class Upscaler:
"""An instance of this class is a upscaler that will
upscale all images in the given directory.
Raises:
Exception -- all exceptions
ArgumentError -- if argument is not valid
"""
def __init__(
self,
input_path: pathlib.Path or list,
output_path: pathlib.Path,
driver_settings: dict,
ffmpeg_settings: dict,
gifski_settings: dict,
driver: str = "waifu2x_caffe",
scale_ratio: float = None,
scale_width: int = None,
scale_height: int = None,
processes: int = 1,
video2x_cache_directory: pathlib.Path = pathlib.Path(
tempfile.gettempdir()) / "video2x",
extracted_frame_format: str = "png",
output_file_name_format_string:
str = "{original_file_name}_output{extension}",
image_output_extension: str = ".png",
video_output_extension: str = ".mp4",
preserve_frames: bool = False,
):
# required parameters
self.input = input_path
self.output = output_path
self.driver_settings = driver_settings
self.ffmpeg_settings = ffmpeg_settings
self.gifski_settings = gifski_settings
# optional parameters
self.driver = driver
self.scale_ratio = scale_ratio
self.scale_width = scale_width
self.scale_height = scale_height
self.processes = processes
self.video2x_cache_directory = video2x_cache_directory
self.extracted_frame_format = extracted_frame_format
self.output_file_name_format_string = output_file_name_format_string
self.image_output_extension = image_output_extension
self.video_output_extension = video_output_extension
self.preserve_frames = preserve_frames
# other internal members and signals
self.running = False
self.current_processing_starting_time = time.time()
self.total_frames_upscaled = 0
self.total_frames = 0
self.total_files = 0
self.total_processed = 0
self.scaling_jobs = []
self.current_pass = 0
self.current_input_file = pathlib.Path()
self.last_frame_upscaled = pathlib.Path()
def create_temp_directories(self):
"""create temporary directories"""
# if cache directory unspecified, use %TEMP%\video2x
if self.video2x_cache_directory is None:
self.video2x_cache_directory = (
pathlib.Path(tempfile.gettempdir()) / "video2x")
# if specified cache path exists and isn't a directory
if (self.video2x_cache_directory.exists()
and not self.video2x_cache_directory.is_dir()):
Avalon.error(
_("Specified or default cache directory is a file/link"))
raise FileExistsError(
"Specified or default cache directory is a file/link")
# if cache directory doesn't exist, try creating it
if not self.video2x_cache_directory.exists():
try:
Avalon.debug_info(
_("Creating cache directory {}").format(
self.video2x_cache_directory))
self.video2x_cache_directory.mkdir(parents=True, exist_ok=True)
except Exception as exception:
Avalon.error(
_("Unable to create {}").format(
self.video2x_cache_directory))
raise exception
# create temp directories for extracted frames and upscaled frames
self.extracted_frames = pathlib.Path(
tempfile.mkdtemp(dir=self.video2x_cache_directory))
Avalon.debug_info(
_("Extracted frames are being saved to: {}").format(
self.extracted_frames))
self.upscaled_frames = pathlib.Path(
tempfile.mkdtemp(dir=self.video2x_cache_directory))
Avalon.debug_info(
_("Upscaled frames are being saved to: {}").format(
self.upscaled_frames))
def cleanup_temp_directories(self):
"""delete temp directories when done"""
if not self.preserve_frames:
for directory in [
self.extracted_frames,
self.upscaled_frames,
self.video2x_cache_directory,
]:
try:
# avalon framework cannot be used if python is shutting down
# therefore, plain print is used
print(
_("Cleaning up cache directory: {}").format(directory))
shutil.rmtree(directory)
except FileNotFoundError:
pass
except OSError:
print(_("Unable to delete: {}").format(directory))
traceback.print_exc()
def _check_arguments(self):
if isinstance(self.input, list):
if self.output.exists() and not self.output.is_dir():
Avalon.error(_("Input and output path type mismatch"))
Avalon.error(
_("Input is multiple files but output is not directory"))
raise ArgumentError("input output path type mismatch")
for input_path in self.input:
if not input_path.is_file() and not input_path.is_dir():
Avalon.error(
_("Input path {} is neither a file nor a directory").
format(input_path))
raise FileNotFoundError(
f"{input_path} is neither file nor directory")
with contextlib.suppress(FileNotFoundError):
if input_path.samefile(self.output):
Avalon.error(
_("Input directory and output directory cannot be the same"
))
raise FileExistsError(
"input directory and output directory are the same"
)
# if input is a file
elif self.input.is_file():
if self.output.is_dir():
Avalon.error(_("Input and output path type mismatch"))
Avalon.error(_("Input is single file but output is directory"))
raise ArgumentError("input output path type mismatch")
if self.output.suffix == "":
Avalon.error(_("No suffix found in output file path"))
Avalon.error(_("Suffix must be specified"))
raise ArgumentError("no output file suffix specified")
# if input is a directory
elif self.input.is_dir():
if self.output.is_file():
Avalon.error(_("Input and output path type mismatch"))
Avalon.error(
_("Input is directory but output is existing single file"))
raise ArgumentError("input output path type mismatch")
with contextlib.suppress(FileNotFoundError):
if self.input.samefile(self.output):
Avalon.error(
_("Input directory and output directory cannot be the same"
))
raise FileExistsError(
"input directory and output directory are the same")
# if input is neither
else:
Avalon.error(_("Input path is neither a file nor a directory"))
raise FileNotFoundError(
f"{self.input} is neither file nor directory")
# check FFmpeg settings
ffmpeg_path = pathlib.Path(self.ffmpeg_settings["ffmpeg_path"])
if not ((pathlib.Path(ffmpeg_path / "ffmpeg.exe").is_file()
and pathlib.Path(ffmpeg_path / "ffprobe.exe").is_file()) or
(pathlib.Path(ffmpeg_path / "ffmpeg").is_file()
and pathlib.Path(ffmpeg_path / "ffprobe").is_file())):
Avalon.error(
_("FFmpeg or FFprobe cannot be found under the specified path")
)
Avalon.error(_("Please check the configuration file settings"))
raise FileNotFoundError(self.ffmpeg_settings["ffmpeg_path"])
# check if driver settings
driver_settings = copy.deepcopy(self.driver_settings)
driver_path = driver_settings.pop("path")
# check if driver path exists
if not (pathlib.Path(driver_path).is_file()
or pathlib.Path(f"{driver_path}.exe").is_file()):
Avalon.error(
_("Specified driver executable directory doesn't exist"))
Avalon.error(_("Please check the configuration file settings"))
raise FileNotFoundError(driver_path)
# parse driver arguments using driver's parser
# the parser will throw AttributeError if argument doesn't satisfy constraints
try:
driver_arguments = []
for key in driver_settings.keys():
value = driver_settings[key]
if value is None or value is False:
continue
else:
if len(key) == 1:
driver_arguments.append(f"-{key}")
else:
driver_arguments.append(f"--{key}")
# true means key is an option
if value is not True:
driver_arguments.append(str(value))
DriverWrapperMain = getattr(
importlib.import_module(f"wrappers.{self.driver}"),
"WrapperMain")
DriverWrapperMain.parse_arguments(driver_arguments)
except AttributeError as e:
Avalon.error(
_("Failed to parse driver argument: {}").format(e.args[0]))
raise e
def _upscale_frames(self, input_directory: pathlib.Path,
output_directory: pathlib.Path):
"""Upscale video frames with waifu2x-caffe
This function upscales all the frames extracted
by ffmpeg using the waifu2x-caffe binary.
Args:
input_directory (pathlib.Path): directory containing frames to upscale
output_directory (pathlib.Path): directory which upscaled frames should be exported to
Raises:
UnrecognizedDriverError: raised when the given driver is not recognized
e: re-raised exception after an exception has been captured and finished processing in this scope
"""
# initialize waifu2x driver
if self.driver not in AVAILABLE_DRIVERS:
raise UnrecognizedDriverError(
_("Unrecognized driver: {}").format(self.driver))
# list all images in the extracted frames
frames = [(input_directory / f) for f in input_directory.iterdir()
if f.is_file]
# if we have less images than processes,
# create only the processes necessary
if len(frames) < self.processes:
self.processes = len(frames)
# create a directory for each process and append directory
# name into a list
process_directories = []
for process_id in range(self.processes):
process_directory = input_directory / str(process_id)
process_directories.append(process_directory)
# delete old directories and create new directories
if process_directory.is_dir():
shutil.rmtree(process_directory)
process_directory.mkdir(parents=True, exist_ok=True)
# waifu2x-converter-cpp will perform multi-threading within its own process
if self.driver in [
"waifu2x_converter_cpp",
"waifu2x_ncnn_vulkan",
"srmd_ncnn_vulkan",
"realsr_ncnn_vulkan",
"anime4kcpp",
]:
process_directories = [input_directory]
else:
# evenly distribute images into each directory
# until there is none left in the directory
for image in frames:
# move image
image.rename(process_directories[0] / image.name)
# rotate list
process_directories = (process_directories[-1:] +
process_directories[:-1])
# create driver processes and start them
for process_directory in process_directories:
self.process_pool.append(
self.driver_object.upscale(process_directory,
output_directory))
# start progress bar in a different thread
Avalon.debug_info(_("Starting progress monitor"))
self.progress_monitor = ProgressMonitor(self, process_directories)
self.progress_monitor.start()
# create the clearer and start it
Avalon.debug_info(_("Starting upscaled image cleaner"))
self.image_cleaner = ImageCleaner(input_directory, output_directory,
len(self.process_pool))
self.image_cleaner.start()
# wait for all process to exit
try:
self._wait()
except (Exception, KeyboardInterrupt, SystemExit) as e:
# cleanup
Avalon.debug_info(_("Killing progress monitor"))
self.progress_monitor.stop()
Avalon.debug_info(_("Killing upscaled image cleaner"))
self.image_cleaner.stop()
raise e
# if the driver is waifu2x-converter-cpp
# images need to be renamed to be recognizable for FFmpeg
if self.driver == "waifu2x_converter_cpp":
for image in [
f for f in output_directory.iterdir() if f.is_file()
]:
renamed = re.sub(
f"_\\[.*\\]\\[x(\\d+(\\.\\d+)?)\\]\\.{self.extracted_frame_format}",
f".{self.extracted_frame_format}",
str(image.name),
)
(output_directory / image).rename(output_directory / renamed)
# upscaling done, kill helper threads
Avalon.debug_info(_("Killing progress monitor"))
self.progress_monitor.stop()
Avalon.debug_info(_("Killing upscaled image cleaner"))
self.image_cleaner.stop()
def _terminate_subprocesses(self):
Avalon.warning(_("Terminating all processes"))
for process in self.process_pool:
process.terminate()
def _wait(self):
"""wait for subprocesses in process pool to complete"""
Avalon.debug_info(_("Main process waiting for subprocesses to exit"))
try:
# while process pool not empty
while self.process_pool:
# if stop signal received, terminate all processes
if self.running is False:
raise SystemExit
for process in self.process_pool:
process_status = process.poll()
# if process finished
if process_status is None:
continue
# if return code is not 0
elif process_status != 0:
Avalon.error(
_("Subprocess {} exited with code {}").format(
process.pid, process_status))
raise subprocess.CalledProcessError(
process_status, process.args)
else:
Avalon.debug_info(
_("Subprocess {} exited with code {}").format(
process.pid, process_status))
self.process_pool.remove(process)
time.sleep(0.1)
except (KeyboardInterrupt, SystemExit) as e:
Avalon.warning(_("Stop signal received"))
self._terminate_subprocesses()
raise e
except (Exception, subprocess.CalledProcessError) as e:
Avalon.error(_("Subprocess execution ran into an error"))
self._terminate_subprocesses()
raise e
def run(self):
"""Main controller for Video2X
This function controls the flow of video conversion
and handles all necessary functions.
"""
# external stop signal when called in a thread
self.running = True
# define process pool to contain processes
self.process_pool = []
# load driver modules
DriverWrapperMain = getattr(
importlib.import_module(f"wrappers.{self.driver}"), "WrapperMain")
self.driver_object = DriverWrapperMain(self.driver_settings)
# load options from upscaler class into driver settings
self.driver_object.load_configurations(self)
# initialize FFmpeg object
self.ffmpeg_object = Ffmpeg(
self.ffmpeg_settings,
extracted_frame_format=self.extracted_frame_format)
# define processing queue
self.processing_queue = queue.Queue()
Avalon.info(_("Loading files into processing queue"))
Avalon.debug_info(_("Input path(s): {}").format(self.input))
# make output directory if the input is a list or a directory
if isinstance(self.input, list) or self.input.is_dir():
self.output.mkdir(parents=True, exist_ok=True)
input_files = []
# if input is single directory
# put it in a list for compability with the following code
if not isinstance(self.input, list):
input_paths = [self.input]
else:
input_paths = self.input
# flatten directories into file paths
for input_path in input_paths:
# if the input path is a single file
# add the file's path object to input_files
if input_path.is_file():
input_files.append(input_path)
# if the input path is a directory
# add all files under the directory into the input_files (non-recursive)
elif input_path.is_dir():
input_files.extend(
[f for f in input_path.iterdir() if f.is_file()])
output_paths = []
for input_path in input_files:
# get file type
# try python-magic if it's available
try:
input_file_mime_type = magic.from_file(str(
input_path.absolute()),
mime=True)
input_file_type = input_file_mime_type.split("/")[0]
input_file_subtype = input_file_mime_type.split("/")[1]
except Exception:
input_file_mime_type = input_file_type = input_file_subtype = ""
# if python-magic doesn't determine the file to be an image/video file
# fall back to mimetypes to guess the file type based on the extension
if input_file_type not in ["image", "video"]:
# in case python-magic fails to detect file type
# try guessing file mime type with mimetypes
input_file_mime_type = mimetypes.guess_type(input_path.name)[0]
input_file_type = input_file_mime_type.split("/")[0]
input_file_subtype = input_file_mime_type.split("/")[1]
Avalon.debug_info(
_("File MIME type: {}").format(input_file_mime_type))
# set default output file suffixes
# if image type is GIF, default output suffix is also .gif
if input_file_mime_type == "image/gif":
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem, extension=".gif")
elif input_file_type == "image":
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem,
extension=self.image_output_extension,
)
elif input_file_type == "video":
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem,
extension=self.video_output_extension,
)
# if file is none of: image, image/gif, video
# skip to the next task
else:
Avalon.error(
_("File {} ({}) neither an image nor a video").format(
input_path, input_file_mime_type))
Avalon.warning(_("Skipping this file"))
continue
# if there is only one input file
# do not modify output file suffix
if isinstance(self.input, pathlib.Path) and self.input.is_file():
output_path = self.output
output_path_id = 0
while str(output_path) in output_paths:
output_path = output_path.parent / pathlib.Path(
f"{output_path.stem}_{output_path_id}{output_path.suffix}")
output_path_id += 1
# record output path
output_paths.append(str(output_path))
# push file information into processing queue
self.processing_queue.put((
input_path.absolute(),
output_path.absolute(),
input_file_mime_type,
input_file_type,
input_file_subtype,
))
# check argument sanity before running
self._check_arguments()
# record file count for external calls
self.total_files = self.processing_queue.qsize()
Avalon.info(_("Loaded files into processing queue"))
# print all files in queue for debugging
for job in self.processing_queue.queue:
Avalon.debug_info(_("Input file: {}").format(job[0].absolute()))
try:
while not self.processing_queue.empty():
# get new job from queue
(
self.current_input_file,
output_path,
input_file_mime_type,
input_file_type,
input_file_subtype,
) = self.processing_queue.get()
# get current job starting time for GUI calculations
self.current_processing_starting_time = time.time()
# get video information JSON using FFprobe
Avalon.info(_("Reading file information"))
file_info = self.ffmpeg_object.probe_file_info(
self.current_input_file)
# create temporary directories for storing frames
self.create_temp_directories()
# start handling input
# if input file is a static image
if input_file_type == "image" and input_file_subtype != "gif":
Avalon.info(_("Starting upscaling image"))
# copy original file into the pre-processing directory
shutil.copy(
self.current_input_file,
self.extracted_frames / self.current_input_file.name,
)
width = int(file_info["streams"][0]["width"])
height = int(file_info["streams"][0]["height"])
framerate = self.total_frames = 1
# elif input_file_mime_type == 'image/gif' or input_file_type == 'video':
else:
Avalon.info(_("Starting upscaling video/GIF"))
# find index of video stream
video_stream_index = None
for stream in file_info["streams"]:
if stream["codec_type"] == "video":
video_stream_index = stream["index"]
break
# exit if no video stream found
if video_stream_index is None:
Avalon.error(_("Aborting: No video stream found"))
raise StreamNotFoundError("no video stream found")
# get average frame rate of video stream
framerate = float(
Fraction(file_info["streams"][video_stream_index]
["r_frame_rate"]))
width = int(
file_info["streams"][video_stream_index]["width"])
height = int(
file_info["streams"][video_stream_index]["height"])
# get total number of frames
Avalon.info(
_("Getting total number of frames in the file"))
# if container stores total number of frames in nb_frames, fetch it directly
if "nb_frames" in file_info["streams"][video_stream_index]:
self.total_frames = int(
file_info["streams"][video_stream_index]
["nb_frames"])
# otherwise call FFprobe to count the total number of frames
else:
self.total_frames = self.ffmpeg_object.get_number_of_frames(
self.current_input_file, video_stream_index)
# calculate scale width/height/ratio and scaling jobs if required
Avalon.info(_("Calculating scaling parameters"))
# create a local copy of the global output settings
output_scale = self.scale_ratio
output_width = self.scale_width
output_height = self.scale_height
# calculate output width and height if scale ratio is specified
if output_scale is not None:
output_width = int(
math.ceil(width * output_scale / 2.0) * 2)
output_height = int(
math.ceil(height * output_scale / 2.0) * 2)
else:
# scale keeping aspect ratio is only one of width/height is given
if output_width == 0 or output_width is None:
output_width = output_height / height * width
elif output_height == 0 or output_height is None:
output_height = output_width / width * height
output_width = int(math.ceil(output_width / 2.0) * 2)
output_height = int(math.ceil(output_height / 2.0) * 2)
# calculate required minimum scale ratio
output_scale = max(output_width / width,
output_height / height)
# if driver is one of the drivers that doesn't support arbitrary scaling ratio
# TODO: more documentations on this block
if self.driver in DRIVER_FIXED_SCALING_RATIOS:
# select the optimal driver scaling ratio to use
supported_scaling_ratios = sorted(
DRIVER_FIXED_SCALING_RATIOS[self.driver])
remaining_scaling_ratio = math.ceil(output_scale)
self.scaling_jobs = []
# if the scaling ratio is 1.0
# apply the smallest scaling ratio available
if remaining_scaling_ratio == 1:
self.scaling_jobs.append(supported_scaling_ratios[0])
else:
while remaining_scaling_ratio > 1:
for ratio in supported_scaling_ratios:
if ratio >= remaining_scaling_ratio:
self.scaling_jobs.append(ratio)
remaining_scaling_ratio /= ratio
break
else:
found = False
for i in supported_scaling_ratios:
for j in supported_scaling_ratios:
if i * j >= remaining_scaling_ratio:
self.scaling_jobs.extend([i, j])
remaining_scaling_ratio /= i * j
found = True
break
if found is True:
break
if found is False:
self.scaling_jobs.append(
supported_scaling_ratios[-1])
remaining_scaling_ratio /= supported_scaling_ratios[
-1]
else:
self.scaling_jobs = [output_scale]
# print file information
Avalon.debug_info(_("Framerate: {}").format(framerate))
Avalon.debug_info(_("Width: {}").format(width))
Avalon.debug_info(_("Height: {}").format(height))
Avalon.debug_info(
_("Total number of frames: {}").format(self.total_frames))
Avalon.debug_info(_("Output width: {}").format(output_width))
Avalon.debug_info(_("Output height: {}").format(output_height))
Avalon.debug_info(
_("Required scale ratio: {}").format(output_scale))
Avalon.debug_info(
_("Upscaling jobs queue: {}").format(self.scaling_jobs))
# extract frames from video
if input_file_mime_type == "image/gif" or input_file_type == "video":
self.process_pool.append(
(self.ffmpeg_object.extract_frames(
self.current_input_file, self.extracted_frames)))
self._wait()
# if driver is waifu2x-caffe
# pass pixel format output depth information
if self.driver == "waifu2x_caffe":
# get a dict of all pixel formats and corresponding bit depth
pixel_formats = self.ffmpeg_object.get_pixel_formats()
# try getting pixel format's corresponding bti depth
try:
self.driver_settings["output_depth"] = pixel_formats[
self.ffmpeg_object.pixel_format]
except KeyError:
Avalon.error(
_("Unsupported pixel format: {}").format(
self.ffmpeg_object.pixel_format))
raise UnsupportedPixelError(
f"unsupported pixel format {self.ffmpeg_object.pixel_format}"
)
# upscale images one by one using waifu2x
Avalon.info(_("Starting to upscale extracted frames"))
upscale_begin_time = time.time()
self.current_pass = 1
if self.driver == "waifu2x_caffe":
self.driver_object.set_scale_resolution(
output_width, output_height)
else:
self.driver_object.set_scale_ratio(self.scaling_jobs[0])
self._upscale_frames(self.extracted_frames,
self.upscaled_frames)
for job in self.scaling_jobs[1:]:
self.current_pass += 1
self.driver_object.set_scale_ratio(job)
shutil.rmtree(self.extracted_frames)
shutil.move(self.upscaled_frames, self.extracted_frames)
self.upscaled_frames.mkdir(parents=True, exist_ok=True)
self._upscale_frames(self.extracted_frames,
self.upscaled_frames)
Avalon.info(_("Upscaling completed"))
Avalon.info(
_("Average processing speed: {} seconds per frame").format(
self.total_frames /
(time.time() - upscale_begin_time)))
# downscale frames with Lanczos
Avalon.info(_("Lanczos downscaling frames"))
shutil.rmtree(self.extracted_frames)
shutil.move(self.upscaled_frames, self.extracted_frames)
self.upscaled_frames.mkdir(parents=True, exist_ok=True)
for image in tqdm(
[
i for i in self.extracted_frames.iterdir()
if i.is_file()
and i.name.endswith(self.extracted_frame_format)
],
ascii=True,
desc=_("Downscaling"),
):
image_object = Image.open(image)
# if the image dimensions are not equal to the output size
# resize the image using Lanczos
if (image_object.width, image_object.height) != (
output_width,
output_height,
):
image_object.resize(
(output_width, output_height), Image.LANCZOS).save(
self.upscaled_frames / image.name)
image_object.close()
# if the image's dimensions are already equal to the output size
# move image to the finished directory
else:
image_object.close()
shutil.move(image, self.upscaled_frames / image.name)
# start handling output
# output can be either GIF or video
if input_file_type == "image" and input_file_subtype != "gif":
Avalon.info(_("Exporting image"))
# there should be only one image in the directory
shutil.move(
[
f for f in self.upscaled_frames.iterdir()
if f.is_file()
][0],
output_path,
)
# elif input_file_mime_type == 'image/gif' or input_file_type == 'video':
else:
# if the desired output is gif file
if output_path.suffix.lower() == ".gif":
Avalon.info(
_("Converting extracted frames into GIF image"))
gifski_object = Gifski(self.gifski_settings)
self.process_pool.append(
gifski_object.make_gif(
self.upscaled_frames,
output_path,
framerate,
self.extracted_frame_format,
output_width,
output_height,
))
self._wait()
Avalon.info(_("Conversion completed"))
# if the desired output is video
else:
# frames to video
Avalon.info(
_("Converting extracted frames into video"))
self.process_pool.append(
self.ffmpeg_object.assemble_video(
framerate, self.upscaled_frames))
# f'{scale_width}x{scale_height}'
self._wait()
Avalon.info(_("Conversion completed"))
try:
# migrate audio tracks and subtitles
Avalon.info(
_("Migrating audio, subtitles and other streams to upscaled video"
))
self.process_pool.append(
self.ffmpeg_object.migrate_streams(
self.current_input_file,
output_path,
self.upscaled_frames,
))
self._wait()
# if failed to copy streams
# use file with only video stream
except subprocess.CalledProcessError:
traceback.print_exc()
Avalon.error(_("Failed to migrate streams"))
Avalon.warning(
_("Trying to output video without additional streams"
))
if input_file_mime_type == "image/gif":
# copy will overwrite destination content if exists
shutil.copy(
self.upscaled_frames /
self.ffmpeg_object.intermediate_file_name,
output_path,
)
else:
# construct output file path
output_file_name = f"{output_path.stem}{self.ffmpeg_object.intermediate_file_name.suffix}"
output_video_path = (output_path.parent /
output_file_name)
# if output file already exists
# create temporary directory in output folder
# temporary directories generated by tempfile are guaranteed to be unique
# and won't conflict with other files
if output_video_path.exists():
Avalon.error(_("Output video file exists"))
temporary_directory = pathlib.Path(
tempfile.mkdtemp(
dir=output_path.parent))
output_video_path = (temporary_directory /
output_file_name)
Avalon.info(
_("Created temporary directory to contain file"
))
# move file to new destination
Avalon.info(
_("Writing intermediate file to: {}").
format(output_video_path.absolute()))
shutil.move(
self.upscaled_frames /
self.ffmpeg_object.intermediate_file_name,
output_video_path,
)
# increment total number of files processed
self.cleanup_temp_directories()
self.processing_queue.task_done()
self.total_processed += 1
except (Exception, KeyboardInterrupt, SystemExit) as e:
with contextlib.suppress(ValueError, AttributeError):
self.cleanup_temp_directories()
self.running = False
raise e
# signal upscaling completion
self.running = False
class Upscaler:
""" An instance of this class is a upscaler that will
upscale all images in the given directory.
Raises:
Exception -- all exceptions
ArgumentError -- if argument is not valid
"""
def __init__(self,
input_path: pathlib.Path or list,
output_path: pathlib.Path,
driver_settings: dict,
ffmpeg_settings: dict,
gifski_settings: dict,
driver: str = 'waifu2x_caffe',
scale_ratio: float = None,
processes: int = 1,
video2x_cache_directory: pathlib.Path = pathlib.Path(
tempfile.gettempdir()) / 'video2x',
extracted_frame_format: str = 'png',
output_file_name_format_string:
str = '{original_file_name}_output{extension}',
image_output_extension: str = '.png',
video_output_extension: str = '.mp4',
preserve_frames: bool = False):
# required parameters
self.input = input_path
self.output = output_path
self.driver_settings = driver_settings
self.ffmpeg_settings = ffmpeg_settings
self.gifski_settings = gifski_settings
# optional parameters
self.driver = driver
self.scale_ratio = scale_ratio
self.processes = processes
self.video2x_cache_directory = video2x_cache_directory
self.extracted_frame_format = extracted_frame_format
self.output_file_name_format_string = output_file_name_format_string
self.image_output_extension = image_output_extension
self.video_output_extension = video_output_extension
self.preserve_frames = preserve_frames
# other internal members and signals
self.running = False
self.total_frames_upscaled = 0
self.total_frames = 0
self.total_files = 0
self.total_processed = 0
self.current_input_file = pathlib.Path()
self.last_frame_upscaled = pathlib.Path()
def create_temp_directories(self):
"""create temporary directories
"""
# if cache directory unspecified, use %TEMP%\video2x
if self.video2x_cache_directory is None:
self.video2x_cache_directory = pathlib.Path(
tempfile.gettempdir()) / 'video2x'
# if specified cache path exists and isn't a directory
if self.video2x_cache_directory.exists(
) and not self.video2x_cache_directory.is_dir():
Avalon.error(
_('Specified or default cache directory is a file/link'))
raise FileExistsError(
'Specified or default cache directory is a file/link')
# if cache directory doesn't exist, try creating it
if not self.video2x_cache_directory.exists():
try:
Avalon.debug_info(
_('Creating cache directory {}').format(
self.video2x_cache_directory))
self.video2x_cache_directory.mkdir(parents=True, exist_ok=True)
except Exception as exception:
Avalon.error(
_('Unable to create {}').format(
self.video2x_cache_directory))
raise exception
# create temp directories for extracted frames and upscaled frames
self.extracted_frames = pathlib.Path(
tempfile.mkdtemp(dir=self.video2x_cache_directory))
Avalon.debug_info(
_('Extracted frames are being saved to: {}').format(
self.extracted_frames))
self.upscaled_frames = pathlib.Path(
tempfile.mkdtemp(dir=self.video2x_cache_directory))
Avalon.debug_info(
_('Upscaled frames are being saved to: {}').format(
self.upscaled_frames))
def cleanup_temp_directories(self):
"""delete temp directories when done
"""
if not self.preserve_frames:
for directory in [
self.extracted_frames, self.upscaled_frames,
self.video2x_cache_directory
]:
try:
# avalon framework cannot be used if python is shutting down
# therefore, plain print is used
print(
_('Cleaning up cache directory: {}').format(directory))
shutil.rmtree(directory)
except FileNotFoundError:
pass
except OSError:
print(_('Unable to delete: {}').format(directory))
traceback.print_exc()
def _check_arguments(self):
if isinstance(self.input, list):
if self.output.exists() and not self.output.is_dir():
Avalon.error(_('Input and output path type mismatch'))
Avalon.error(
_('Input is multiple files but output is not directory'))
raise ArgumentError('input output path type mismatch')
for input_path in self.input:
if not input_path.is_file() and not input_path.is_dir():
Avalon.error(
_('Input path {} is neither a file nor a directory').
format(input_path))
raise FileNotFoundError(
f'{input_path} is neither file nor directory')
with contextlib.suppress(FileNotFoundError):
if input_path.samefile(self.output):
Avalon.error(
_('Input directory and output directory cannot be the same'
))
raise FileExistsError(
'input directory and output directory are the same'
)
# if input is a file
elif self.input.is_file():
if self.output.is_dir():
Avalon.error(_('Input and output path type mismatch'))
Avalon.error(_('Input is single file but output is directory'))
raise ArgumentError('input output path type mismatch')
if self.output.suffix == '':
Avalon.error(_('No suffix found in output file path'))
Avalon.error(_('Suffix must be specified'))
raise ArgumentError('no output file suffix specified')
# if input is a directory
elif self.input.is_dir():
if self.output.is_file():
Avalon.error(_('Input and output path type mismatch'))
Avalon.error(
_('Input is directory but output is existing single file'))
raise ArgumentError('input output path type mismatch')
with contextlib.suppress(FileNotFoundError):
if self.input.samefile(self.output):
Avalon.error(
_('Input directory and output directory cannot be the same'
))
raise FileExistsError(
'input directory and output directory are the same')
# if input is neither
else:
Avalon.error(_('Input path is neither a file nor a directory'))
raise FileNotFoundError(
f'{self.input} is neither file nor directory')
# check FFmpeg settings
ffmpeg_path = pathlib.Path(self.ffmpeg_settings['ffmpeg_path'])
if not ((pathlib.Path(ffmpeg_path / 'ffmpeg.exe').is_file()
and pathlib.Path(ffmpeg_path / 'ffprobe.exe').is_file()) or
(pathlib.Path(ffmpeg_path / 'ffmpeg').is_file()
and pathlib.Path(ffmpeg_path / 'ffprobe').is_file())):
Avalon.error(
_('FFmpeg or FFprobe cannot be found under the specified path')
)
Avalon.error(_('Please check the configuration file settings'))
raise FileNotFoundError(self.ffmpeg_settings['ffmpeg_path'])
# check if driver settings
driver_settings = copy.deepcopy(self.driver_settings)
driver_path = driver_settings.pop('path')
# check if driver path exists
if not (pathlib.Path(driver_path).is_file()
or pathlib.Path(f'{driver_path}.exe').is_file()):
Avalon.error(
_('Specified driver executable directory doesn\'t exist'))
Avalon.error(_('Please check the configuration file settings'))
raise FileNotFoundError(driver_path)
# parse driver arguments using driver's parser
# the parser will throw AttributeError if argument doesn't satisfy constraints
try:
driver_arguments = []
for key in driver_settings.keys():
value = driver_settings[key]
if value is None or value is False:
continue
else:
if len(key) == 1:
driver_arguments.append(f'-{key}')
else:
driver_arguments.append(f'--{key}')
# true means key is an option
if value is not True:
driver_arguments.append(str(value))
DriverWrapperMain = getattr(
importlib.import_module(f'wrappers.{self.driver}'),
'WrapperMain')
DriverWrapperMain.parse_arguments(driver_arguments)
except AttributeError as e:
Avalon.error(
_('Failed to parse driver argument: {}').format(e.args[0]))
raise e
# waifu2x-caffe scale_ratio, scale_width and scale_height check
if self.driver == 'waifu2x_caffe':
if (driver_settings['scale_width'] != 0
and driver_settings['scale_height'] == 0
or driver_settings['scale_width'] == 0
and driver_settings['scale_height'] != 0):
Avalon.error(
_('Only one of scale_width and scale_height is specified for waifu2x-caffe'
))
raise AttributeError(
'only one of scale_width and scale_height is specified for waifu2x-caffe'
)
# if scale_width and scale_height are specified, ensure scale_ratio is None
elif self.driver_settings[
'scale_width'] != 0 and self.driver_settings[
'scale_height'] != 0:
self.driver_settings['scale_ratio'] = None
# if scale_width and scale_height not specified
# ensure they are None, not 0
else:
self.driver_settings['scale_width'] = None
self.driver_settings['scale_height'] = None
def _upscale_frames(self):
""" Upscale video frames with waifu2x-caffe
This function upscales all the frames extracted
by ffmpeg using the waifu2x-caffe binary.
Arguments:
w2 {Waifu2x Object} -- initialized waifu2x object
"""
# initialize waifu2x driver
if self.driver not in AVAILABLE_DRIVERS:
raise UnrecognizedDriverError(
_('Unrecognized driver: {}').format(self.driver))
# list all images in the extracted frames
frames = [(self.extracted_frames / f)
for f in self.extracted_frames.iterdir() if f.is_file]
# if we have less images than processes,
# create only the processes necessary
if len(frames) < self.processes:
self.processes = len(frames)
# create a directory for each process and append directory
# name into a list
process_directories = []
for process_id in range(self.processes):
process_directory = self.extracted_frames / str(process_id)
process_directories.append(process_directory)
# delete old directories and create new directories
if process_directory.is_dir():
shutil.rmtree(process_directory)
process_directory.mkdir(parents=True, exist_ok=True)
# waifu2x-converter-cpp will perform multi-threading within its own process
if self.driver in [
'waifu2x_converter_cpp', 'waifu2x_ncnn_vulkan',
'srmd_ncnn_vulkan', 'realsr_ncnn_vulkan', 'anime4kcpp'
]:
process_directories = [self.extracted_frames]
else:
# evenly distribute images into each directory
# until there is none left in the directory
for image in frames:
# move image
image.rename(process_directories[0] / image.name)
# rotate list
process_directories = process_directories[
-1:] + process_directories[:-1]
# create driver processes and start them
for process_directory in process_directories:
self.process_pool.append(
self.driver_object.upscale(process_directory,
self.upscaled_frames))
# start progress bar in a different thread
Avalon.debug_info(_('Starting progress monitor'))
self.progress_monitor = ProgressMonitor(self, process_directories)
self.progress_monitor.start()
# create the clearer and start it
Avalon.debug_info(_('Starting upscaled image cleaner'))
self.image_cleaner = ImageCleaner(self.extracted_frames,
self.upscaled_frames,
len(self.process_pool))
self.image_cleaner.start()
# wait for all process to exit
try:
self._wait()
except (Exception, KeyboardInterrupt, SystemExit) as e:
# cleanup
Avalon.debug_info(_('Killing progress monitor'))
self.progress_monitor.stop()
Avalon.debug_info(_('Killing upscaled image cleaner'))
self.image_cleaner.stop()
raise e
# if the driver is waifu2x-converter-cpp
# images need to be renamed to be recognizable for FFmpeg
if self.driver == 'waifu2x_converter_cpp':
for image in [
f for f in self.upscaled_frames.iterdir() if f.is_file()
]:
renamed = re.sub(
f'_\\[.*\\]\\[x(\\d+(\\.\\d+)?)\\]\\.{self.extracted_frame_format}',
f'.{self.extracted_frame_format}', str(image.name))
(self.upscaled_frames / image).rename(self.upscaled_frames /
renamed)
# upscaling done, kill helper threads
Avalon.debug_info(_('Killing progress monitor'))
self.progress_monitor.stop()
Avalon.debug_info(_('Killing upscaled image cleaner'))
self.image_cleaner.stop()
def _terminate_subprocesses(self):
Avalon.warning(_('Terminating all processes'))
for process in self.process_pool:
process.terminate()
def _wait(self):
""" wait for subprocesses in process pool to complete
"""
Avalon.debug_info(_('Main process waiting for subprocesses to exit'))
try:
# while process pool not empty
while self.process_pool:
# if stop signal received, terminate all processes
if self.running is False:
raise SystemExit
for process in self.process_pool:
process_status = process.poll()
# if process finished
if process_status is None:
continue
# if return code is not 0
elif process_status != 0:
Avalon.error(
_('Subprocess {} exited with code {}').format(
process.pid, process_status))
raise subprocess.CalledProcessError(
process_status, process.args)
else:
Avalon.debug_info(
_('Subprocess {} exited with code {}').format(
process.pid, process_status))
self.process_pool.remove(process)
time.sleep(0.1)
except (KeyboardInterrupt, SystemExit) as e:
Avalon.warning(_('Stop signal received'))
self._terminate_subprocesses()
raise e
except (Exception, subprocess.CalledProcessError) as e:
Avalon.error(_('Subprocess execution ran into an error'))
self._terminate_subprocesses()
raise e
def run(self):
""" Main controller for Video2X
This function controls the flow of video conversion
and handles all necessary functions.
"""
# external stop signal when called in a thread
self.running = True
# define process pool to contain processes
self.process_pool = []
# load driver modules
DriverWrapperMain = getattr(
importlib.import_module(f'wrappers.{self.driver}'), 'WrapperMain')
self.driver_object = DriverWrapperMain(self.driver_settings)
# load options from upscaler class into driver settings
self.driver_object.load_configurations(self)
# initialize FFmpeg object
self.ffmpeg_object = Ffmpeg(
self.ffmpeg_settings,
extracted_frame_format=self.extracted_frame_format)
# define processing queue
self.processing_queue = queue.Queue()
Avalon.info(_('Loading files into processing queue'))
Avalon.debug_info(_('Input path(s): {}').format(self.input))
# make output directory if the input is a list or a directory
if isinstance(self.input, list) or self.input.is_dir():
self.output.mkdir(parents=True, exist_ok=True)
input_files = []
# if input is single directory
# put it in a list for compability with the following code
if not isinstance(self.input, list):
input_paths = [self.input]
else:
input_paths = self.input
# flatten directories into file paths
for input_path in input_paths:
# if the input path is a single file
# add the file's path object to input_files
if input_path.is_file():
input_files.append(input_path)
# if the input path is a directory
# add all files under the directory into the input_files (non-recursive)
elif input_path.is_dir():
input_files.extend(
[f for f in input_path.iterdir() if f.is_file()])
output_paths = []
for input_path in input_files:
# get file type
# try python-magic if it's available
try:
input_file_mime_type = magic.from_file(str(
input_path.absolute()),
mime=True)
input_file_type = input_file_mime_type.split('/')[0]
input_file_subtype = input_file_mime_type.split('/')[1]
except Exception:
input_file_type = input_file_subtype = None
# in case python-magic fails to detect file type
# try guessing file mime type with mimetypes
if input_file_type not in ['image', 'video']:
input_file_mime_type = mimetypes.guess_type(input_path.name)[0]
input_file_type = input_file_mime_type.split('/')[0]
input_file_subtype = input_file_mime_type.split('/')[1]
# set default output file suffixes
# if image type is GIF, default output suffix is also .gif
if input_file_mime_type == 'image/gif':
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem, extension='.gif')
elif input_file_type == 'image':
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem,
extension=self.image_output_extension)
elif input_file_type == 'video':
output_path = self.output / self.output_file_name_format_string.format(
original_file_name=input_path.stem,
extension=self.video_output_extension)
# if file is none of: image, image/gif, video
# skip to the next task
else:
Avalon.error(
_('File {} ({}) neither an image nor a video').format(
input_path, input_file_mime_type))
Avalon.warning(_('Skipping this file'))
continue
# if there is only one input file
# do not modify output file suffix
if isinstance(self.input, pathlib.Path) and self.input.is_file():
output_path = self.output
output_path_id = 0
while str(output_path) in output_paths:
output_path = output_path.parent / pathlib.Path(
f'{output_path.stem}_{output_path_id}{output_path.suffix}')
output_path_id += 1
# record output path
output_paths.append(str(output_path))
# push file information into processing queue
self.processing_queue.put(
(input_path.absolute(), output_path.absolute(),
input_file_mime_type, input_file_type, input_file_subtype))
# check argument sanity before running
self._check_arguments()
# record file count for external calls
self.total_files = self.processing_queue.qsize()
Avalon.info(_('Loaded files into processing queue'))
# print all files in queue for debugging
for job in self.processing_queue.queue:
Avalon.debug_info(_('Input file: {}').format(job[0].absolute()))
try:
while not self.processing_queue.empty():
# get new job from queue
self.current_input_file, output_path, input_file_mime_type, input_file_type, input_file_subtype = self.processing_queue.get(
)
# start handling input
# if input file is a static image
if input_file_type == 'image' and input_file_subtype != 'gif':
Avalon.info(_('Starting to upscale image'))
self.process_pool.append(
self.driver_object.upscale(self.current_input_file,
output_path))
self._wait()
Avalon.info(_('Upscaling completed'))
# static images don't require GIF or video encoding
# go to the next task
self.processing_queue.task_done()
self.total_processed += 1
continue
# if input file is a image/gif file or a video
elif input_file_mime_type == 'image/gif' or input_file_type == 'video':
self.create_temp_directories()
# get video information JSON using FFprobe
Avalon.info(_('Reading video information'))
video_info = self.ffmpeg_object.probe_file_info(
self.current_input_file)
# analyze original video with FFprobe and retrieve framerate
# width, height = info['streams'][0]['width'], info['streams'][0]['height']
# find index of video stream
video_stream_index = None
for stream in video_info['streams']:
if stream['codec_type'] == 'video':
video_stream_index = stream['index']
break
# exit if no video stream found
if video_stream_index is None:
Avalon.error(_('Aborting: No video stream found'))
raise StreamNotFoundError('no video stream found')
# get average frame rate of video stream
framerate = float(
Fraction(video_info['streams'][video_stream_index]
['r_frame_rate']))
Avalon.info(_('Framerate: {}').format(framerate))
# self.ffmpeg_object.pixel_format = video_info['streams'][video_stream_index]['pix_fmt']
# extract frames from video
self.process_pool.append(
(self.ffmpeg_object.extract_frames(
self.current_input_file, self.extracted_frames)))
self._wait()
# if driver is waifu2x-caffe
# pass pixel format output depth information
if self.driver == 'waifu2x_caffe':
# get a dict of all pixel formats and corresponding bit depth
pixel_formats = self.ffmpeg_object.get_pixel_formats()
# try getting pixel format's corresponding bti depth
try:
self.driver_settings[
'output_depth'] = pixel_formats[
self.ffmpeg_object.pixel_format]
except KeyError:
Avalon.error(
_('Unsupported pixel format: {}').format(
self.ffmpeg_object.pixel_format))
raise UnsupportedPixelError(
f'unsupported pixel format {self.ffmpeg_object.pixel_format}'
)
# width/height will be coded width/height x upscale factor
# original_width = video_info['streams'][video_stream_index]['width']
# original_height = video_info['streams'][video_stream_index]['height']
# scale_width = int(self.scale_ratio * original_width)
# scale_height = int(self.scale_ratio * original_height)
# upscale images one by one using waifu2x
Avalon.info(_('Starting to upscale extracted frames'))
self._upscale_frames()
Avalon.info(_('Upscaling completed'))
# start handling output
# output can be either GIF or video
# if the desired output is gif file
if output_path.suffix.lower() == '.gif':
Avalon.info(
_('Converting extracted frames into GIF image'))
gifski_object = Gifski(self.gifski_settings)
self.process_pool.append(
gifski_object.make_gif(self.upscaled_frames,
output_path, framerate,
self.extracted_frame_format))
self._wait()
Avalon.info(_('Conversion completed'))
# if the desired output is video
else:
# frames to video
Avalon.info(_('Converting extracted frames into video'))
self.process_pool.append(
self.ffmpeg_object.assemble_video(
framerate, self.upscaled_frames))
# f'{scale_width}x{scale_height}'
self._wait()
Avalon.info(_('Conversion completed'))
try:
# migrate audio tracks and subtitles
Avalon.info(
_('Migrating audio, subtitles and other streams to upscaled video'
))
self.process_pool.append(
self.ffmpeg_object.migrate_streams(
self.current_input_file, output_path,
self.upscaled_frames))
self._wait()
# if failed to copy streams
# use file with only video stream
except subprocess.CalledProcessError:
traceback.print_exc()
Avalon.error(_('Failed to migrate streams'))
Avalon.warning(
_('Trying to output video without additional streams'
))
if input_file_mime_type == 'image/gif':
# copy will overwrite destination content if exists
shutil.copy(
self.upscaled_frames /
self.ffmpeg_object.intermediate_file_name,
output_path)
else:
# construct output file path
output_file_name = f'{output_path.stem}{self.ffmpeg_object.intermediate_file_name.suffix}'
output_video_path = output_path.parent / output_file_name
# if output file already exists
# create temporary directory in output folder
# temporary directories generated by tempfile are guaranteed to be unique
# and won't conflict with other files
if output_video_path.exists():
Avalon.error(_('Output video file exists'))
temporary_directory = pathlib.Path(
tempfile.mkdtemp(dir=output_path.parent))
output_video_path = temporary_directory / output_file_name
Avalon.info(
_('Created temporary directory to contain file'
))
# move file to new destination
Avalon.info(
_('Writing intermediate file to: {}').format(
output_video_path.absolute()))
shutil.move(
self.upscaled_frames /
self.ffmpeg_object.intermediate_file_name,
output_video_path)
# increment total number of files processed
self.cleanup_temp_directories()
self.processing_queue.task_done()
self.total_processed += 1
except (Exception, KeyboardInterrupt, SystemExit) as e:
with contextlib.suppress(ValueError, AttributeError):
self.cleanup_temp_directories()
self.running = False
raise e
# signal upscaling completion
self.running = False