Lomanic/NewLeaf
1
0
Fork 0
mirror of https://git.sr.ht/~cadence/NewLeaf synced 2024-11-21 23:27:29 +00:00
Code Issues Releases Wiki Activity

Remove extraneous " align:start position:0%" on auto-generated captions

Browse Source
This commit is contained in:
Lomanic 2021-04-05 02:06:45 +02:00 committed by Cadence Ember
parent 1d52fca3a0
commit be8a2dad5f
No known key found for this signature in database
GPG Key ID: BC1C2C61CF521B17
1 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
extractors/captions.py
View File

@ -1,3 +1,4 @@
import re
import requests import requests
from extractors.video import extract_video from extractors.video import extract_video
from tools.converters import escape_html_textcontent, get_subtitle_api_url from tools.converters import escape_html_textcontent, get_subtitle_api_url
@ -20,6 +21,9 @@ def extract_captions_from_dict(captions, *, lang=None, label=None):
url = next(caption["second__remoteUrl"] for caption in captions["captions"] if caption["languageCode"] == lang or caption["label"] == label) url = next(caption["second__remoteUrl"] for caption in captions["captions"] if caption["languageCode"] == lang or caption["label"] == label)
with requests.get(url) as r: with requests.get(url) as r:
r.raise_for_status() r.raise_for_status()
# remove extraneous " align:start position:0%" on timestamps lines on auto-generated captions
if (lang and "auto-generated" in lang) or (label and "auto-generated" in label):
return re.sub(r"^([0-9:.]+ --> [0-9:.]+).*$", r"\1", r.content.decode("utf8"), flags=re.MULTILINE)
return r return r
# List of captions directly from youtube, but no automatic # List of captions directly from youtube, but no automatic