mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-13 14:44:29 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

add video subtitles and description to full-text index

Browse source
This commit is contained in:
Nick Sweeting 2021-04-10 07:22:20 -04:00
parent 1e3c1e3c83
commit 193df5c8d3
3 changed files with 15 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

1
archivebox/config.py
View file

@ -126,6 +126,7 @@ CONFIG_SCHEMA: Dict[str, ConfigDefaultDict] = {
'--write-sub', '--write-sub',
'--all-subs', '--all-subs',
'--write-auto-sub', '--write-auto-sub',
'--convert-subs=srt',
'--yes-playlist', '--yes-playlist',
'--continue', '--continue',
'--ignore-errors', '--ignore-errors',

13
archivebox/extractors/media.py
View file

@ -70,11 +70,24 @@ def save_media(link: Link, out_dir: Optional[Path]=None, timeout: int=MEDIA_TIME
finally: finally:
timer.end() timer.end()
# add video description and subtitles to full-text index
index_texts = [
text_file.read_text(encoding='utf-8').strip()
for text_file in (
*output_path.glob('*.description'),
*output_path.glob('*.srt'),
*output_path.glob('*.vtt'),
*output_path.glob('*.lrc'),
*output_path.glob('*.lrc'),
)
]
return ArchiveResult( return ArchiveResult(
cmd=cmd, cmd=cmd,
pwd=str(out_dir), pwd=str(out_dir),
cmd_version=YOUTUBEDL_VERSION, cmd_version=YOUTUBEDL_VERSION,
output=output, output=output,
status=status, status=status,
index_texts=index_texts,
**timer.stats, **timer.stats,
) )