administration/yt-dlp
1
0
Fork 0
mirror of https://github.com/yt-dlp/yt-dlp.git synced 2024-11-12 20:23:11 +00:00
Code Issues Projects Releases Wiki Activity

[PromoDJ] Fix paid music metadata

Browse Source
This commit is contained in:
DmitryScaletta 2024-02-15 12:34:50 +03:00
parent c837d90e12
commit e6f3e6de0e
No known key found for this signature in database
GPG Key ID: 167A65222EDD4C2A
1 changed files with 3 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
yt_dlp/extractor/promodj.py
View File

@ -769,7 +769,6 @@ class PromoDJIE(PromoDJBaseIE):
}, },
}] }]
_IS_PAID_RE = r'<b>Цена:</b>'
# examples: MP3, 320 Кбит | MP4, 20157 Кбит | WAV, 1412 Кбит | AVI, 1731 Кбит | ASF, 6905 Кбит | FLAC, 1509 Кбит # examples: MP3, 320 Кбит | MP4, 20157 Кбит | WAV, 1412 Кбит | AVI, 1731 Кбит | ASF, 6905 Кбит | FLAC, 1509 Кбит
# https://regex101.com/r/2AuaxB/1 # https://regex101.com/r/2AuaxB/1
_FORMATS_RE = r'(?:<a\s+href=\"(?P<url>[^\"]+)\">)?\s*\w+, (?P<bitrate>\d+) Кбит' _FORMATS_RE = r'(?:<a\s+href=\"(?P<url>[^\"]+)\">)?\s*\w+, (?P<bitrate>\d+) Кбит'
@ -835,7 +834,7 @@ class PromoDJIE(PromoDJBaseIE):
# download links can be missing # download links can be missing
# best quality format always comes first # best quality format always comes first
formats_from_html = re.findall(self._FORMATS_RE, meta_html) formats_from_html = re.findall(self._FORMATS_RE, meta_html)
is_paid = re.search(self._IS_PAID_RE, meta_html) is_paid = '<b>Цена:</b>' in meta_html
# size field describes best quality # size field describes best quality
size = self._parse_ru_size(*re.search(self._SIZE_RE, meta_html).groups()) size = self._parse_ru_size(*re.search(self._SIZE_RE, meta_html).groups())
if type == 'videos': if type == 'videos':
@ -848,13 +847,13 @@ class PromoDJIE(PromoDJBaseIE):
'size': size, 'size': size,
'quality': 1, 'quality': 1,
}) })
else: elif not is_paid:
for i, match in enumerate(formats_from_html): for i, match in enumerate(formats_from_html):
url, bitrate = match url, bitrate = match
is_last = i == len(formats_from_html) - 1 is_last = i == len(formats_from_html) - 1
if is_last: if is_last:
metadata['formats'][0]['abr'] = int(bitrate) metadata['formats'][0]['abr'] = int(bitrate)
elif url_or_none(url) and not is_paid: elif url_or_none(url):
metadata['formats'].append({ metadata['formats'].append({
'url': url, 'url': url,
'abr': int(bitrate), 'abr': int(bitrate),