Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

refactor get_subtitle_sources #136

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
Open

refactor get_subtitle_sources #136

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
2897105
refactor get_subtitle_sources
zrose584 May 21, 2022
6c5d1b2
captions and plyr: choose correct default track
zrose584 May 22, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
98 changes: 43 additions & 55 deletions youtube/watch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -168,7 +168,7 @@ def video_rank(src):



def make_caption_src(info, lang, auto=False, trans_lang=None):
def make_caption_src(info, lang, score, auto=False, trans_lang=None):
label = lang
if auto:
label += ' (Automatic)'
Expand All @@ -179,6 +179,9 @@ def make_caption_src(info, lang, auto=False, trans_lang=None):
'label': label,
'srclang': trans_lang[0:2] if trans_lang else lang[0:2],
'on': False,
'score': score,
'lang': lang,
'auto': auto,
}

def lang_in(lang, sequence):
Expand All @@ -195,83 +198,68 @@ def lang_eq(lang1, lang2):
return False
return lang1[0:2] == lang2[0:2]

def equiv_lang_in(lang, sequence):
'''Extracts a language in sequence which is equivalent to lang.
e.g. if lang is en, extracts en-GB from sequence.
Necessary because if only a specific variant like en-GB is available, can't ask Youtube for simply en. Need to get the available variant.'''
lang = lang[0:2]
for l in sequence:
if l[0:2] == lang:
return l
return None

def get_subtitle_sources(info):
'''Returns these sources, ordered from least to most intelligible:
native_video_lang (Automatic)
foreign_langs (Manual)

native_video_lang (Automatic) -> pref_lang
foreign_langs (Manual) -> pref_lang
native_video_lang (Manual) -> pref_lang

pref_lang (Automatic)
pref_lang (Manual)'''
sources = []
if not yt_data_extract.captions_available(info):
return []
pref_lang = settings.subtitles_language

user_langs = settings.subtitles_language.split(',')
native_video_lang = None
if info['automatic_caption_languages']:
native_video_lang = info['automatic_caption_languages'][0]

highest_fidelity_is_manual = False
kManual, kNativeVideoLang, kUserSpeaksIt = 1, 2, 4
Copy link
Owner

@user234683 user234683 Jul 30, 2022 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Shouldn't this be kNativeVideoLang, kManual, kUserSpeaksIt = 1, 2, 4?

Manually created subtitles in a different language than the video are always going to be higher quality than automatically generated ones. Comparing the current scores to the comment at the top of the code with the ideal ordering:

010    native_video_lang (Automatic)
001    foreign_langs (Manual)

110    native_video_lang (Automatic) -> pref_lang
101    foreign_langs (Manual) -> pref_lang
111    native_video_lang (Manual) -> pref_lang

110   pref_lang (Automatic)
111   pref_lang (Manual) (101 if pref_lang != native_video_lang)'''

Swapping the last two bits will fix it

Additionally, these constants should be changed to NATIVE_VIDEO_LANG, MANUAL, etc. to match the convention in the rest of the code that constants are indicated with uppercase snakecase; we dont use hungarian notation in the code.

def score_lang(lang):
score = 0
if lang_in(lang, user_langs):
score |= kUserSpeaksIt
if lang_eq(lang, native_video_lang):
score |= kNativeVideoLang
return score

# Sources are added in very specific order outlined above
# More intelligible sources are put further down to avoid browser bug when there are too many languages
# (in firefox, it is impossible to select a language near the top of the list because it is cut off)
for lang in info['manual_caption_languages']:
score = score_lang(lang) | kManual
sources.append(make_caption_src(info, lang, score))

# native_video_lang (Automatic)
if native_video_lang and not lang_eq(native_video_lang, pref_lang):
sources.append(make_caption_src(info, native_video_lang, auto=True))
for lang in info['automatic_caption_languages']:
score = score_lang(lang)
if score:
sources.append(make_caption_src(info, lang, score, auto=True))

# foreign_langs (Manual)
for lang in info['manual_caption_languages']:
if not lang_eq(lang, pref_lang):
sources.append(make_caption_src(info, lang))

if (lang_in(pref_lang, info['translation_languages'])
and not lang_in(pref_lang, info['automatic_caption_languages'])
and not lang_in(pref_lang, info['manual_caption_languages'])):
# native_video_lang (Automatic) -> pref_lang
if native_video_lang and not lang_eq(pref_lang, native_video_lang):
sources.append(make_caption_src(info, native_video_lang, auto=True, trans_lang=pref_lang))

# foreign_langs (Manual) -> pref_lang
for lang in info['manual_caption_languages']:
if not lang_eq(lang, native_video_lang) and not lang_eq(lang, pref_lang):
sources.append(make_caption_src(info, lang, trans_lang=pref_lang))

# native_video_lang (Manual) -> pref_lang
if lang_in(native_video_lang, info['manual_caption_languages']):
sources.append(make_caption_src(info, native_video_lang, trans_lang=pref_lang))

# pref_lang (Automatic)
if lang_in(pref_lang, info['automatic_caption_languages']):
sources.append(make_caption_src(info, equiv_lang_in(pref_lang, info['automatic_caption_languages']), auto=True))

# pref_lang (Manual)
if lang_in(pref_lang, info['manual_caption_languages']):
sources.append(make_caption_src(info, equiv_lang_in(pref_lang, info['manual_caption_languages'])))
highest_fidelity_is_manual = True

if sources and sources[-1]['srclang'] == pref_lang:
# set as on by default since it's manual a default-on subtitles mode is in settings
if highest_fidelity_is_manual and settings.subtitles_mode > 0:
if not any(s['score'] & kUserSpeaksIt for s in sources):
for user_lang in user_langs:
if not lang in info['translation_languages']:
continue

for s in sources[:]:
lang, score, auto = s['lang'], s['score'] | kUserSpeaksIt, s['auto']
sources.append(make_caption_src(info, lang, score, auto=auto, trans_lang=user_lang))

sources = sorted(sources, key=lambda x: x['score'])

if len(sources) == 0:
assert len(info['automatic_caption_languages']) == 0 and len(info['manual_caption_languages']) == 0

if sources and settings.subtitles_mode > 0 and sources[-1]['score'] & kUserSpeaksIt:
if sources[-1]['score'] & kManual:
# set as on by default since it's manual a default-on subtitles mode is in settings
sources[-1]['on'] = True
# set as on by default since settings indicate to set it as such even if it's not manual
elif settings.subtitles_mode == 2:
# set as on by default since settings indicate to set it as such even if it's not manual
sources[-1]['on'] = True

if len(sources) == 0:
assert len(info['automatic_caption_languages']) == 0 and len(info['manual_caption_languages']) == 0
if settings.video_player == 1: # Plyr
sources[-1]['on'] = True
Copy link
Owner

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why are we always turning captions on for Plyr only?


return sources

Expand Down