mirror of
https://github.com/lrnselfreliance/wrolpi.git
synced 2026-10-26 19:01:39 -04:00
86893ef7f1
* Videos are marked as validated so they will not be tried again. All previous bulk methods are combined into one. * Passing json to fixture factory to avoid writing the files in every test. * Adding log to explain if refresh is slow. * Generating video poster if one cannot be found. * Generating video poster if no poster can be found. * Videos without a video file (yet to be downloaded) are not validated. * Video title is extracted from the video file name as a last resort. * Keeping existing view count because it may be from the channel, which will be more recent. * Downloads should only be downloaded after they are due, not the same day. * Setting manager on new downloads. * Checking if downloading is disabled in loop. * Adding log when refresh is complete. * Not generating posters by default. This is to avoid deleting files. * Not converting posters unless channel has that enabled. * Not logging on valid because all posters will be checked. * Validating videos in upsert. * Using video name parser in upsert. * Using validate_video in upsert so that data is only read once from disk. * Attaching manager in download_manager.get_download. * Adding 22 to preferred video format for odd 720p resolution. Signed-off-by: Roland <roland@learningselfreliance.com>
65 lines
1.8 KiB
Python
Executable file
65 lines
1.8 KiB
Python
Executable file
#! /usr/bin/env python3
|
|
import pathlib
|
|
from pathlib import Path
|
|
from typing import Generator, Union
|
|
|
|
import srt
|
|
import webvtt
|
|
|
|
from wrolpi.common import logger
|
|
from wrolpi.media_path import MediaPath
|
|
from .models import Video
|
|
|
|
|
|
def get_caption_text(caption_path: Union[str, Path]) -> Generator:
|
|
"""
|
|
Return all text from each caption of a caption file.
|
|
"""
|
|
if str(caption_path).endswith('vtt'):
|
|
for caption in webvtt.read(caption_path):
|
|
text = str(caption.text).strip()
|
|
yield text
|
|
else:
|
|
with open(caption_path, 'rt') as fh:
|
|
contents = fh.read()
|
|
for subtitle in srt.parse(contents):
|
|
yield subtitle.content
|
|
|
|
|
|
def get_unique_caption_lines(caption_path: Union[str, Path]) -> Generator:
|
|
"""
|
|
Return all unique lines from each caption of a caption file.
|
|
"""
|
|
last_line = None
|
|
for text in get_caption_text(caption_path):
|
|
for line in text.split('\n'):
|
|
if line and line != last_line:
|
|
last_line = line
|
|
yield line
|
|
|
|
|
|
def get_video_captions(video: Video):
|
|
"""
|
|
Parse video captions from the video's captions file.
|
|
"""
|
|
if isinstance(video.caption_path, MediaPath):
|
|
caption_path = video.caption_path.path
|
|
else:
|
|
caption_path = pathlib.Path(video.caption_path)
|
|
|
|
try:
|
|
lines = get_unique_caption_lines(str(caption_path))
|
|
block = '\n'.join(lines)
|
|
return block
|
|
except UnicodeDecodeError:
|
|
# Failed to decode the caption file
|
|
# TODO handle this error
|
|
pass
|
|
except webvtt.errors.MalformedFileError:
|
|
# File format is broken
|
|
pass
|
|
except webvtt.errors.MalformedCaptionError:
|
|
# Captions form is broken somehow
|
|
pass
|
|
|
|
logger.debug(f'Failed to parse caption file {caption_path}')
|