mirrors/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2025-06-28 01:18:30 +00:00
Code Issues Projects Releases Wiki Activity

[pp/ffmpeg] improve WebVTT filtering for MKV

Browse Source
This commit is contained in:
Marcos Del Sol Vives 2025-05-20 11:20:47 +02:00
parent b343713d97
commit 004f028a46
No known key found for this signature in database
1 changed files with 15 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
yt_dlp/postprocessor/ffmpeg.py
View File

@ -619,6 +619,7 @@ def run(self, info):
sub_langs, sub_names, sub_filenames = [], [], [] sub_langs, sub_names, sub_filenames = [], [], []
webm_vtt_warn = False webm_vtt_warn = False
mp4_ass_warn = False mp4_ass_warn = False
files_to_delete = []
for lang, sub_info in subtitles.items(): for lang, sub_info in subtitles.items():
if not os.path.exists(sub_info.get('filepath', '')): if not os.path.exists(sub_info.get('filepath', '')):
@ -632,20 +633,25 @@ def run(self, info):
# If the WebVTT contains empty text cues, it'll corrupt remuxed Matroska files, dropping frames # If the WebVTT contains empty text cues, it'll corrupt remuxed Matroska files, dropping frames
sub_langs.append(lang) sub_langs.append(lang)
sub_names.append(sub_info.get('name')) sub_names.append(sub_info.get('name'))
sub_filename = sub_info['filepath'] original_file = sub_info['filepath']
if not self._already_have_subtitle:
files_to_delete.append(original_file)
with open(sub_info['filepath'], 'rb') as f: with open(original_file, 'rb') as f:
vtt_bytes = f.read() vtt_bytes = f.read()
blocks = list(webvtt.parse_fragment(vtt_bytes)) blocks = list(webvtt.parse_fragment(vtt_bytes))
nonempty_blocks = [block for block in blocks if not isinstance(block, webvtt.CueBlock) or block.text.strip() != ''] nonempty_blocks = [block for block in blocks if not isinstance(block, webvtt.CueBlock) or block.text.strip() != '']
if len(blocks) != len(nonempty_blocks): if len(blocks) != len(nonempty_blocks):
self.report_warning(f'Filtering WebVTT for {lang} subtitle due to ffmpeg bug #11493') self.report_warning(f'Filtering WebVTT for {lang} subtitle due to ffmpeg bug #11493')
sub_filename = f'{sub_filename}_filtered.vtt' temp_file = prepend_extension(original_file, 'filtered')
with open(sub_filename, 'w') as f: with open(temp_file, 'w', encoding='utf-8', newline='\n') as f:
for block in nonempty_blocks: for block in nonempty_blocks:
block.write_into(f) block.write_into(f)
sub_filenames.append(sub_filename) files_to_delete.append(temp_file)
sub_filenames.append(temp_file)
else:
sub_filenames.append(original_file)
elif ext == 'webm': elif ext == 'webm':
if not webm_vtt_warn: if not webm_vtt_warn:
webm_vtt_warn = True webm_vtt_warn = True
@ -653,7 +659,10 @@ def run(self, info):
else: else:
sub_langs.append(lang) sub_langs.append(lang)
sub_names.append(sub_info.get('name')) sub_names.append(sub_info.get('name'))
sub_filenames.append(sub_info['filepath']) filepath = sub_info['filepath']
sub_filenames.append(filepath)
if not self._already_have_subtitle:
files_to_delete.append(filepath)
if not mp4_ass_warn and ext == 'mp4' and sub_ext == 'ass': if not mp4_ass_warn and ext == 'mp4' and sub_ext == 'ass':
mp4_ass_warn = True mp4_ass_warn = True
self.report_warning('ASS subtitles cannot be properly embedded in mp4 files; expect issues') self.report_warning('ASS subtitles cannot be properly embedded in mp4 files; expect issues')
@ -682,7 +691,6 @@ def run(self, info):
self.run_ffmpeg_multiple_files(input_files, temp_filename, opts) self.run_ffmpeg_multiple_files(input_files, temp_filename, opts)
os.replace(temp_filename, filename) os.replace(temp_filename, filename)
files_to_delete = [] if self._already_have_subtitle else sub_filenames
return files_to_delete, info return files_to_delete, info