yt-dlp/yt_dlp/extractor/piapro.py
sepro add96eb9f8
[cleanup] Add more ruff rules (#10149)
Authored by: seproDev

Reviewed-by: bashonly <88596187+bashonly@users.noreply.github.com>
Reviewed-by: Simon Sawicki <contact@grub4k.xyz>
2024-06-12 01:09:58 +02:00

112 lines
4.2 KiB
Python
Raw Permalink Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

import urllib.parse
from .common import InfoExtractor
from ..utils import (
ExtractorError,
clean_html,
get_element_by_class,
parse_duration,
parse_filesize,
str_to_int,
unified_timestamp,
urlencode_postdata,
)
class PiaproIE(InfoExtractor):
_NETRC_MACHINE = 'piapro'
_VALID_URL = r'https?://piapro\.jp/(?:t|content)/(?P<id>[\w-]+)/?'
_TESTS = [{
'url': 'https://piapro.jp/t/NXYR',
'md5': 'f7c0f760913fb1d44a1c45a4af793909',
'info_dict': {
'id': 'NXYR',
'ext': 'mp3',
'uploader': 'wowaka',
'uploader_id': 'wowaka',
'title': '裏表ラバーズ',
'description': 'http://www.nicovideo.jp/watch/sm8082467',
'duration': 189.0,
'timestamp': 1251785475,
'thumbnail': r're:^https?://.*\.(?:png|jpg)$',
'upload_date': '20090901',
'view_count': int,
},
}, {
'note': 'There are break lines in description, mandating (?s) flag',
'url': 'https://piapro.jp/t/9cSd',
'md5': '952bb6d1e8de95050206408a87790676',
'info_dict': {
'id': '9cSd',
'ext': 'mp3',
'title': '青に溶けた風船 / 初音ミク',
'description': 'md5:d395a9bd151447631a5a1460bc7f9132',
'uploader': 'シアン・キノ',
'duration': 229.0,
'timestamp': 1644030039,
'upload_date': '20220205',
'view_count': int,
'thumbnail': r're:^https?://.*\.(?:png|jpg)$',
'uploader_id': 'cyankino',
},
}, {
'url': 'https://piapro.jp/content/hcw0z3a169wtemz6',
'only_matching': True,
}, {
'url': 'https://piapro.jp/t/-SO-',
'only_matching': True,
}]
_login_status = False
def _perform_login(self, username, password):
login_ok = True
login_form_strs = {
'_username': username,
'_password': password,
'_remember_me': 'on',
'login': 'ログイン',
}
self._request_webpage('https://piapro.jp/login/', None)
urlh = self._request_webpage(
'https://piapro.jp/login/exe', None,
note='Logging in', errnote='Unable to log in',
data=urlencode_postdata(login_form_strs))
if urlh is False:
login_ok = False
else:
parts = urllib.parse.urlparse(urlh.url)
if parts.path != '/':
login_ok = False
if not login_ok:
self.report_warning(
'unable to log in: bad username or password')
self._login_status = login_ok
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
category_id = self._search_regex(r'categoryId=(.+)">', webpage, 'category ID')
if category_id not in ('1', '2', '21', '22', '23', '24', '25'):
raise ExtractorError('The URL does not contain audio.', expected=True)
def extract_info(name, description):
return self._search_regex(rf'{name}[:]\s*([\d\s,:/]+)\s*</p>', webpage, description, default=None)
return {
'id': video_id,
'title': clean_html(get_element_by_class('contents_title', webpage)),
'description': clean_html(get_element_by_class('contents_description', webpage)),
'uploader': clean_html(get_element_by_class('contents_creator_txt', webpage)),
'uploader_id': self._search_regex(
r'<a\s+href="/([^"]+)"', get_element_by_class('contents_creator', webpage), 'uploader id', default=None),
'timestamp': unified_timestamp(extract_info('投稿日', 'timestamp'), False),
'duration': parse_duration(extract_info('長さ', 'duration')),
'view_count': str_to_int(extract_info('閲覧数', 'view count')),
'thumbnail': self._html_search_meta('twitter:image', webpage),
'filesize_approx': parse_filesize((extract_info('サイズ', 'size') or '').replace(',', '')),
'url': self._search_regex(r'\"url\":\s*\"(.*?)\"', webpage, 'url'),
'ext': 'mp3',
'vcodec': 'none',
}