +31,6 @@ def _playsoundWin(sound, block = True):
     sound = '"' + _canonicalizePath(sound) + '"'
 
     from ctypes import create_unicode_buffer, windll, wintypes
-    from time   import sleep
     windll.winmm.mciSendStringW.argtypes = [wintypes.LPCWSTR, wintypes.LPWSTR, wintypes.UINT, wintypes.HANDLE]
     windll.winmm.mciGetErrorStringW.argtypes = [wintypes.DWORD, wintypes.LPWSTR, wintypes.UINT]
 

tests/unit/test_cli.py

@@ -175,7 +175,7 @@ def test_cli_normalization_threshold_argument(common_expected_args):
             mock_separator.assert_called_once_with(**expected_args)
 
 
-# Test using normalization_threshold argument
+# Test using amplification_threshold argument
 def test_cli_amplification_threshold_argument(common_expected_args):
     test_args = ["cli.py", "test_audio.mp3", "--amplification=0.75"]
     with patch("sys.argv", test_args):
@@ -243,36 +243,6 @@ def test_cli_use_autocast_argument(common_expected_args):
             mock_separator.assert_called_once_with(**expected_args)
 
 
-# Test using custom_output_names argument
-def test_cli_Vocals_output_name_argument(common_expected_args):
-    custom_vocals_names = {"Vocals": "vocals_output"}
-    test_args = ["cli.py", "test_audio.mp3", f"--custom_output_names={json.dumps(custom_vocals_names)}"]
-    with patch("sys.argv", test_args):
-        with patch("audio_separator.separator.Separator") as mock_separator:
-            mock_separator_instance = mock_separator.return_value
-            mock_separator_instance.separate.return_value = ["output_file.mp3"]
-            main()
-
-            # Assertions
-            mock_separator.assert_called_once_with(**common_expected_args)
-            mock_separator_instance.separate.assert_called_once_with("test_audio.mp3", custom_output_names=custom_vocals_names)
-
-
-# Test using custom_output_names argument
-def test_cli_Instrumental_output_name_argument(common_expected_args):
-    custom_instrumental_names = {"Instrumental": "instrumental_output"}
-    test_args = ["cli.py", "test_audio.mp3", f"--custom_output_names={json.dumps(custom_instrumental_names)}"]
-    with patch("sys.argv", test_args):
-        with patch("audio_separator.separator.Separator") as mock_separator:
-            mock_separator_instance = mock_separator.return_value
-            mock_separator_instance.separate.return_value = ["output_file.mp3"]
-            main()
-
-            # Assertions
-            mock_separator.assert_called_once_with(**common_expected_args)
-            mock_separator_instance.separate.assert_called_once_with("test_audio.mp3", custom_output_names=custom_instrumental_names)
-
-
 # Test using custom_output_names arguments
 def test_cli_custom_output_names_argument(common_expected_args):
     custom_names = {

tests/unit/test_stft.py

@@ -1,7 +1,7 @@
 import unittest
 import numpy as np
 import torch
-from unittest.mock import Mock, patch
+from unittest.mock import Mock
 from audio_separator.separator.uvr_lib_v5.stft import STFT
 
 # Short-Time Fourier Transform (STFT) Process Overview:
@@ -121,29 +121,16 @@ def test_prepare_for_istft(self):
         # Assertions
         self.assertEqual(complex_tensor.shape, expected_shape)
 
-    def test_inverse_device_handling(self):
+    def test_inverse_stft(self):
         # Create a mock tensor with the correct input shape
         input_tensor = torch.rand(1, 2, 1025, 32)  # shape matching output of STFT
 
-        # Initialize STFT
-        stft = STFT(logger=MockLogger(), n_fft=2048, hop_length=512, dim_f=1025, device="cpu")
-
         # Apply inverse STFT
-        output_tensor = stft.inverse(input_tensor)
+        output_tensor = self.stft.inverse(input_tensor)
 
         # Check if the output tensor is on the CPU
         self.assertEqual(output_tensor.device.type, "cpu")
 
-    def test_inverse_output_shape(self):
-        # Create a mock tensor
-        input_tensor = torch.rand(1, 2, 1025, 32)  # shape matching output of STFT
-
-        # Initialize STFT
-        stft = STFT(logger=MockLogger(), n_fft=2048, hop_length=512, dim_f=1025, device="cpu")
-
-        # Apply inverse STFT
-        output_tensor = stft.inverse(input_tensor)
-
         # Expected output shape: (Batch size, Channel dimension, Time dimension)
         expected_shape = (1, 2, 7936)  # Calculated based on STFT parameters