yt-dlp/youtube_dlc/extractor/storyfire.py

256 lines
9.1 KiB
Python
Raw Normal View History

2020-08-17 00:07:14 +00:00
# coding: utf-8
from __future__ import unicode_literals
import itertools
from .common import InfoExtractor
class StoryFireIE(InfoExtractor):
_VALID_URL = r'(?:(?:https?://(?:www\.)?storyfire\.com/video-details)|(?:https://storyfire.app.link))/(?P<id>[^/\s]+)'
_TESTS = [{
'url': 'https://storyfire.com/video-details/5df1d132b6378700117f9181',
'md5': '560953bfca81a69003cfa5e53ac8a920',
'info_dict': {
'id': '5df1d132b6378700117f9181',
'ext': 'mp4',
'title': 'Buzzfeed Teaches You About Memes',
'uploader_id': 'ntZAJFECERSgqHSxzonV5K2E89s1',
'timestamp': 1576129028,
'description': 'Mocking Buzzfeed\'s meme lesson. Reuploaded from YouTube because of their new policies',
'uploader': 'whang!',
'upload_date': '20191212',
},
'params': {'format': 'bestvideo'} # There are no merged formats in the playlist.
}, {
'url': 'https://storyfire.app.link/5GxAvWOQr8', # Alternate URL format, with unrelated short ID
'md5': '7a2dc6d60c4889edfed459c620fe690d',
'info_dict': {
'id': '5f1e11ecd78a57b6c702001d',
'ext': 'm4a',
'title': 'Weird Nintendo Prototype Leaks',
'description': 'A stream taking a look at some weird Nintendo Prototypes with Luigi in Mario 64 and weird Yoshis',
'timestamp': 1595808576,
'upload_date': '20200727',
'uploader': 'whang!',
'uploader_id': 'ntZAJFECERSgqHSxzonV5K2E89s1',
},
'params': {'format': 'bestaudio'} # Verifying audio extraction
}]
_aformats = {
'audio-medium-audio': {'acodec': 'aac', 'abr': 125, 'preference': -10},
'audio-high-audio': {'acodec': 'aac', 'abr': 254, 'preference': -1},
}
def _real_extract(self, url):
video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id)
# Extracting the json blob is mandatory to proceed with extraction.
jsontext = self._html_search_regex(
r'<script id="__NEXT_DATA__" type="application/json">(.+?)</script>',
webpage, 'json_data')
json = self._parse_json(jsontext, video_id)
# The currentVideo field in the json is mandatory
# because it contains the only link to the m3u playlist
video = json['props']['initialState']['video']['currentVideo']
videourl = video['vimeoVideoURL'] # Video URL is mandatory
# Extract other fields from the json in an error tolerant fashion
# ID may be incorrect (on short URL format), correct it.
parsed_id = video.get('_id')
if parsed_id:
video_id = parsed_id
title = video.get('title')
description = video.get('description')
thumbnail = video.get('storyImage')
views = video.get('views')
likes = video.get('likesCount')
comments = video.get('commentsCount')
duration = video.get('videoDuration')
publishdate = video.get('publishDate') # Apparently epoch time, day only
uploader = video.get('username')
uploader_id = video.get('hostID')
# Construct an uploader URL
uploader_url = None
if uploader_id:
uploader_url = "https://storyfire.com/user/%s/video" % uploader_id
# Collect root playlist to determine formats
formats = self._extract_m3u8_formats(
videourl, video_id, 'mp4', 'm3u8_native')
# Modify formats to fill in missing information about audio codecs
for format in formats:
aformat = self._aformats.get(format['format_id'])
if aformat:
format['acodec'] = aformat['acodec']
format['abr'] = aformat['abr']
format['preference'] = aformat['preference']
format['ext'] = 'm4a'
self._sort_formats(formats)
return {
'id': video_id,
'title': title,
'description': description,
'ext': "mp4",
'url': videourl,
'formats': formats,
'thumbnail': thumbnail,
'view_count': views,
'like_count': likes,
'comment_count': comments,
'duration': duration,
'timestamp': publishdate,
'uploader': uploader,
'uploader_id': uploader_id,
'uploader_url': uploader_url,
}
class StoryFireUserIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?storyfire\.com/user/(?P<id>[^/\s]+)/video'
_TESTS = [{
'url': 'https://storyfire.com/user/ntZAJFECERSgqHSxzonV5K2E89s1/video',
'info_dict': {
'id': 'ntZAJFECERSgqHSxzonV5K2E89s1',
'title': 'whang!',
},
'playlist_mincount': 18
}, {
'url': 'https://storyfire.com/user/UQ986nFxmAWIgnkZQ0ftVhq4nOk2/video',
'info_dict': {
'id': 'UQ986nFxmAWIgnkZQ0ftVhq4nOk2',
'title': 'McJuggerNuggets',
},
'playlist_mincount': 143
}]
# Generator for fetching playlist items
def _enum_videos(self, baseurl, user_id, firstjson):
totalVideos = int(firstjson['videosCount'])
haveVideos = 0
json = firstjson
for page in itertools.count(1):
for video in json['videos']:
id = video['_id']
url = "https://storyfire.com/video-details/%s" % id
haveVideos += 1
yield {
'_type': 'url',
'id': id,
'url': url,
'ie_key': 'StoryFire',
'title': video.get('title'),
'description': video.get('description'),
'view_count': video.get('views'),
'comment_count': video.get('commentsCount'),
'duration': video.get('videoDuration'),
'timestamp': video.get('publishDate'),
}
# Are there more pages we could fetch?
if haveVideos < totalVideos:
pageurl = baseurl + ("%i" % haveVideos)
json = self._download_json(pageurl, user_id,
note='Downloading page %s' % page)
# Are there any videos in the new json?
videos = json.get('videos')
if not videos or len(videos) == 0:
break # no videos
else:
break # We have fetched all the videos, stop
def _real_extract(self, url):
user_id = self._match_id(url)
baseurl = "https://storyfire.com/app/publicVideos/%s?skip=" % user_id
# Download first page to ensure it can be downloaded, and get user information if available.
firstpage = baseurl + "0"
firstjson = self._download_json(firstpage, user_id)
title = None
videos = firstjson.get('videos')
if videos and len(videos):
title = videos[1].get('username')
return {
'_type': 'playlist',
'entries': self._enum_videos(baseurl, user_id, firstjson),
'id': user_id,
'title': title,
}
class StoryFireSeriesIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?storyfire\.com/write/series/stories/(?P<id>[^/\s]+)'
_TESTS = [{
'url': 'https://storyfire.com/write/series/stories/-Lq6MsuIHLODO6d2dDkr/',
'info_dict': {
'id': '-Lq6MsuIHLODO6d2dDkr',
},
'playlist_mincount': 13
}, {
'url': 'https://storyfire.com/write/series/stories/the_mortal_one/',
'info_dict': {
'id': 'the_mortal_one',
},
'playlist_count': 0 # This playlist has entries, but no videos.
}, {
'url': 'https://storyfire.com/write/series/stories/story_time',
'info_dict': {
'id': 'story_time',
},
'playlist_mincount': 10
}]
# Generator for returning playlist items
# This object is substantially different than the one in the user videos page above
def _enum_videos(self, jsonlist):
for video in jsonlist:
id = video['_id']
if video.get('hasVideo'): # Boolean element
url = "https://storyfire.com/video-details/%s" % id
yield {
'_type': 'url',
'id': id,
'url': url,
'ie_key': 'StoryFire',
'title': video.get('title'),
'description': video.get('description'),
'view_count': video.get('views'),
'likes_count': video.get('likesCount'),
'comment_count': video.get('commentsCount'),
'duration': video.get('videoDuration'),
'timestamp': video.get('publishDate'),
}
def _real_extract(self, url):
list_id = self._match_id(url)
listurl = "https://storyfire.com/app/seriesStories/%s/list" % list_id
json = self._download_json(listurl, list_id)
return {
'_type': 'playlist',
'entries': self._enum_videos(json),
'id': list_id
}