before_code stringlengths 14 465k | reviewer_comment stringlengths 16 64.5k | after_code stringlengths 9 467k | diff_context stringlengths 0 97k | file_path stringlengths 5 226 | comment_line int32 0 26 | language stringclasses 37
values | quality_score float32 0.07 1 | comment_type stringclasses 9
values | comment_length int32 16 64.5k | before_lines int32 1 17.2k | after_lines int32 1 12.1k | is_negative bool 2
classes | pr_title stringlengths 1 308 | pr_number int32 1 299k | repo_name stringclasses 533
values | repo_stars int64 321 419k | repo_language stringclasses 27
values | reviewer_username stringlengths 0 39 | author_username stringlengths 2 39 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
conda update -n base -c conda-forge conda -y
conda update --all -y
conda clean --index-cache -y
conda clean --packages --tarballs -y
conda create --prefix "$SCRIPT_DIR/$PYTHON_ENV" python=$PYTHON_VERSION -y || return 1
source "$CONDA_ENV" || return 1
conda activate "$SCRIPT_DIR/$PYTHON_ENV" || return 1
... | The output redirection uses '>nul' which is Windows syntax. On Unix-like systems (bash), the correct redirection should be '>/dev/null' or '2>&1' for stderr. The '>nul' redirection will create a file named 'nul' in the current directory instead of suppressing output. | conda update -n base -c conda-forge conda -y
conda update --all -y
conda clean --index-cache -y
conda clean --packages --tarballs -y
conda create --prefix "$SCRIPT_DIR/$PYTHON_ENV" python=$PYTHON_VERSION -y || return 1
source "$CONDA_ENV" || return 1
conda activate "$SCRIPT_DIR/$PYTHON_ENV" || return 1
... | @@ -626,7 +626,7 @@ function check_docker {
function install_python_packages {
echo "[ebook2audiobook] Installing dependencies..."
python3 -m pip cache purge > /dev/null 2>&1
- python3 -m pip install --upgrade pip pip setuptools wheel >nul 2>&1
+ python3 -m pip install --upgrade pip setuptools wheel >nul 2>&1 | ebook2audiobook.sh | 26 | Shell | 0.643 | suggestion | 267 | 51 | 51 | false | v25.12.18 | 1,356 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
class BackgroundDetector:
def __init__(self, wav_file:str):
self.wav_file = wav_file
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = Model.from_pretrained(
default_voice_detection_model,
cache_dir=tts_dir
)
self.pipeline ... | The `_get_pipeline()` method lacks a docstring explaining its purpose and lazy initialization behavior. This is important for maintainability, especially since this is a non-trivial pattern that defers expensive model loading. |
def __init__(self, wav_file: str):
self.wav_file = wav_file
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
self.total_duration = librosa.get_duration(path=self.wav_file)
self._pipeline = None
def _get_pipeline(self):
if self._pipeline is not None:
return self._pipeline
mo... | @@ -9,34 +8,49 @@
class BackgroundDetector:
- def __init__(self, wav_file:str):
- self.wav_file = wav_file
- device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
- model = Model.from_pretrained(
- default_voice_detection_model,
- cache_dir=tts_dir
- ... | lib/classes/background_detector.py | 26 | Python | 0.571 | performance | 226 | 33 | 47 | false | v25.12.18 | 1,358 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
class BackgroundDetector:
def __init__(self, wav_file:str):
self.wav_file = wav_file
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model = Model.from_pretrained(
default_voice_detection_model,
cache_dir=tts_dir
)
self.pipeline ... | The lazy initialization in `_get_pipeline()` is not thread-safe. If multiple threads call `detect()` simultaneously on the same instance, the pipeline could be initialized multiple times, potentially causing race conditions. Consider using a lock or ensuring single-threaded usage is documented. |
def __init__(self, wav_file: str):
self.wav_file = wav_file
self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
self.total_duration = librosa.get_duration(path=self.wav_file)
self._pipeline = None
def _get_pipeline(self):
if self._pipeline is not None:
return self._pipeline
mo... | @@ -9,34 +8,49 @@
class BackgroundDetector:
- def __init__(self, wav_file:str):
- self.wav_file = wav_file
- device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
- model = Model.from_pretrained(
- default_voice_detection_model,
- cache_dir=tts_dir
- ... | lib/classes/background_detector.py | 26 | Python | 0.786 | bug | 295 | 33 | 47 | false | v25.12.18 | 1,358 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
import librosa
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
class BackgroundDetector:
def __init__(self, wav_file: str):
self.wav_file = wav_file
self.device = torch.device("cuda" if... | The docstring claims the pipeline is "safe for multithreading and multiprocessing" but this is misleading. The caching mechanism with threading.Lock only addresses thread safety during pipeline initialization, not during concurrent inference calls. Additionally, the global module-level cache will not work correctly acr... | import librosa
import threading
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
_PIPELINE_CACHE = {}
_PIPELINE_LOCK = threading.Lock()
class BackgroundDetector:
def __init__(self, wav_fi... | @@ -1,38 +1,57 @@
import torch
import librosa
+import threading
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
+
+_PIPELINE_CACHE = {}
+_PIPELINE_LOCK = threading.Lock()
+
+
class B... | lib/classes/background_detector.py | 26 | Python | 0.643 | performance | 583 | 51 | 51 | false | v25.12.18 | 1,359 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
class BackgroundDetector:
def __init__(self, wav_file: str):
self.wav_file = wav_file
self.device = torch.device("cuda" if torch.cuda.is_a... | The cache key uses only device type (e.g., "cuda" or "cpu"), which doesn't account for multiple GPUs. If instances are created with different CUDA devices (e.g., cuda:0, cuda:1), they will all receive the same cached pipeline, potentially causing device mismatch errors. The key should incorporate the full device identi... |
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
_PIPELINE_CACHE = {}
_PIPELINE_LOCK = threading.Lock()
class BackgroundDetector:
def __init__(self, wav_file: str):
self.wav_file = wav_... | @@ -1,38 +1,57 @@
import torch
import librosa
+import threading
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
+
+_PIPELINE_CACHE = {}
+_PIPELINE_LOCK = threading.Lock()
+
+
class B... | lib/classes/background_detector.py | 26 | Python | 1 | performance | 633 | 51 | 51 | false | v25.12.18 | 1,359 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
import torch
import librosa
import threading
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
_PIPELINE_CACHE = {}
_PIPELINE_LOCK = threading.Lock()
class BackgroundDetector:
def __init__... | The removal of the docstring from `_get_pipeline` eliminates important documentation about thread safety and caching behavior. This method implements a double-checked locking pattern for pipeline caching, and the docstring explaining this pattern should be retained to help future maintainers understand why this approac... | import torch
import librosa
import threading
from pyannote.audio import Model
from pyannote.audio.pipelines import VoiceActivityDetection
from lib.conf import tts_dir
from lib.models import default_voice_detection_model
_PIPELINE_CACHE = {}
_PIPELINE_LOCK = threading.Lock()
class BackgroundDetector:
def __init__(... | @@ -20,55 +19,37 @@ def __init__(self, wav_file: str):
self.total_duration = librosa.get_duration(path=self.wav_file)
def _get_pipeline(self) -> VoiceActivityDetection: | lib/classes/background_detector.py | 21 | Python | 1 | performance | 1,227 | 46 | 46 | false | v25.12.18 | 1,360 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
import subprocess, re, sys, gradio as gr
class SubprocessPipe:
def __init__(self, cmd:str, is_gui_process:bool, total_duration:float, msg:str='Processing'):
self.cmd = cmd
self.is_gui_process = is_gui_process
self.total_duration = total_duration
self.msg = msg
self.process =... | The 'tqdm' module is imported but never used directly in the code. The code only parses tqdm output from subprocess stdout using regex. Consider removing this unused import.
```suggestion
import os, subprocess, re, sys, gradio as gr
``` | import os, subprocess, re, sys, tqdm, gradio as gr
class SubprocessPipe:
def __init__(self, cmd:str, is_gui_process:bool, total_duration:float, msg:str='Processing'):
self.cmd = cmd
self.is_gui_process = is_gui_process
self.total_duration = total_duration
self.msg = msg
self... | @@ -1,4 +1,4 @@
-import subprocess, re, sys, gradio as gr
+import os, subprocess, re, sys, tqdm, gradio as gr | lib/classes/subprocess_pipe.py | 1 | Python | 0.857 | suggestion | 236 | 26 | 26 | false | v25.12.19 | 1,363 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
... | The error message when proc_pipe returns False is unclear. It should indicate why the subprocess failed rather than just stating the file path. Consider adding context like 'Demucs subprocess failed for file: {self.wav_file}'.
```suggestion
error = f"_demucs_voice() error: Demucs subprocess failed f... | msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
... | @@ -96,26 +96,25 @@ def _demucs_voice(self)->tuple[bool, str]:
self.wav_file
]
try:
- process = subprocess.run(cmd, check = True)
- self.voice_track = os.path.join(self.demucs_dir, self.voice_track)
- msg = 'Voice track isolation su... | lib/classes/voice_extractor.py | 26 | Python | 1 | bug | 349 | 51 | 51 | false | v25.12.19 | 1,363 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
try:
if device_info_str:
device_info = json.loads(device_info_str)
if device_info:
print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
... | The conda install command uses '-d' flag which should be '-c' for specifying a channel. The correct flag for channel selection is '-c' or '--channel', not '-d'.
```suggestion
subprocess.check_call(['conda', 'install', '-c', 'conda-forge', 'scikit-learn', '-y'])
``` | print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
if device_info['tag'] not in ['cpu', 'unknown', 'unsupported']:
m = re.search(r'\+(.+)$', to... | @@ -736,7 +732,7 @@ def install_device_packages(self, device_info_str:str)->int:
subprocess.check_call([sys.executable, '-m', 'pip', 'install', '--upgrade', '--no-cache-dir', torch_pkg])
subprocess.check_call([sys.executable, '-m', 'pip',... | lib/classes/device_installer.py | 26 | Python | 0.857 | suggestion | 306 | 51 | 51 | false | v25.12.19 | 1,363 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if key in _PIPELINE_CACHE:
return _PIPELINE_CACHE[key]
with _PIPELINE_LOCK:
if key in _PIPELINE_CACHE:
return _PIPELINE_CACHE[key]
model = Model.from_pretrained(
default_voice_detection_model,
cache_dir=tts_dir
)
pipeline = VoiceActivityDetection(segmentation=model)
pipeline.instantiat... | The 'onset' and 'offset' parameters were removed from the pipeline instantiation without explanation. If these parameters are no longer needed, this change is fine, but if they were providing important thresholds for voice activity detection, their removal could affect detection accuracy.
```suggestion
pipe... | Return a cached voice activity detection pipeline for the current device.
This method uses a per-device cache backed by a global dictionary
(`_PIPELINE_CACHE`) and protects lazy initialization with a global lock
(`_PIPELINE_LOCK`). The implementation follows a double-checked locking
... | @@ -13,44 +13,57 @@
class BackgroundDetector:
- def __init__(self, wav_file: str):
- self.wav_file = wav_file
- self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
- self.total_duration = librosa.get_duration(path=self.wav_file)
+ def __init__(self, wav_file: str):
+ self.wav_fi... | lib/classes/background_detector.py | 26 | Python | 0.857 | suggestion | 403 | 34 | 47 | false | v25.12.19 | 1,363 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
try:
if device_info_str:
device_info = json.loads(device_info_str)
if device_info:
print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
... | The conda install command has an invalid flag. The `-d` flag is not a valid conda option. This should likely be `-c` to specify the channel. The correct command should be: `conda install -c conda-forge scikit-learn -y`
```suggestion
subprocess.check_call(['conda', 'install', '-c'... | print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
if device_info['tag'] not in ['cpu', 'unknown', 'unsupported']:
m = re.search(r'\+(.+)$', to... | @@ -736,7 +732,7 @@ def install_device_packages(self, device_info_str:str)->int:
subprocess.check_call([sys.executable, '-m', 'pip', 'install', '--upgrade', '--no-cache-dir', torch_pkg])
subprocess.check_call([sys.executable, '-m', 'pip',... | lib/classes/device_installer.py | 26 | Python | 0.857 | suggestion | 363 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
... | The condition `if proc_pipe:` will always evaluate to True because `proc_pipe` is an object instance, not the boolean result of the subprocess execution. The `SubprocessPipe.__init__` method calls `_run_process()` internally but doesn't store or expose its return value. You need to either store the result of `_run_proc... | msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
... | @@ -96,26 +96,25 @@ def _demucs_voice(self)->tuple[bool, str]:
self.wav_file
]
try:
- process = subprocess.run(cmd, check = True)
- self.voice_track = os.path.join(self.demucs_dir, self.voice_track)
- msg = 'Voice track isolation su... | lib/classes/voice_extractor.py | 26 | Python | 0.857 | suggestion | 876 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
raise ValueError(error)
return False,False,error
def _demucs_voice(self)->tuple[bool, str]:
try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
... | The exception is caught and an error message is set, but the exception is not re-raised and execution continues to the next except block. This means `subprocess.CalledProcessError` will never be properly handled. Either remove this except block if it's no longer needed, or ensure the error is properly returned or raise... | raise ValueError(error)
return False,False,error
def _demucs_voice(self)->tuple[bool, str]:
try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
... | @@ -96,26 +96,25 @@ def _demucs_voice(self)->tuple[bool, str]:
self.wav_file
]
try:
- process = subprocess.run(cmd, check = True)
- self.voice_track = os.path.join(self.demucs_dir, self.voice_track)
- msg = 'Voice track isolation su... | lib/classes/voice_extractor.py | 26 | Python | 0.786 | refactor | 322 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
]
try:
process = subprocess.run(cmd, check = True)
self.voice_trac... | The exception is caught and an error message is set, but the exception is not re-raised and execution continues to the outer except block. This means this generic exception handler will never properly handle exceptions. Either remove this except block if it's no longer needed, or ensure the error is properly returned o... | try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
]
try:
proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process... | @@ -96,26 +96,25 @@ def _demucs_voice(self)->tuple[bool, str]:
self.wav_file
]
try:
- process = subprocess.run(cmd, check = True)
- self.voice_track = os.path.join(self.demucs_dir, self.voice_track)
- msg = 'Voice track isolation su... | lib/classes/voice_extractor.py | 26 | Python | 0.643 | refactor | 329 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
import subprocess, re, sys, gradio as gr
class SubprocessPipe:
def __init__(self, cmd:str, is_gui_process:bool, total_duration:float, msg:str='Processing'):
self.cmd = cmd
self.is_gui_process = is_gui_process
self.total_duration = total_duration
self.msg = msg
self.process =... | The `tqdm` module is imported but never used in the code. The code only parses tqdm output via regex patterns, but doesn't actually use the tqdm library itself. Consider removing this unused import.
```suggestion
import os, subprocess, re, sys, gradio as gr
``` | import os, subprocess, re, sys, tqdm, gradio as gr
class SubprocessPipe:
def __init__(self, cmd:str, is_gui_process:bool, total_duration:float, msg:str='Processing'):
self.cmd = cmd
self.is_gui_process = is_gui_process
self.total_duration = total_duration
self.msg = msg
self... | @@ -1,4 +1,4 @@
-import subprocess, re, sys, gradio as gr
+import os, subprocess, re, sys, tqdm, gradio as gr | lib/classes/subprocess_pipe.py | 1 | Python | 0.857 | suggestion | 261 | 26 | 26 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if key in _PIPELINE_CACHE:
return _PIPELINE_CACHE[key]
with _PIPELINE_LOCK:
if key in _PIPELINE_CACHE:
return _PIPELINE_CACHE[key]
model = Model.from_pretrained(
default_voice_detection_model,
cache_dir=tts_dir
)
pipeline = VoiceActivityDetection(segmentation=model)
pipeline.instantiat... | The `onset` and `offset` parameters have been removed from the pipeline configuration. These parameters control the threshold for voice activity detection onset and offset. Removing them may significantly change the behavior of voice activity detection. If this change is intentional, ensure that the default values used... | Return a cached voice activity detection pipeline for the current device.
This method uses a per-device cache backed by a global dictionary
(`_PIPELINE_CACHE`) and protects lazy initialization with a global lock
(`_PIPELINE_LOCK`). The implementation follows a double-checked locking
... | @@ -13,44 +13,57 @@
class BackgroundDetector:
- def __init__(self, wav_file: str):
- self.wav_file = wav_file
- self.device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
- self.total_duration = librosa.get_duration(path=self.wav_file)
+ def __init__(self, wav_file: str):
+ self.wav_fi... | lib/classes/background_detector.py | 26 | Python | 0.857 | suggestion | 529 | 34 | 47 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
def _demucs_voice(self)->tuple[bool, str]:
try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
]
try:
process = subprocess.run(cm... | The exception is caught and an error message is set, but the exception is not re-raised and execution continues to the next except block. This means `FileNotFoundError` will never be properly handled. Either remove this except block if it's no longer needed, or ensure the error is properly returned or raised.
```sugges... |
def _demucs_voice(self)->tuple[bool, str]:
try:
cmd = [
"demucs",
"--verbose",
"--two-stems=vocals",
"--out", self.output_dir,
self.wav_file
]
try:
proc_pipe = SubprocessPipe(... | @@ -96,26 +96,25 @@ def _demucs_voice(self)->tuple[bool, str]:
self.wav_file
]
try:
- process = subprocess.run(cmd, check = True)
- self.voice_track = os.path.join(self.demucs_dir, self.voice_track)
- msg = 'Voice track isolation su... | lib/classes/voice_extractor.py | 26 | Python | 1 | refactor | 515 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
try:
if device_info_str:
device_info = json.loads(device_info_str)
if device_info:
print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
... | There are two consecutive spaces before the '-y' flag in the command list: `'scikit-learn', '-y'`. While this won't cause a functional issue since these are separate list elements, it's inconsistent with typical code formatting. Consider using single space for consistency.
```suggestion
... | print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
if device_info['tag'] not in ['cpu', 'unknown', 'unsupported']:
m = re.search(r'\+(.+)$', to... | @@ -736,7 +732,7 @@ def install_device_packages(self, device_info_str:str)->int:
subprocess.check_call([sys.executable, '-m', 'pip', 'install', '--upgrade', '--no-cache-dir', torch_pkg])
subprocess.check_call([sys.executable, '-m', 'pip',... | lib/classes/device_installer.py | 26 | Python | 0.857 | suggestion | 419 | 51 | 51 | false | v25.12.19 | 1,364 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
self.voice_file = voice_file
self.voice_name = voice_name
self.voice_track = 'vocals.wav'
self.samplerate = models[session['tts_engine']][session['fine_tuned']]['samplerate']
self.output_dir = self.session['voice_dir']
self.demucs_dir = os.path.join(self.output_dir,'htdem... | The return value from SubprocessPipe is no longer being checked. According to the SubprocessPipe implementation, the _run_process method returns False when the process fails or is stopped, and True only when the process completes successfully with returncode 0. Without checking this return value, the code will proceed ... | self.voice_file = voice_file
self.voice_name = voice_name
self.voice_track = 'vocals.wav'
self.samplerate = models[session['tts_engine']][session['fine_tuned']]['samplerate']
self.output_dir = self.session['voice_dir']
self.demucs_dir = os.path.join(self.output_dir,'htdem... | @@ -46,17 +46,13 @@ def _convert2wav(self)->tuple[bool, str]:
shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', self.voice_file,
'-ac', '1', '-y', self.wav_file
]
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_d... | lib/classes/voice_extractor.py | 26 | Python | 0.929 | bug | 954 | 51 | 51 | false | v25.12.19 | 1,365 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
raise ValueError(error)
return False, error
def _detect_background(self)->tuple[bool,bool,str]:
try:
msg = 'Detecting any background noise or music...'
print(msg)
detector = BackgroundDetector(wav_file = self.wav_file)
status,report = dete... | The return value from SubprocessPipe is no longer being checked. The _run_process method in SubprocessPipe returns False when the process fails and True when it succeeds. Without this check, the code assumes the demucs command always succeeds and immediately proceeds to construct the voice_track path and return success... | try:
msg = 'Detecting any background noise or music...'
print(msg)
detector = BackgroundDetector(wav_file = self.wav_file)
status,report = detector.detect(vad_ratio_thresh = 0.15)
print(report)
if status:
msg = 'Background n... | @@ -95,27 +91,23 @@ def _demucs_voice(self)->tuple[bool, str]:
"--out", self.output_dir,
self.wav_file
]
- try:
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(self.wav_file)... | lib/classes/voice_extractor.py | 26 | Python | 1 | bug | 871 | 51 | 51 | false | v25.12.19 | 1,365 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
DependencyError(e)
return 0
except Exception as e:
error = f"get_audio_duration() Error: Failed to process {filepath}: {e}"
print(error)
return 0
def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=None)->tuple[bool, str]... | The return value from SubprocessPipe is no longer being checked. The SubprocessPipe._run_process method returns False on failure and True on success. Without checking this return value, the code will proceed to check file existence even if ffmpeg failed, potentially producing misleading error messages. The original log... | error = ''
try:
src_file = src_file or self.voice_track
proc_file = proc_file or self.proc_voice_file
dst_file = dst_file or self.final_voice_file
cmd = [shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', src_file]
filter_complex = (
... | @@ -241,20 +233,16 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg=... | lib/classes/voice_extractor.py | 26 | Python | 0.929 | bug | 835 | 51 | 51 | false | v25.12.19 | 1,365 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
if device_info['tag'] not in ['cpu', 'unknown', 'unsupported']:
m = re.search(r'\+(.+)$', to... | The version check logic applies numpy constraint for torch versions <= 2.2.2. Consider adding a comment explaining why this constraint is necessary (likely due to compatibility issues between newer numpy versions and older torch versions) to help future maintainers understand this requirement.
```suggestion
... | print(f'---> Hardware detected: {device_info}')
torch_version = self.get_package_version('torch')
if torch_version:
if device_info['tag'] not in ['cpu', 'unknown', 'unsupported']:
m = re.search(r'\+(.+)$', to... | @@ -732,7 +732,9 @@ def install_device_packages(self, device_info_str:str)->int:
subprocess.check_call([sys.executable, '-m', 'pip', 'install', '--upgrade', '--no-cache-dir', torch_pkg])
subprocess.check_call([sys.executable, '-m', 'pip',... | lib/classes/device_installer.py | 26 | Python | 0.857 | bug | 673 | 51 | 51 | false | v25.12.19 | 1,366 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
proc_file = proc_file or self.proc_voice_file
dst_file = dst_file or self.final_voice_file
cmd = [shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', src_file]
filter_complex = (
'agate=threshold=-25dB:ratio=1.4:attack=10:release=250,'
... | The error message attempts to decode stderr, but if SubprocessPipe doesn't raise CalledProcessError (since it returns a boolean instead of raising exceptions), this catch block may never be reached. Additionally, the variable 'e' might not have a stderr attribute if SubprocessPipe behaves differently than expected. Con... | 'equalizer=f=150:t=q:w=2:g=1,'
'equalizer=f=250:t=q:w=2:g=-3,'
'equalizer=f=3000:t=q:w=2:g=2,'
'equalizer=f=5500:t=q:w=2:g=-4,'
'equalizer=f=9000:t=q:w=2:g=-2,'
'highpass=f=63[audio]'
)
cmd += [
... | @@ -241,20 +233,16 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg=... | lib/classes/voice_extractor.py | 26 | Python | 1 | suggestion | 749 | 51 | 51 | false | v25.12.19 | 1,366 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
self.proc_voice_file = os.path.join(self.session['voice_dir'], f'{self.voice_name}_proc.wav')
self.final_voice_file = os.path.join(self.session['voice_dir'], f'{self.voice_name}.wav')
self.silence_threshold = -60
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.... | The refactored code no longer checks the return value of SubprocessPipe. If SubprocessPipe fails and returns False (or raises an exception), the code will continue to check if the file exists. However, based on SubprocessPipe's implementation, it doesn't raise exceptions on failure - it returns a boolean. The code shou... | self.proc_voice_file = os.path.join(self.session['voice_dir'], f'{self.voice_name}_proc.wav')
self.final_voice_file = os.path.join(self.session['voice_dir'], f'{self.voice_name}.wav')
self.silence_threshold = -60
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.... | @@ -46,17 +46,13 @@ def _convert2wav(self)->tuple[bool, str]:
shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', self.voice_file,
'-ac', '1', '-y', self.wav_file
]
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_d... | lib/classes/voice_extractor.py | 26 | Python | 0.643 | bug | 454 | 51 | 51 | false | v25.12.19 | 1,366 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
def _detect_background(self)->tuple[bool,bool,str]:
try:
msg = 'Detecting any background noise or music...'
print(msg)
detector = BackgroundDetector(wav_file = self.wav_file)
status,report = detector.detect(vad_ratio_thresh = 0.15)
print(report)
... | The refactored code no longer checks the return value of SubprocessPipe. If SubprocessPipe fails and returns False, the code will continue to set self.voice_track and return success. Based on SubprocessPipe's implementation, it returns a boolean to indicate success/failure but doesn't raise exceptions. The code should ... | detector = BackgroundDetector(wav_file = self.wav_file)
status,report = detector.detect(vad_ratio_thresh = 0.15)
print(report)
if status:
msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No... | @@ -95,27 +91,23 @@ def _demucs_voice(self)->tuple[bool, str]:
"--out", self.output_dir,
self.wav_file
]
- try:
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(self.wav_file)... | lib/classes/voice_extractor.py | 26 | Python | 0.643 | bug | 378 | 51 | 51 | false | v25.12.19 | 1,366 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
try:
src_file = src_file or self.voice_track
proc_file = proc_file or self.proc_voice_file
dst_file = dst_file or self.final_voice_file
cmd = [shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', src_file]
filter_complex = (
'agate... | The refactored code no longer checks the return value of SubprocessPipe. If SubprocessPipe fails and returns False, the code will continue to check if the file exists. Based on SubprocessPipe's implementation, it returns a boolean to indicate success/failure but doesn't raise exceptions. The code should check the retur... | 'acompressor=threshold=-20dB:ratio=2:attack=80:release=200:makeup=1dB,'
'loudnorm=I=-14:TP=-3:LRA=7:linear=true,'
'equalizer=f=150:t=q:w=2:g=1,'
'equalizer=f=250:t=q:w=2:g=-3,'
'equalizer=f=3000:t=q:w=2:g=2,'
'equalizer=f=55... | @@ -241,20 +233,16 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- proc_pipe = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg=... | lib/classes/voice_extractor.py | 26 | Python | 0.643 | bug | 370 | 51 | 51 | false | v25.12.19 | 1,366 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | Error message is incomplete and provides no actionable information. Consider including details about what went wrong, such as process exit code or stderr output if available from the SubprocessPipe object.
```suggestion
# SubprocessPipe indicates a failure; include any available diagnostics.
... |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | @@ -46,13 +46,17 @@ def _convert2wav(self)->tuple[bool, str]:
shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', self.voice_file,
'-ac', '1', '-y', self.wav_file
]
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self... | lib/classes/voice_extractor.py | 26 | Python | 1 | bug | 1,199 | 51 | 51 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
raise ValueError(error)
return False,False,error
def _demucs_voice(self)->tuple[bool... | Error message is incomplete and provides no actionable information. Consider including details about what went wrong, such as process exit code or stderr output if available from the SubprocessPipe object.
```suggestion
details_parts = []
returncode = getattr(proc, "returncode", None)
... | print(report)
if status:
msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
... | @@ -91,10 +95,13 @@ def _demucs_voice(self)->tuple[bool, str]:
"--out", self.output_dir,
self.wav_file
]
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(self.wav_file), msg='Denoising')
- se... | lib/classes/voice_extractor.py | 26 | Python | 1 | bug | 917 | 51 | 51 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
cmd += [
'-filter_complex', filter_complex,
'-map', '[audio]',
'-ar', f'{default_audio_proc_samplerate}',
'-y', proc_file
]
try:
SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_du... | Error message is incomplete and provides no actionable information. Consider including details about what went wrong, such as process exit code or stderr output if available from the SubprocessPipe object. | 'equalizer=f=250:t=q:w=2:g=-3,'
'equalizer=f=3000:t=q:w=2:g=2,'
'equalizer=f=5500:t=q:w=2:g=-4,'
'equalizer=f=9000:t=q:w=2:g=-2,'
'highpass=f=63[audio]'
)
cmd += [
'-filter_complex', filter_complex,
... | @@ -233,23 +239,34 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg='Normalize')... | lib/classes/voice_extractor.py | 26 | Python | 0.643 | bug | 205 | 51 | 51 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
Exec=$SCRIPT_DIR/ebook2audiobook.sh
Icon=$ICON_PATH
Terminal=true
Categories=Utility;
EOF
chmod +x "$MENU_ENTRY"
mkdir -p "$HOME/Desktop" 2>&1 > /dev/null
cp "$MENU_ENTRY" "$DESKTOP_SHORTCUT"
chmod +x "$DESKTOP_SHORTCUT"
if command -v update-desktop-database >/dev/null 2>&1; then
update-desktop-database ~/.local... | The logic appears inverted. The function `has_no_display` returns 0 (success/true) when a display IS present, so `! has_no_display` will be true when there's NO display, which contradicts the intended behavior of checking for desktop app viability. Consider changing to `|| has_no_display` without the negation. | open_desktop_app
return 0
fi
mkdir -p "$HOME/.local/share/applications"
cat > "$MENU_ENTRY" <<EOF
[Desktop Entry]
Type=Application
Name=ebook2audiobook
Exec=$SCRIPT_DIR/ebook2audiobook.sh
Icon=$ICON_PATH
Terminal=true
Categories=Utility;
EOF
chmod +x "$MENU_ENTRY"
mkdir -p "$HOME/Desktop" 2>&1 > /dev/null
cp ... | @@ -304,12 +312,12 @@ EOF
}
function check_desktop_app {
- if [[ " ${ARGS[*]} " == *" --headless "* || has_no_display -eq 1 ]]; then
+ if [[ " ${ARGS[*]} " == *" --headless "* ]] || ! has_no_display; then | ebook2audiobook.sh | 26 | Shell | 0.643 | suggestion | 311 | 51 | 51 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
#!/usr/bin/env bash
if [[ "$OSTYPE" == darwin* && -z "$SWITCHED_TO_ZSH" && "$(ps -p $$ -o comm=)" != "zsh" ]]; then
export SWITCHED_TO_ZSH=1
exec env zsh "$0" "$@"
fi
if [[ -n "$BASH_SOURCE" ]]; then
script_path="${BASH_SOURCE[0]}"
elif [[ -n "$ZSH_VERSION" ]]; then
script_path="${(%):-%x}"
else
script_path="$0"... | Variable CURRENT_PYVENV should be quoted consistently with the defensive quoting pattern used elsewhere. Change to `\"$CURRENT_PYVENV\"`.
```suggestion
echo -e "\e[31m=============== Error: Current python virtual environment detected: \"$CURRENT_PYVENV\"..\e[0m"
``` | exit 1
fi
build_docker_image "$device_info_str" || exit 1
elif [[ "$DOCKER_DEVICE_STR" != "" ]];then
install_python_packages || exit 1
install_device_packages "${DOCKER_DEVICE_STR}" || exit 1
check_sitecustomized || exit 1
fi
elif [[ "$SCRIPT_MODE" == "$NATIVE" ]]; then
# Check if running in a... | @@ -766,21 +791,22 @@ else
fi
elif [[ "$SCRIPT_MODE" == "$NATIVE" ]]; then
# Check if running in a Conda or Python virtual environment
- if [[ -n "$CONDA_DEFAULT_ENV" ]]; then
- current_pyvenv="$CONDA_PREFIX"
- elif [[ -n "$VIRTUAL_ENV" ]]; then
+ if [[ -n "${CONDA_DEFAULT_ENV:-}" ]]; then
+ current_pyve... | ebook2audiobook.sh | 26 | Shell | 1 | suggestion | 269 | 803 | 46 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
#!/usr/bin/env bash
if [[ "$OSTYPE" == darwin* && -z "$SWITCHED_TO_ZSH" && "$(ps -p $$ -o comm=)" != "zsh" ]]; then
export SWITCHED_TO_ZSH=1
exec env zsh "$0" "$@"
fi
if [[ -n "$BASH_SOURCE" ]]; then
script_path="${BASH_SOURCE[0]}"
elif [[ -n "$ZSH_VERSION" ]]; then
script_path="${(%):-%x}"
else
script_path="$0"... | The condition checks if CURRENT_PYVENV is empty, but CURRENT_PYVENV is never assigned before this check (it's initialized as empty at line 5). This should check `current_pyvenv` (lowercase) which is set in lines 795 and 797, then assign to CURRENT_PYVENV.
```suggestion
if [[ -z "${current_pyvenv:-}" ]]; then
PYTHO... | fi
device_info_str="$(check_device_info "${SCRIPT_MODE}")"
if [[ "$device_info_str" == "" ]]; then
echo "check_device_info() error: result is empty"
exit 1
fi
build_docker_image "$device_info_str" || exit 1
elif [[ "$DOCKER_DEVICE_STR" != "" ]];then
install_python_packages || exit 1
insta... | @@ -766,21 +791,22 @@ else
fi
elif [[ "$SCRIPT_MODE" == "$NATIVE" ]]; then
# Check if running in a Conda or Python virtual environment
- if [[ -n "$CONDA_DEFAULT_ENV" ]]; then
- current_pyvenv="$CONDA_PREFIX"
- elif [[ -n "$VIRTUAL_ENV" ]]; then
+ if [[ -n "${CONDA_DEFAULT_ENV:-}" ]]; then
+ current_pyve... | ebook2audiobook.sh | 26 | Shell | 1 | suggestion | 662 | 803 | 50 | false | v25.12.19 | 1,367 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
else:
if proc_file != dst_file:
os.replace(proc_file, dst_file)
shutil.rmtree(self.demucs_dir, ignore_errors = True)
msg = 'Audio normalization successful!'
return True, msg
except sub... | Missing a period after the f-string interpolation placeholder. The error message should have proper punctuation between the exception message and the description.
```suggestion
error = f'normalize_audio() FileNotFoundError: {e}. Input file or FFmpeg PATH not found!'
``` | ]
try:
proc = SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg='Normalize')
if not os.path.exists(proc_file) or os.path.getsize(proc_file) == 0:
error = f'normalize_audio(... | @@ -233,23 +239,34 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg='Normalize')... | lib/classes/voice_extractor.py | 26 | Python | 1 | suggestion | 282 | 43 | 51 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
#!/usr/bin/env bash
if [[ "$OSTYPE" == darwin* && -z "$SWITCHED_TO_ZSH" && "$(ps -p $$ -o comm=)" != "zsh" ]]; then
export SWITCHED_TO_ZSH=1
exec env zsh "$0" "$@"
fi
if [[ -n "$BASH_SOURCE" ]]; then
script_path="${BASH_SOURCE[0]}"
elif [[ -n "$ZSH_VERSION" ]]; then
script_path="${(%):-%x}"
else
script_path="$0"... | The condition checks if CURRENT_PYVENV is empty, but then assigns to CURRENT_PYVENV (uppercase), while earlier in the code at lines 794-797 it assigns to 'current_pyvenv' (lowercase). This inconsistency will cause the variable check to always pass since CURRENT_PYVENV won't be set by the earlier assignments. Either all... | exit 1
fi
build_docker_image "$device_info_str" || exit 1
elif [[ "$DOCKER_DEVICE_STR" != "" ]];then
install_python_packages || exit 1
install_device_packages "${DOCKER_DEVICE_STR}" || exit 1
check_sitecustomized || exit 1
fi
elif [[ "$SCRIPT_MODE" == "$NATIVE" ]]; then
# Check if running in a... | @@ -766,21 +791,22 @@ else
fi
elif [[ "$SCRIPT_MODE" == "$NATIVE" ]]; then
# Check if running in a Conda or Python virtual environment
- if [[ -n "$CONDA_DEFAULT_ENV" ]]; then
- current_pyvenv="$CONDA_PREFIX"
- elif [[ -n "$VIRTUAL_ENV" ]]; then
+ if [[ -n "${CONDA_DEFAULT_ENV:-}" ]]; then
+ current_pyve... | ebook2audiobook.sh | 26 | Shell | 0.5 | style | 391 | 803 | 46 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | The error message references 'normalize_audio()' but this code is in the '_convert2wav' method. The error message should reference the correct function name '_convert2wav()'.
```suggestion
error = f'_convert2wav() SubprocessPipe Error:'
``` |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | @@ -46,13 +46,17 @@ def _convert2wav(self)->tuple[bool, str]:
shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', self.voice_file,
'-ac', '1', '-y', self.wav_file
]
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self... | lib/classes/voice_extractor.py | 26 | Python | 1 | suggestion | 260 | 51 | 51 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | The error message ends with a colon but provides no additional information. Consider either adding specific error details or removing the trailing colon to make the message complete.
```suggestion
error = 'normalize_audio() SubprocessPipe Error'
``` |
def _validate_format(self)->tuple[bool,str]:
file_extension = os.path.splitext(self.voice_file)[1].lower()
if file_extension in voice_formats:
msg = 'Input file is valid'
return True,msg
error = f'Unsupported file format: {file_extension}. Supported formats are: {", ... | @@ -46,13 +46,17 @@ def _convert2wav(self)->tuple[bool, str]:
shutil.which('ffmpeg'), '-hide_banner', '-nostats', '-i', self.voice_file,
'-ac', '1', '-y', self.wav_file
]
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self... | lib/classes/voice_extractor.py | 26 | Python | 1 | suggestion | 269 | 51 | 51 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
return True,status,msg
except Exception as e:
error = f'_detect_background() error: {e}'
raise ValueError(error)
return False,False,error
def _demucs_voice(self)->tuple[bool, str]:
try:
cmd = [
"demucs",
"--... | The error message ends with a colon but provides no additional information. Consider either adding specific error details or removing the trailing colon to make the message complete. | if status:
msg = 'Background noise or music detected. Proceeding voice extraction...'
else:
msg = 'No background noise or music detected. Skipping separation...'
return True,status,msg
except Exception as e:
error = f'_detect_backgr... | @@ -91,10 +95,13 @@ def _demucs_voice(self)->tuple[bool, str]:
"--out", self.output_dir,
self.wav_file
]
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(self.wav_file), msg='Denoising')
- se... | lib/classes/voice_extractor.py | 26 | Python | 0.571 | suggestion | 182 | 51 | 51 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
cmd += [
'-filter_complex', filter_complex,
'-map', '[audio]',
'-ar', f'{default_audio_proc_samplerate}',
'-y', proc_file
]
try:
SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_du... | The error message ends with a colon but provides no additional information. Consider either adding specific error details or removing the trailing colon to make the message complete.
```suggestion
error = 'normalize_audio() SubprocessPipe Error'
``` | 'equalizer=f=250:t=q:w=2:g=-3,'
'equalizer=f=3000:t=q:w=2:g=2,'
'equalizer=f=5500:t=q:w=2:g=-4,'
'equalizer=f=9000:t=q:w=2:g=-2,'
'highpass=f=63[audio]'
)
cmd += [
'-filter_complex', filter_complex,
... | @@ -233,23 +239,34 @@ def normalize_audio(self, src_file:str=None, proc_file:str=None, dst_file:str=No
'-y', proc_file
]
try:
- SubprocessPipe(cmd, is_gui_process=self.session['is_gui_process'], total_duration=self._get_audio_duration(src_file), msg='Normalize')... | lib/classes/voice_extractor.py | 26 | Python | 1 | suggestion | 273 | 51 | 51 | false | v25.12.19 | 1,368 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
total_bytes += model_size_bytes(model)
except Exception:
pass
gb = total_bytes / (1024 ** 3)
return round(gb, 2)
def load_xtts_builtin_list()->dict:
try:
if len(xtts_builtin_speakers_list) > 0:
return xtts_builtin_speakers_list
speakers_path = hf_hub_download(repo_id=models[TT... | The new apply_cuda_policy function should include type hints for its parameters to maintain consistency with the rest of the codebase. Consider adding:
```python
def apply_cuda_policy(using_gpu: bool, enough_vram: bool, seed: int) -> None:
```
```suggestion
def apply_cuda_policy(using_gpu: bool, enough_vram: bool, seed... | total_bytes += model_size_bytes(model)
except Exception:
pass
gb = total_bytes / (1024 ** 3)
return round(gb, 2)
def load_xtts_builtin_list()->dict:
try:
if len(xtts_builtin_speakers_list) > 0:
return xtts_builtin_speakers_list
speakers_path = hf_hub_download(repo_id=models[TT... | @@ -57,6 +57,26 @@ def load_xtts_builtin_list()->dict:
"load_xtts_builtin_list() failed"
) from error
+def apply_cuda_policy(using_gpu, enough_vram, seed): | lib/classes/tts_engines/common/utils.py | 26 | Python | 1 | suggestion | 339 | 51 | 51 | false | v25.12.20 | 1,382 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
total_bytes += model_size_bytes(model)
except Exception:
pass
gb = total_bytes / (1024 ** 3)
return round(gb, 2)
def load_xtts_builtin_list()->dict:
try:
if len(xtts_builtin_speakers_list) > 0:
return xtts_builtin_speakers_list
speakers_path = hf_hub_download(repo_id=models[TT... | The function parameters are missing type hints. Add type annotations for the parameters (using_gpu: bool, enough_vram: bool, seed: int) to maintain consistency with other functions in this file and improve code clarity.
```suggestion
def apply_cuda_policy(using_gpu: bool, enough_vram: bool, seed: int) -> None:
``` | total_bytes += model_size_bytes(model)
except Exception:
pass
gb = total_bytes / (1024 ** 3)
return round(gb, 2)
def load_xtts_builtin_list()->dict:
try:
if len(xtts_builtin_speakers_list) > 0:
return xtts_builtin_speakers_list
speakers_path = hf_hub_download(repo_id=models[TT... | @@ -57,6 +57,26 @@ def load_xtts_builtin_list()->dict:
"load_xtts_builtin_list() failed"
) from error
+def apply_cuda_policy(using_gpu, enough_vram, seed): | lib/classes/tts_engines/common/utils.py | 26 | Python | 0.929 | suggestion | 315 | 51 | 51 | false | v25.12.20 | 1,383 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if session['audiobook'] is not None:
return gr.update(choices=audiobook_options, value=session['audiobook'])
else:
return gr.update(choices=audiobook_options, value=audiobook_options[0][1])
... | The indentation change introduces a logic bug. The session restoration code is now nested inside the else block, which means it only executes when data.get('id') returns a truthy value. However, when a new UUID is generated in line 1604 (when data.get('id') is falsy), the session restoration will be skipped entirely. T... | if session['audiobook'] in [option[1] for option in audiobook_options]
else None
)
if len(audiobook_options) > 0:
if session['audiobook'] is not None:
retur... | @@ -1600,9 +1604,9 @@ def change_gr_restore_session(data:DictProxy|None, state:dict, req:gr.Request)->
session = context.set_session(str(uuid.uuid4()))
else:
session = context.set_session(data.get('id'))
- if len(active_sessions) ... | lib/gradio.py | 26 | Python | 1 | bug | 718 | 51 | 51 | false | v25.12.25 | 1,386 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if session['audiobook'] is not None:
return gr.update(choices=audiobook_options, value=session['audiobook'])
else:
return gr.update(choices=audiobook_options, value=audiobook_options[0][1])
... | The indentation change moves the session restoration logic inside the `else` block, which means `restore_session_from_data` will only be called when `data.get('id')` exists. Previously, restoration would occur for both new and existing sessions (based on the `active_sessions` and `data['status']` conditions). This chan... | if session['audiobook'] in [option[1] for option in audiobook_options]
else None
)
if len(audiobook_options) > 0:
if session['audiobook'] is not None:
retur... | @@ -1600,9 +1604,9 @@ def change_gr_restore_session(data:DictProxy|None, state:dict, req:gr.Request)->
session = context.set_session(str(uuid.uuid4()))
else:
session = context.set_session(data.get('id'))
- if len(active_sessions) ... | lib/gradio.py | 26 | Python | 0.857 | bug | 670 | 51 | 51 | false | v25.12.25 | 1,387 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
[laughter]
[laughs]
[sighs]
[music]
[gasps]
[clears throat]
— or ... for hesitations
♪ for song lyrics
C... | The conditional logic adds `speaker_wav` to `tts_dyn_params` only when the .pth file doesn't exist, but `speaker_wav` is already being passed unconditionally on line 191. This creates redundancy and potential confusion about which value takes precedence. Consider either removing the unconditional parameter or clarifyin... | [laughter]
[laughs]
[sighs]
[music]
[gasps]
[clears throat]
— or ... for hesitations
♪ for song lyrics
C... | @@ -165,6 +165,9 @@ def convert(self, sentence_index:int, sentence:str)->bool:
"""
pth_voice_dir = os.path.join(bark_dir, speaker)
pth_voice_file = os.path.join(bark_dir, speaker, f'{speaker}.pth')
+ tts_dyn_params = {}
+ ... | lib/classes/tts_engines/bark.py | 26 | Python | 0.643 | suggestion | 344 | 51 | 51 | false | v25.12.25 | 1,388 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if len(audiobook_options) > 0:
if session['audiobook'] is not None:
return gr.update(choices=audiobook_options, value=session['audiobook'])
else:
return gr.update(choices=audio... | The condition `data.get('status', None) is None` is redundant. The `.get()` method already returns `None` by default when the key doesn't exist, so explicitly passing `None` as the second argument is unnecessary. Simplify to `data.get('status') is None`.
```suggestion
if len(active_sessions) == 0 or... | session['audiobook'] = (
session['audiobook']
if session['audiobook'] in [option[1] for option in audiobook_options]
else None
)
if len(audiobook_options) > 0:
... | @@ -1604,9 +1609,9 @@ def change_gr_restore_session(data:DictProxy|None, state:dict, req:gr.Request)->
session = context.set_session(str(uuid.uuid4()))
else:
session = context.set_session(data.get('id'))
- if len(active_sessio... | lib/gradio.py | 26 | Python | 0.857 | refactor | 363 | 51 | 51 | false | v25.12.25 | 1,389 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
[laughter]
[laughs]
[sighs]
[music]
[gasps]
[clears throat]
— or ... for hesitations
♪ for song lyrics
C... | The variable `tts_dyn_params` could be named more descriptively to clarify its purpose as conditional parameters. Consider renaming to `conditional_params` or `voice_fallback_params` to better convey that it contains fallback parameters when the voice file is missing. | [laughter]
[laughs]
[sighs]
[music]
[gasps]
[clears throat]
— or ... for hesitations
♪ for song lyrics
C... | @@ -165,6 +165,9 @@ def convert(self, sentence_index:int, sentence:str)->bool:
"""
pth_voice_dir = os.path.join(bark_dir, speaker)
pth_voice_file = os.path.join(bark_dir, speaker, f'{speaker}.pth')
+ tts_dyn_params = {}
+ ... | lib/classes/tts_engines/bark.py | 26 | Python | 0.643 | suggestion | 268 | 51 | 51 | false | v25.12.25 | 1,389 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if session:
models = load_engine_presets(session['tts_engine'])
if method == 'confirm_voice_del':
selected_name = Path(voice_path).stem
pattern = re.sub(r'\.wav$', '*.wav', voi... | The refactored logic has a subtle behavior change. In the original code, if voice_options[0][1] was None (which is possible since line 1164 shows ('Default', None) can be added to voice_options), session['voice'] would be set to None. The new code unconditionally sets session['voice'] to models[session['fine_tuned']]['... |
def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, method:str|None=None)->tuple:
try:
nonlocal models
if method is not None:
session = context.get_session(id)
if session:
... | @@ -1067,8 +1073,10 @@ def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, me
shutil.rmtree(custom_model, ignore_errors=True)
msg = f'Custom model {selected_name} deleted!'
... | lib/gradio.py | 26 | Python | 0.857 | refactor | 829 | 51 | 51 | false | v25.12.25 | 1,408 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
... | The audio_sentence variable is referenced but never defined. The line that extracts audio_sentence from the result (audio_sentence = result.get('wav') if isinstance(result, dict) else None) was removed in this change. This will cause a NameError when this code executes. You need to add the extraction of audio_sentence ... | "xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
... | @@ -207,7 +207,16 @@ def _check_xtts_builtin_speakers(self, voice_path:str, speaker:str)->str|bool:
speaker_embedding=speaker_embedding,
**fine_tuned_params,
)
- audio_sentence = result.get('wav') if is... | lib/classes/tts_engines/common/utils.py | 26 | Python | 0.643 | refactor | 357 | 51 | 51 | false | v25.12.25 | 1,411 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
pattern_speaker = re.compile(r"^.*?_speaker_(\d+)$")
bark_options = [
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_spe... | Potential IndexError if voice_options is empty. The code accesses voice_options[0][1] without checking if the list is empty. This could occur if no voice files are found in any of the directories. Consider adding a check: 'if voice_options and voice_options[0][1] is not None' or handle the empty case explicitly before ... | pattern_speaker = re.compile(r"^.*?_speaker_(\d+)$")
bark_options = [
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_spe... | @@ -1163,10 +1163,12 @@ def update_gr_voice_list(id:str)->dict:
if session['tts_engine'] in [TTS_ENGINES['VITS'], TTS_ENGINES['FAIRSEQ'], TTS_ENGINES['TACOTRON2'], TTS_ENGINES['YOURTTS']]:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0... | lib/gradio.py | 26 | Python | 1 | suggestion | 541 | 51 | 51 | false | v25.12.25 | 1,412 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
bark_options = [
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_speaker_*.npz")
]
voice_options ... | The comparison 'session['voice'] not in voice_options' is incorrect. The session['voice'] is a string path, but voice_options is a list of tuples in the format (label, path). This comparison will always be True because it's comparing a string to tuples. The correct check should iterate through the values: 'session['voi... | bark_options = [
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_speaker_*.npz")
]
voice_options ... | @@ -1163,10 +1163,12 @@ def update_gr_voice_list(id:str)->dict:
if session['tts_engine'] in [TTS_ENGINES['VITS'], TTS_ENGINES['FAIRSEQ'], TTS_ENGINES['TACOTRON2'], TTS_ENGINES['YOURTTS']]:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0... | lib/gradio.py | 26 | Python | 0.857 | bug | 561 | 51 | 51 | false | v25.12.25 | 1,412 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if self.session['free_vram_gb'] > models_loaded_size_gb:
loaded_tts[key] = engine
return engine
except Exception as e:
error = f'_load_checkpoint() error: {e}'
print(error)
return None
def _load_engine_zs(self)-... | Potential UnboundLocalError: proc_voice_path is used in the exception handler (line 253) but is only defined inside a nested conditional block (line 228). If an exception occurs before proc_voice_path is defined, this will raise UnboundLocalError. Initialize proc_voice_path to an empty string or None at the beginning o... | if self.session['free_vram_gb'] > models_loaded_size_gb:
loaded_tts[key] = engine
return engine
except Exception as e:
error = f'_load_checkpoint() error: {e}'
print(error)
return None
def _load_engine_zs(self)-... | @@ -150,6 +150,7 @@ def _load_engine_zs(self)->Any:
raise ValueError(error)
def _check_xtts_builtin_speakers(self, voice_path:str, speaker:str)->str|bool:
+ new_voice_path = '' | lib/classes/tts_engines/common/utils.py | 26 | Python | 1 | suggestion | 450 | 51 | 51 | false | v25.12.25 | 1,412 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
if session['tts_engine'] == TTS_ENGINES['BARK']:
bark_visible = visible_gr_tab_bark_params
return (
gr.update(value=show_rating(session['tts_engine'])),
... | Redundant condition: The else block at line 1343-1344 sets visible_custom_model to False, which is already set to False at line 1340. This redundant assignment can be removed or the condition can be simplified.
```suggestion
``` | else:
if session['tts_engine'] == TTS_ENGINES['BARK']:
bark_visible = visible_gr_tab_bark_params
return (
gr.update(value=show_rating(session['tts_engine'])),
gr.update... | @@ -1337,14 +1336,12 @@ def change_gr_fine_tuned_list(selected:str, id:str)->tuple:
if selected:
session = context.get_session(id)
if session:
- models = load_engine_presets(session['tts_engine'])
session['fine_tu... | lib/gradio.py | 26 | Python | 0.857 | suggestion | 229 | 51 | 51 | false | v25.12.25 | 1,412 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
]
voice_options = builtin_options + eng_options + bark_options
session['voice_dir'] = os.path.join(voices_dir, '__sessions', f"voice-{session['id']}", session['language'])
os.makedirs(session['voice_dir'], exist_ok=T... | Missing fallback logic: When new_voice_path doesn't exist, the voice path is left unchanged, which could point to an invalid path. The original code had a fallback to models[session['fine_tuned']]['voice'] when the new_voice_path doesn't exist. Consider adding an else clause to handle this case, or setting session['voi... | ]
voice_options = builtin_options + eng_options + bark_options
session['voice_dir'] = os.path.join(voices_dir, '__sessions', f"voice-{session['id']}", session['language'])
os.makedirs(session['voice_dir'], exist_ok=T... | @@ -1163,10 +1163,12 @@ def update_gr_voice_list(id:str)->dict:
if session['tts_engine'] in [TTS_ENGINES['VITS'], TTS_ENGINES['FAIRSEQ'], TTS_ENGINES['TACOTRON2'], TTS_ENGINES['YOURTTS']]:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0... | lib/gradio.py | 26 | Python | 1 | suggestion | 1,177 | 51 | 51 | false | v25.12.25 | 1,412 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.... | The bare except clause `except Exception:` silently catches all exceptions without logging them. This makes debugging difficult if there are unexpected errors accessing the models dictionary or session configuration. Consider logging the exception or at least being more specific about which exceptions to catch.
```sugg... | if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.... | @@ -1164,11 +1164,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 1 | bug | 559 | 51 | 51 | false | v25.12.25 | 1,413 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_file()
... | The error message variable name 'error' conflicts with potential naming conventions. Consider renaming to 'error_msg' or 'error_message' to avoid confusion with the exception object 'e'. | if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_file()
... | @@ -1177,8 +1177,9 @@ def update_gr_voice_list(id:str)->dict:
fine_tuned_cfg = models.get(session.get('fine_tuned'))
if isinstance(fine_tuned_cfg, dict):
fallback_voice = fine_tuned_cfg.get('vo... | lib/gradio.py | 26 | Python | 0.571 | suggestion | 186 | 51 | 51 | false | v25.12.25 | 1,414 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
fine_tuned_params = {
key.removeprefix("xtts_"): cast_type(self.session[key])
for key, cast_type in {
"xtts_temperature": float,
#"xtts_codec_temperature": float,
... | If `engine.inference()` returns `None` or a non-dictionary type, calling `result.get('wav')` will raise an AttributeError. Add validation to ensure `result` is a dictionary before calling `.get()`.
```suggestion
)
if not isinstance(result, dict):
... | fine_tuned_params = {
key.removeprefix("xtts_"): cast_type(self.session[key])
for key, cast_type in {
"xtts_temperature": float,
#"xtts_codec_temperature": float,
... | @@ -202,13 +202,14 @@ def _check_xtts_builtin_speakers(self, voice_path:str, speaker:str)->str|bool:
if self.session.get(key) is not None
}
with torch.no_grad():
- audio_sentence = engine.inference(
+ ... | lib/classes/tts_engines/common/utils.py | 26 | Python | 0.929 | suggestion | 506 | 51 | 51 | false | v25.12.25 | 1,414 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
version_str = toolkit_version_parse(version_out)
cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cm... | The comparison `tag_letters == 'cu'` is incorrect. `re.match()` returns a Match object (or None), not a string. Use `tag_letters.group()` to extract the matched string, or compare using `tag_letters.group() if tag_letters else None`.
```suggestion
tag_letters = re.match(r"[a-zA-Z]+", forced_tag)
... | version_str = toolkit_version_parse(version_out)
cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cm... | @@ -542,8 +542,15 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- name = 'cpu'
- tag = 'cpu'
+ forced_tag = os.environ.get("DEVICE_TAG")
+ if forced_tag:
+ tag_letters =... | lib/classes/device_installer.py | 26 | Python | 0.857 | bug | 390 | 51 | 51 | false | README Typos | 1,420 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cmp == 0:
devices['XPU']['found'] = True
... | This nested ternary expression is difficult to read and maintain. Consider refactoring to a dictionary mapping or if-elif chain for better clarity.
```suggestion
if tag_letters == 'cu':
name = 'cuda'
elif tag_letters == 'rocm':
name = 'rocm'
... | cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cmp == 0:
devices['XPU']['found'] = True
... | @@ -542,8 +542,15 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- name = 'cpu'
- tag = 'cpu'
+ forced_tag = os.environ.get("DEVICE_TAG")
+ if forced_tag:
+ tag_letters =... | lib/classes/device_installer.py | 26 | Python | 0.857 | refactor | 606 | 51 | 51 | false | README Typos | 1,420 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# ============================================================
eli... | The `re.match()` function returns a Match object, not a string. Comparing `tag_letters == 'cu'` will always be False. Extract the matched group using `tag_letters.group()` if a match is found, or use `re.match(r'^(cu|rocm|jetson|xpu|mps)', forced_tag)` and check `match.group(1)` to get the actual matched string.
```sug... | tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# ============================================================
eli... | @@ -542,8 +542,15 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- name = 'cpu'
- tag = 'cpu'
+ forced_tag = os.environ.get("DEVICE_TAG")
+ if forced_tag:
+ tag_letters =... | lib/classes/device_installer.py | 26 | Python | 0.857 | refactor | 1,073 | 51 | 51 | false | README Typos | 1,421 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cmp == 0:
devices['XPU']['found'] = True
... | The nested ternary operators create a single line that is difficult to read and maintain. Consider refactoring this into a dictionary mapping or an if-elif-else chain for better readability.
```suggestion
tag_match = re.match(r"[a-zA-Z]+", forced_tag)
if tag_match:
ta... | cmp = toolkit_version_compare(version_str, xpu_version_range)
if cmp == -1 or cmp == 1:
msg = f'XPU {version_str} out of supported range {xpu_version_range}. Falling back to CPU.'
elif cmp == 0:
devices['XPU']['found'] = True
... | @@ -542,8 +542,15 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- name = 'cpu'
- tag = 'cpu'
+ forced_tag = os.environ.get("DEVICE_TAG")
+ if forced_tag:
+ tag_letters =... | lib/classes/device_installer.py | 26 | Python | 0.857 | refactor | 714 | 51 | 51 | false | README Typos | 1,421 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
elif cmp == 0:
devices['XPU']['found'] = True
name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ================================================... | The newly added line assumes that `tag_letters` is not None, but `re.match()` can return None if the pattern doesn't match. If `forced_tag` contains no letters (e.g., "123" or special characters only), this will raise an AttributeError when trying to call `.group(0)`. Consider checking if `tag_letters` is not None befo... | elif cmp == 0:
devices['XPU']['found'] = True
name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ================================================... | @@ -545,6 +545,7 @@ def has_working_xpu():
forced_tag = os.environ.get("DEVICE_TAG")
if forced_tag:
tag_letters = re.match(r"[a-zA-Z]+", forced_tag)
+ tag_letters = tag_letters.group(0).lower()
name = 'cuda' if tag_letters == 'cu' else 'rocm' if... | lib/classes/device_installer.py | 26 | Python | 1 | bug | 987 | 51 | 51 | false | v25.12.26 | 1,422 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
elif cmp == 0:
devices['XPU']['found'] = True
name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ================================================... | The code assumes that `re.match()` will always return a match object, but it can return `None` if the pattern doesn't match. This will cause an `AttributeError: 'NoneType' object has no attribute 'group'` when calling `.group(0).lower()` on a None value. You should add a None check before accessing the match object, or... | elif cmp == 0:
devices['XPU']['found'] = True
name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ================================================... | @@ -545,6 +545,7 @@ def has_working_xpu():
forced_tag = os.environ.get("DEVICE_TAG")
if forced_tag:
tag_letters = re.match(r"[a-zA-Z]+", forced_tag)
+ tag_letters = tag_letters.group(0).lower()
name = 'cuda' if tag_letters == 'cu' else 'rocm' if... | lib/classes/device_installer.py | 26 | Python | 1 | suggestion | 1,038 | 51 | 51 | false | v25.12.26 | 1,423 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
print(f"Installing the right library packages for {device_info['name']}...")
os_env = device_info['os']
arch = device_info['arch']
tag = device_info['tag']
... | This commented-out code should be removed rather than left in the codebase. Leaving large blocks of commented code reduces code maintainability and can cause confusion about whether the code might be needed in the future.
```suggestion
``` | print(f"Installing the right library packages for {device_info['name']}...")
os_env = device_info['os']
arch = device_info['arch']
tag = device_info['tag']
... | @@ -734,10 +734,11 @@ def install_device_packages(self, device_info_str:str)->int:
torchaudio_pkg = f'{url}/cpu/torchaudio-{torch_version_base}-{torchaudio_tag_py}-{os_env}_{arch}.whl'
subprocess.check_call([sys.executable, '-m', 'pip', '... | lib/classes/device_installer.py | 26 | Python | 0.857 | suggestion | 240 | 51 | 51 | false | v25.12.26 | 1,425 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
os_env = device_info['os']
arch = device_info['arch']
tag = device_info['tag']
url = torch_matrix[device_info['tag']]['url']
toolkit_version... | The new installation command is missing the '--upgrade' flag that was present in the previous implementation (line 740). This could lead to situations where existing packages are not updated when they should be. Consider adding '--upgrade' to ensure consistent behavior with other installation paths in this function (se... | os_env = device_info['os']
arch = device_info['arch']
tag = device_info['tag']
url = torch_matrix[device_info['tag']]['url']
toolkit_version... | @@ -734,10 +734,11 @@ def install_device_packages(self, device_info_str:str)->int:
torchaudio_pkg = f'{url}/cpu/torchaudio-{torch_version_base}-{torchaudio_tag_py}-{os_env}_{arch}.whl'
subprocess.check_call([sys.executable, '-m', 'pip', '... | lib/classes/device_installer.py | 26 | Python | 0.857 | performance | 627 | 50 | 51 | false | v25.12.26 | 1,425 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
-t "$DOCKER_IMG_NAME" \
. || return 1
fi
if [[ -n "$cmd_options" ]]; then
cmd_extra="$cmd_options "
fi
echo "Docker image ready! to run your docker: "
echo "GUI mode:"
echo " docker run ${cmd_extra}--rm -it -p 7860:7860 $DOCKER_IMG_NAME"
echo "Headless mode:"
echo " docker run ${cmd_extra}--rm -it -v \"... | The TAG variable is set here but not exported or declared as global. The `build_docker_image` function called at line 793 depends on this TAG variable being available. In bash, variables are inherited by functions in the same script, but it would be clearer to either export TAG or pass it as a parameter to make the dep... | fi
echo "Docker image ready! to run your docker: "
echo "GUI mode:"
echo " docker run ${cmd_extra}--rm -it -p 7860:7860 $DOCKER_IMG_NAME"
echo "Headless mode:"
echo " docker run ${cmd_extra}--rm -it -v \"/my/real/ebooks/folder/absolute/path:/app/ebooks\" -v \"/my/real/output/folder/absolute/path:/app/audiobooks\"... | @@ -784,16 +779,17 @@ else
if [[ "$SCRIPT_MODE" == "$BUILD_DOCKER" ]]; then
if [[ "$DOCKER_DEVICE_STR" == "" ]]; then
check_docker || exit 1
- if docker image inspect "$DOCKER_IMG_NAME" >/dev/null 2>&1; then
- echo "[STOP] Docker image '$DOCKER_IMG_NAME' already exists. Aborting build."
- echo "Delete ... | ebook2audiobook.sh | 26 | Shell | 0.786 | suggestion | 481 | 51 | 51 | false | v25.12.27 | 1,428 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
url = torch_matrix[device_info['tag']]['url']
toolkit_version = "".join(c for c in tag if c.isdigit())
if device_info['name'] == devices['JETSON']['proc']:
py_major, py_min... | The `--force` flag is not a valid pip option. The correct flag for forcing reinstallation is `--force-reinstall`. This will cause the pip command to fail with an "unrecognized arguments" error. | os_env = device_info['os']
arch = device_info['arch']
tag = device_info['tag']
url = torch_matrix[device_info['tag']]['url']
toolkit_version... | @@ -738,7 +741,7 @@ def install_device_packages(self, device_info_str:str)->int:
#torch_pkg = f'{url}/{tag}/torch-{torch_version_base}%2B{tag}-{tag_py}-{os_env}_{arch}.whl'
#torchaudio_pkg = f'{url}/{tag}/torchaudio-{torch_version_base}%2... | lib/classes/device_installer.py | 26 | Python | 0.714 | bug | 193 | 48 | 51 | false | v25.12.27 | 1,430 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# ============================================================
elif sys.platform == 'darwin' and arch in ('arm64', 'aarc... | The forced device tag logic has been moved outside the "if tag is None" block, which changes the behavior. Previously, DEVICE_TAG would only be checked when no device was detected. Now it will override any detected device, even if hardware was successfully detected. This makes the DEVICE_TAG environment variable more f... | msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# ============================================================
elif sys.platform == 'darwin' and arch in ('arm64', 'aarc... | @@ -542,17 +542,20 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- forced_tag = os.environ.get("DEVICE_TAG")
- if forced_tag:
- tag_letters = re.match(r"[a-zA-Z]+", forced_tag)
+ n... | lib/classes/device_installer.py | 26 | Python | 0.5 | suggestion | 483 | 51 | 51 | false | v25.12.27 | 1,432 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# =======================================... | This deeply nested ternary expression is difficult to read and maintain. Consider refactoring to use a dictionary mapping or if-elif chain for better readability and maintainability.
```suggestion
tag_to_name = {
'cu': 'cuda',
'rocm': 'rocm',
'... | name = 'xpu'
tag = 'xpu'
else:
msg = 'Intel GPU detected but Intel oneAPI Base Toolkit not installed.'
# ============================================================
# APPLE MPS
# =======================================... | @@ -542,17 +542,20 @@ def has_working_xpu():
# CPU
# ============================================================
if tag is None:
- forced_tag = os.environ.get("DEVICE_TAG")
- if forced_tag:
- tag_letters = re.match(r"[a-zA-Z]+", forced_tag)
+ n... | lib/classes/device_installer.py | 26 | Python | 0.857 | refactor | 487 | 51 | 51 | false | v25.12.27 | 1,432 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
alert_exception(error, id)
return gr.update(), gr.update(visible=False), gr.update(visible=False)
def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, method:str|None=None)->tuple:
try:
nonlocal models
... | The condition `session['custom_model'] in session['voice']` will raise a TypeError if `session['custom_model']` is None, since the `in` operator cannot be used with None as the left operand. This should check if both values are not None, or use a safer comparison like checking if `session['voice']` is not None and then... | alert_exception(error, id)
return gr.update(), gr.update(visible=False), gr.update(visible=False)
def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, method:str|None=None)->tuple:
try:
nonlocal models
... | @@ -1073,10 +1073,7 @@ def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, me
shutil.rmtree(custom_model, ignore_errors=True)
msg = f'Custom model {selected_name} deleted!'
... | lib/gradio.py | 26 | Python | 0.786 | suggestion | 381 | 51 | 51 | false | v25.12.28 | 1,438 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
for name, details in models.items()
if details.get("lang") in ("multi", session["language"])
]
if session['fine_tuned'] in fine_tuned_options:
fine_tuned = session['fine_tuned']
... | The `voice_options` variable used here is a module-level variable that is only updated inside `update_gr_voice_list()`. When `change_gr_language()` is called, `voice_options` will contain voice options from the previous language, not the current one being switched to. This means the code will assign a voice from the wr... | ]
if session['fine_tuned'] in fine_tuned_options:
fine_tuned = session['fine_tuned']
else:
fine_tuned = default_fine_tuned
session['fine_tuned'] = fine_tuned
... | @@ -1256,6 +1254,9 @@ def change_gr_language(selected:str, id:str)->tuple:
if session:
prev = session['language']
session['language'] = selected
+ if session['voice'] is not None:
+ if voice_op... | lib/gradio.py | 26 | Python | 0.786 | bug | 521 | 51 | 51 | false | v25.12.28 | 1,438 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
alert_exception(error, id)
return gr.update(), gr.update(visible=False), gr.update(visible=False)
def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, method:str|None=None)->tuple:
try:
nonlocal models
... | This condition will raise a TypeError if `session['voice']` is None, since you cannot use the `in` operator with None. The check should verify that `session['voice']` is not None before using the `in` operator, or handle the case where either value could be None.
```suggestion
if session... | alert_exception(error, id)
return gr.update(), gr.update(visible=False), gr.update(visible=False)
def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, method:str|None=None)->tuple:
try:
nonlocal models
... | @@ -1073,10 +1073,7 @@ def confirm_deletion(voice_path:str, custom_model:str, audiobook:str, id:str, me
shutil.rmtree(custom_model, ignore_errors=True)
msg = f'Custom model {selected_name} deleted!'
... | lib/gradio.py | 26 | Python | 1 | suggestion | 442 | 51 | 51 | false | v25.12.28 | 1,439 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
for name, details in models.items()
if details.get("lang") in ("multi", session["language"])
]
if session['fine_tuned'] in fine_tuned_options:
fine_tuned = session['fine_tuned']
... | This code references `voice_options` before it has been updated for the new language. The `update_gr_voice_list(id)` call on line 1265 populates `voice_options` with voices for the newly selected language, but this check occurs before that update. This means `voice_options` will contain voices from the previous languag... | ]
if session['fine_tuned'] in fine_tuned_options:
fine_tuned = session['fine_tuned']
else:
fine_tuned = default_fine_tuned
session['fine_tuned'] = fine_tuned
... | @@ -1256,6 +1254,9 @@ def change_gr_language(selected:str, id:str)->tuple:
if session:
prev = session['language']
session['language'] = selected
+ if session['voice'] is not None:
+ if voice_op... | lib/gradio.py | 26 | Python | 0.786 | bug | 751 | 51 | 51 | false | v25.12.28 | 1,439 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.... | The same bounds checking issue exists here. Accessing parts[idx + 1] on line 1176 without verifying that idx + 1 is within bounds can raise an IndexError if "voices" is the last component in the path.
```suggestion
if "voices" in parts:
idx = p... | for f in session_voice_dir.rglob(file_pattern)
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
... | @@ -1161,24 +1162,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 0.929 | suggestion | 936 | 51 | 51 | false | v25.12.28 | 1,440 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_speaker_*.npz")
]
voice_options = builtin_options + eng_options + bark_options
... | The nested condition on line 1166 lacks a None check for session['voice_dir']. If session['voice_dir'] is None, the 'in' operation will raise a TypeError. Although session['voice_dir'] is set earlier in the function, it should be checked here to be defensive, or the logic should be restructured to guarantee it's not No... | bark_options = [
(pattern_speaker.sub(r"Speaker \1", f.stem), str(f.with_suffix(".wav")))
for f in speakers_path.rglob(f"{lang}_speaker_*.npz")
]
voice_options ... | @@ -1161,24 +1162,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 1 | suggestion | 447 | 51 | 51 | false | v25.12.28 | 1,440 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_file()
... | The condition on line 1178 is redundant. The else block at line 1178 already guarantees that session['voice'] is None (from the if condition on line 1165), so the additional check 'if session['voice'] is None' on line 1179 is unnecessary.
```suggestion
if voice_options and voice_options[0][1... | ]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_f... | @@ -1161,24 +1162,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 0.786 | suggestion | 338 | 51 | 51 | false | v25.12.28 | 1,440 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.... | The code assumes that "voices" exists in the path parts without validation. If the voice path doesn't contain "voices" as a directory component, parts.index("voices") on line 1170 will raise a ValueError, which will be caught by the generic exception handler but could lead to unclear error messages. Consider adding val... | for f in session_voice_dir.rglob(file_pattern)
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
... | @@ -1161,24 +1162,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 1 | suggestion | 1,147 | 51 | 51 | false | v25.12.28 | 1,440 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.... | The code attempts to access parts[idx + 1] without verifying that idx + 1 is within bounds. If "voices" is the last directory component in the path, this will raise an IndexError. Add a bounds check to ensure idx + 1 is less than len(parts) before accessing it.
```suggestion
if idx +... | for f in session_voice_dir.rglob(file_pattern)
if f.is_file()
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
... | @@ -1161,24 +1162,22 @@ def update_gr_voice_list(id:str)->dict:
voice_options = [('Default', None)] + sorted(voice_options, key=lambda x: x[0].lower())
else:
voice_options = sorted(voice_options, key=lambda x: x[0].lower())
- ... | lib/gradio.py | 26 | Python | 0.929 | suggestion | 839 | 51 | 51 | false | v25.12.28 | 1,440 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_f... | The added guard checks prevent potential IndexError when "voices" is not in parts or when idx + 1 exceeds the list length. However, the logic doesn't handle the case where the path structure is invalid but "voices" exists. If idx + 1 >= len(parts) and the if condition fails, the session voice is left unchanged, which m... | ]
if session.get('custom_model_dir'):
voice_options.extend(
(f.stem, str(f))
for f in Path(session['custom_model_dir']).rglob('*.wav')
if f.is_f... | @@ -1167,16 +1167,18 @@ def update_gr_voice_list(id:str)->dict:
if not any(v[1] == session['voice'] for v in voice_options):
voice_path = Path(session['voice'])
parts = list(voice_path.parts)
- ... | lib/gradio.py | 26 | Python | 1 | bug | 850 | 51 | 51 | false | v25.12.28 | 1,446 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
if self.session.get(key) is not None
}
with torch.no_grad():
result = self.eng... | The detach().cpu() call at line 154 is redundant for tensors created at line 145 (which already calls detach().cpu()), but it handles the other branches (lines 147-149) where tensors are created without detach().cpu(). However, for numpy arrays and lists converted to tensors, detach() is unnecessary as they don't have ... | #"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
if self.session.get(key) is not None
}
with torch.no_grad():
result = self.eng... | @@ -151,6 +151,7 @@ def convert(self, sentence_index:int, sentence:str)->bool:
error = f"Unsupported XTTSv2 wav type: {type(audio_sentence)}"
print(error)
return False
+ audio_tensor = audio_tensor.detach().cpu... | lib/classes/tts_engines/xtts.py | 26 | Python | 1 | suggestion | 490 | 51 | 51 | false | v25.12.28 | 1,446 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
sentence,
#speaker_wav=self.params['voice_path'],
speaker=speaker,
voice_dir=pth_voice_dir,
**fine_tuned_params
)
"""
... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
elif isins... | sentence,
#speaker_wav=self.params['voice_path'],
speaker=speaker,
voice_dir=pth_voice_dir,
**fine_tuned_params
)
"""
... | @@ -202,10 +202,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/bark.py | 26 | Python | 0.929 | suggestion | 467 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
audio_sentence = self.engine_zs.voice_conversion(
source_wav=source_wav,
target_wav=target_wav
)
else:
error = f'Engine {self.tts_zs_key} is None'
... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | audio_sentence = self.engine_zs.voice_conversion(
source_wav=source_wav,
target_wav=target_wav
)
else:
error = f'Engine {self.tts_zs_key} is None'
... | @@ -183,10 +183,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/vits.py | 26 | Python | 0.786 | suggestion | 277 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -183,10 +183,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/vits.py | 26 | Python | 0.786 | suggestion | 290 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
audio_sentence = self.engine_zs.voice_conversion(
source_wav=source_wav,
target_wav=target_wav
)
else:
error = f'Engine {self.tts_zs_key} is None'
... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | audio_sentence = self.engine_zs.voice_conversion(
source_wav=source_wav,
target_wav=target_wav
)
else:
error = f'Engine {self.tts_zs_key} is None'
... | @@ -198,10 +198,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/tacotron.py | 26 | Python | 0.786 | suggestion | 277 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -198,10 +198,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/tacotron.py | 26 | Python | 0.786 | suggestion | 290 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
elif isins... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -170,10 +170,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/fairseq.py | 26 | Python | 0.929 | suggestion | 467 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
else:
if sentence.endswith("'"):
sentence = sentence[:-1]
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pat... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | else:
if sentence.endswith("'"):
sentence = sentence[:-1]
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pat... | @@ -113,10 +113,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/yourtts.py | 26 | Python | 0.786 | suggestion | 277 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pattern = re.compile(r'[—]')
language = self.session['language_iso1'] if self.session['language_iso1'... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
elif isins... | trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pattern = re.compile(r'[—]')
language = self.session['language_iso1'] if self.session['language_iso1'... | @@ -113,10 +113,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/yourtts.py | 26 | Python | 0.929 | suggestion | 467 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
"xtts_repetition_penalty": float,
#"xtts_cvvp_weight": float,
"xtts_top_k": int,
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | "xtts_repetition_penalty": float,
#"xtts_cvvp_weight": float,
"xtts_top_k": int,
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
... | @@ -142,16 +142,17 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_sentence = result.get('wav')
if is_audio_data_valid(audio_sentence):
if isinstance(audio_sentence, torch.Tensor):
- audio_tensor = audio_se... | lib/classes/tts_engines/xtts.py | 26 | Python | 0.786 | suggestion | 277 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | The `.cpu()` call here is redundant. `torch.tensor()` creates a CPU tensor by default unless a device parameter is explicitly specified. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
elif isins... | "xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | @@ -142,16 +142,17 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_sentence = result.get('wav')
if is_audio_data_valid(audio_sentence):
if isinstance(audio_sentence, torch.Tensor):
- audio_tensor = audio_se... | lib/classes/tts_engines/xtts.py | 26 | Python | 0.929 | suggestion | 467 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
elif isinstance(audio_s... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -170,10 +170,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/fairseq.py | 26 | Python | 0.929 | suggestion | 454 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
with torch.no_grad():
"""
result = self.engine.synthesize(
sentence,
#speaker_wav=self.params['voice_path'],
speaker=speaker,
voice_dir=pth_... | The `.cpu()` call here is redundant. `torch.from_numpy()` always creates a CPU tensor since numpy arrays exist only on CPU. While this doesn't break functionality, it adds unnecessary overhead and makes the code less clear about the actual device management.
```suggestion
``` | with torch.no_grad():
"""
result = self.engine.synthesize(
sentence,
#speaker_wav=self.params['voice_path'],
speaker=speaker,
voice_dir=pth_... | @@ -202,10 +202,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/bark.py | 26 | Python | 0.786 | suggestion | 277 | 51 | 51 | false | v25.12.28 | 1,450 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | The `.cpu()` call is redundant here. `torch.from_numpy()` already creates tensors on CPU by default since numpy arrays exist in CPU memory. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_te... | "xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | @@ -142,16 +142,17 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_sentence = result.get('wav')
if is_audio_data_valid(audio_sentence):
if isinstance(audio_sentence, torch.Tensor):
- audio_tensor = audio_se... | lib/classes/tts_engines/xtts.py | 26 | Python | 0.929 | suggestion | 393 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
"xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | The `.cpu()` call is redundant here. `torch.tensor()` creates tensors on CPU by default unless a device is explicitly specified. This unnecessary operation adds overhead without benefit.
```suggestion
``` | "xtts_top_p": float,
"xtts_speed": float,
#"xtts_gpt_cond_len": int,
#"xtts_gpt_batch_size": int,
"xtts_enable_text_splitting": bool
}.items()
... | @@ -142,16 +142,17 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_sentence = result.get('wav')
if is_audio_data_valid(audio_sentence):
if isinstance(audio_sentence, torch.Tensor):
- audio_tensor = audio_se... | lib/classes/tts_engines/xtts.py | 26 | Python | 0.786 | suggestion | 205 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
else:
if sentence.endswith("'"):
sentence = sentence[:-1]
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pat... | The `.cpu()` call is redundant here. `torch.from_numpy()` already creates tensors on CPU by default since numpy arrays exist in CPU memory. This unnecessary operation adds overhead without benefit.
```suggestion
``` | else:
if sentence.endswith("'"):
sentence = sentence[:-1]
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pat... | @@ -113,10 +113,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/yourtts.py | 26 | Python | 0.929 | suggestion | 216 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pattern = re.compile(r'[—]')
language = self.session['language_iso1'] if self.session['language_iso1'... | The `.cpu()` call is redundant here. `torch.tensor()` creates tensors on CPU by default unless a device is explicitly specified. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_tensor = torc... | trim_audio_buffer = 0.002
sentence += '...' if sentence[-1].isalnum() else ''
speaker_argument = {}
not_supported_punc_pattern = re.compile(r'[—]')
language = self.session['language_iso1'] if self.session['language_iso1'... | @@ -113,10 +113,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/yourtts.py | 26 | Python | 0.929 | suggestion | 382 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.from_numpy()` already creates tensors on CPU by default since numpy arrays exist in CPU memory. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_te... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -183,10 +183,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/vits.py | 26 | Python | 0.929 | suggestion | 393 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.tensor()` creates tensors on CPU by default unless a device is explicitly specified. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_tensor = torc... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -183,10 +183,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/vits.py | 26 | Python | 0.929 | suggestion | 382 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.from_numpy()` already creates tensors on CPU by default since numpy arrays exist in CPU memory. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_te... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -198,10 +198,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/tacotron.py | 26 | Python | 0.929 | suggestion | 393 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.tensor()` creates tensors on CPU by default unless a device is explicitly specified. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_tensor = torc... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -198,10 +198,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/tacotron.py | 26 | Python | 0.929 | suggestion | 382 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.from_numpy()` already creates tensors on CPU by default since numpy arrays exist in CPU memory. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_te... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -170,10 +170,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/fairseq.py | 26 | Python | 0.929 | suggestion | 393 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
)
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | The `.cpu()` call is redundant here. `torch.tensor()` creates tensors on CPU by default unless a device is explicitly specified. This unnecessary operation adds overhead without benefit.
```suggestion
elif isinstance(audio_sentence, (list, tuple)):
audio_tensor = torc... | )
else:
error = f'Engine {self.tts_zs_key} is None'
print(error)
return False
if os.path.exists(tmp_in_wav):
os.remove(tmp_in_wav)
... | @@ -170,10 +170,12 @@ def convert(self, sentence_index:int, sentence:str)->bool:
audio_tensor = audio_sentence.detach().cpu().unsqueeze(0)
elif isinstance(audio_sentence, np.ndarray):
audio_tensor = torch.from_numpy(audio_sentence).unsqu... | lib/classes/tts_engines/fairseq.py | 26 | Python | 0.929 | suggestion | 382 | 51 | 51 | false | v25.12.28 | 1,451 | DrewThomasson/ebook2audiobook | 13,365 | Python | Copilot | ROBERT-MCDOWELL |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.