[yt-dlp.git] / yt_dlp / extractor / linkedin.py

# coding: utf-8
from __future__ import unicode_literals

from itertools import zip_longest
import re

from .common import InfoExtractor
from ..utils import (
    clean_html,
    extract_attributes,
    ExtractorError,
    float_or_none,
    get_element_by_class,
    int_or_none,
    srt_subtitles_timecode,
    strip_or_none,
    mimetype2ext,
    try_get,
    urlencode_postdata,
    urljoin,
)


class LinkedInBaseIE(InfoExtractor):
    _NETRC_MACHINE = 'linkedin'
    _logged_in = False

    def _real_initialize(self):
        if self._logged_in:
            return
        email, password = self._get_login_info()
        if email is None:
            return

        login_page = self._download_webpage(
            self._LOGIN_URL, None, 'Downloading login page')
        action_url = urljoin(self._LOGIN_URL, self._search_regex(
            r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page, 'post url',
            default='https://www.linkedin.com/uas/login-submit', group='url'))
        data = self._hidden_inputs(login_page)
        data.update({
            'session_key': email,
            'session_password': password,
        })
        login_submit_page = self._download_webpage(
            action_url, None, 'Logging in',
            data=urlencode_postdata(data))
        error = self._search_regex(
            r'<span[^>]+class="error"[^>]*>\s*(.+?)\s*</span>',
            login_submit_page, 'error', default=None)
        if error:
            raise ExtractorError(error, expected=True)
        LinkedInBaseIE._logged_in = True


class LinkedInLearningBaseIE(LinkedInBaseIE):
    _LOGIN_URL = 'https://www.linkedin.com/uas/login?trk=learning'

    def _call_api(self, course_slug, fields, video_slug=None, resolution=None):
        query = {
            'courseSlug': course_slug,
            'fields': fields,
            'q': 'slugs',
        }
        sub = ''
        if video_slug:
            query.update({
                'videoSlug': video_slug,
                'resolution': '_%s' % resolution,
            })
            sub = ' %dp' % resolution
        api_url = 'https://www.linkedin.com/learning-api/detailedCourses'
        if not self._get_cookies(api_url).get('JSESSIONID'):
            self.raise_login_required()
        return self._download_json(
            api_url, video_slug, 'Downloading%s JSON metadata' % sub, headers={
                'Csrf-Token': self._get_cookies(api_url)['JSESSIONID'].value,
            }, query=query)['elements'][0]

    def _get_urn_id(self, video_data):
        urn = video_data.get('urn')
        if urn:
            mobj = re.search(r'urn:li:lyndaCourse:\d+,(\d+)', urn)
            if mobj:
                return mobj.group(1)

    def _get_video_id(self, video_data, course_slug, video_slug):
        return self._get_urn_id(video_data) or '%s/%s' % (course_slug, video_slug)


class LinkedInIE(LinkedInBaseIE):
    _VALID_URL = r'https?://(?:www\.)?linkedin\.com/posts/.+?(?P<id>\d+)'
    _TESTS = [{
        'url': 'https://www.linkedin.com/posts/mishalkhawaja_sendinblueviews-toronto-digitalmarketing-ugcPost-6850898786781339649-mM20',
        'info_dict': {
            'id': '6850898786781339649',
            'ext': 'mp4',
            'title': 'Mishal K. on LinkedIn: #sendinblueviews #toronto #digitalmarketing',
            'description': 'md5:be125430bab1c574f16aeb186a4d5b19',
            'creator': 'Mishal K.'
        },
    }]

    def _real_extract(self, url):
        video_id = self._match_id(url)
        webpage = self._download_webpage(url, video_id)

        title = self._html_search_regex(r'<title>([^<]+)</title>', webpage, 'title')
        description = clean_html(get_element_by_class('share-update-card__update-text', webpage))
        like_count = int_or_none(get_element_by_class('social-counts-reactions__social-counts-numRections', webpage))
        creator = strip_or_none(clean_html(get_element_by_class('comment__actor-name', webpage)))

        sources = self._parse_json(extract_attributes(self._search_regex(r'(<video[^>]+>)', webpage, 'video'))['data-sources'], video_id)
        formats = [{
            'url': source['src'],
            'ext': mimetype2ext(source.get('type')),
            'tbr': float_or_none(source.get('data-bitrate'), scale=1000),
        } for source in sources]

        self._sort_formats(formats)

        return {
            'id': video_id,
            'formats': formats,
            'title': title,
            'like_count': like_count,
            'creator': creator,
            'thumbnail': self._og_search_thumbnail(webpage),
            'description': description,
        }


class LinkedInLearningIE(LinkedInLearningBaseIE):
    IE_NAME = 'linkedin:learning'
    _VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<course_slug>[^/]+)/(?P<id>[^/?#]+)'
    _TEST = {
        'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals/welcome?autoplay=true',
        'md5': 'a1d74422ff0d5e66a792deb996693167',
        'info_dict': {
            'id': '90426',
            'ext': 'mp4',
            'title': 'Welcome',
            'timestamp': 1430396150.82,
            'upload_date': '20150430',
        },
    }

    def json2srt(self, transcript_lines, duration=None):
        srt_data = ''
        for line, (line_dict, next_dict) in enumerate(zip_longest(transcript_lines, transcript_lines[1:])):
            start_time, caption = line_dict['transcriptStartAt'] / 1000, line_dict['caption']
            end_time = next_dict['transcriptStartAt'] / 1000 if next_dict else duration or start_time + 1
            srt_data += '%d\n%s --> %s\n%s\n\n' % (line + 1, srt_subtitles_timecode(start_time),
                                                   srt_subtitles_timecode(end_time),
                                                   caption)
        return srt_data

    def _real_extract(self, url):
        course_slug, video_slug = self._match_valid_url(url).groups()

        formats = []
        for width, height in ((640, 360), (960, 540), (1280, 720)):
            video_data = self._call_api(
                course_slug, 'selectedVideo', video_slug, height)['selectedVideo']

            video_url_data = video_data.get('url') or {}
            progressive_url = video_url_data.get('progressiveUrl')
            if progressive_url:
                formats.append({
                    'format_id': 'progressive-%dp' % height,
                    'url': progressive_url,
                    'ext': 'mp4',
                    'height': height,
                    'width': width,
                    'source_preference': 1,
                })

        title = video_data['title']

        audio_url = video_data.get('audio', {}).get('progressiveUrl')
        if audio_url:
            formats.append({
                'abr': 64,
                'ext': 'm4a',
                'format_id': 'audio',
                'url': audio_url,
                'vcodec': 'none',
            })

        streaming_url = video_url_data.get('streamingUrl')
        if streaming_url:
            formats.extend(self._extract_m3u8_formats(
                streaming_url, video_slug, 'mp4',
                'm3u8_native', m3u8_id='hls', fatal=False))

        # It seems like this would be correctly handled by default
        # However, unless someone can confirm this, the old
        # behaviour is being kept as-is
        self._sort_formats(formats, ('res', 'source_preference'))
        subtitles = {}
        duration = int_or_none(video_data.get('durationInSeconds'))
        transcript_lines = try_get(video_data, lambda x: x['transcript']['lines'], expected_type=list)
        if transcript_lines:
            subtitles['en'] = [{
                'ext': 'srt',
                'data': self.json2srt(transcript_lines, duration)
            }]

        return {
            'id': self._get_video_id(video_data, course_slug, video_slug),
            'title': title,
            'formats': formats,
            'thumbnail': video_data.get('defaultThumbnail'),
            'timestamp': float_or_none(video_data.get('publishedOn'), 1000),
            'duration': duration,
            'subtitles': subtitles,
        }


class LinkedInLearningCourseIE(LinkedInLearningBaseIE):
    IE_NAME = 'linkedin:learning:course'
    _VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<id>[^/?#]+)'
    _TEST = {
        'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals',
        'info_dict': {
            'id': 'programming-foundations-fundamentals',
            'title': 'Programming Foundations: Fundamentals',
            'description': 'md5:76e580b017694eb89dc8e8923fff5c86',
        },
        'playlist_mincount': 61,
    }

    @classmethod
    def suitable(cls, url):
        return False if LinkedInLearningIE.suitable(url) else super(LinkedInLearningCourseIE, cls).suitable(url)

    def _real_extract(self, url):
        course_slug = self._match_id(url)
        course_data = self._call_api(course_slug, 'chapters,description,title')

        entries = []
        for chapter_number, chapter in enumerate(course_data.get('chapters', []), 1):
            chapter_title = chapter.get('title')
            chapter_id = self._get_urn_id(chapter)
            for video in chapter.get('videos', []):
                video_slug = video.get('slug')
                if not video_slug:
                    continue
                entries.append({
                    '_type': 'url_transparent',
                    'id': self._get_video_id(video, course_slug, video_slug),
                    'title': video.get('title'),
                    'url': 'https://www.linkedin.com/learning/%s/%s' % (course_slug, video_slug),
                    'chapter': chapter_title,
                    'chapter_number': chapter_number,
                    'chapter_id': chapter_id,
                    'ie_key': LinkedInLearningIE.ie_key(),
                })

        return self.playlist_result(
            entries, course_slug,
            course_data.get('title'),
            course_data.get('description'))
Commit	Line	Data
aa7e974a RA	1	# coding: utf-8
	2	from __future__ import unicode_literals
	3
8dc831f7	4	from itertools import zip_longest
aa7e974a RA	5	import re
	6
	7	from .common import InfoExtractor
	8	from ..utils import (
402cd603	9	clean_html,
402cd603	10	extract_attributes,
aa7e974a RA	11	ExtractorError,
aa7e974a RA	12	float_or_none,
402cd603	13	get_element_by_class,
aa7e974a	14	int_or_none,
8dc831f7	15	srt_subtitles_timecode,
402cd603	16	strip_or_none,
402cd603	17	mimetype2ext,
8dc831f7	18	try_get,
aa7e974a	19	urlencode_postdata,
c9120294	20	urljoin,
aa7e974a RA	21	)
	22
	23
402cd603	24	class LinkedInBaseIE(InfoExtractor):
aa7e974a	25	_NETRC_MACHINE = 'linkedin'
da483200	26	_logged_in = False
aa7e974a	27
402cd603	28	def _real_initialize(self):
	29	if self._logged_in:
	30	return
	31	email, password = self._get_login_info()
	32	if email is None:
	33	return
	34
	35	login_page = self._download_webpage(
	36	self._LOGIN_URL, None, 'Downloading login page')
	37	action_url = urljoin(self._LOGIN_URL, self._search_regex(
	38	r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page, 'post url',
	39	default='https://www.linkedin.com/uas/login-submit', group='url'))
	40	data = self._hidden_inputs(login_page)
	41	data.update({
	42	'session_key': email,
	43	'session_password': password,
	44	})
	45	login_submit_page = self._download_webpage(
	46	action_url, None, 'Logging in',
	47	data=urlencode_postdata(data))
	48	error = self._search_regex(
	49	r'<span[^>]+class="error"[^>]>\s(.+?)\s*</span>',
	50	login_submit_page, 'error', default=None)
	51	if error:
	52	raise ExtractorError(error, expected=True)
	53	LinkedInBaseIE._logged_in = True
	54
	55
	56	class LinkedInLearningBaseIE(LinkedInBaseIE):
	57	_LOGIN_URL = 'https://www.linkedin.com/uas/login?trk=learning'
	58
aa7e974a RA	59	def _call_api(self, course_slug, fields, video_slug=None, resolution=None):
	60	query = {
	61	'courseSlug': course_slug,
	62	'fields': fields,
	63	'q': 'slugs',
	64	}
	65	sub = ''
	66	if video_slug:
	67	query.update({
	68	'videoSlug': video_slug,
	69	'resolution': '_%s' % resolution,
	70	})
	71	sub = ' %dp' % resolution
	72	api_url = 'https://www.linkedin.com/learning-api/detailedCourses'
da483200	73	if not self._get_cookies(api_url).get('JSESSIONID'):
da483200	74	self.raise_login_required()
aa7e974a RA	75	return self._download_json(
	76	api_url, video_slug, 'Downloading%s JSON metadata' % sub, headers={
	77	'Csrf-Token': self._get_cookies(api_url)['JSESSIONID'].value,
	78	}, query=query)['elements'][0]
	79
91effe22 RA	80	def _get_urn_id(self, video_data):
91effe22 RA	81	urn = video_data.get('urn')
aa7e974a RA	82	if urn:
	83	mobj = re.search(r'urn:li:lyndaCourse:\d+,(\d+)', urn)
	84	if mobj:
	85	return mobj.group(1)
91effe22 RA	86
	87	def _get_video_id(self, video_data, course_slug, video_slug):
	88	return self._get_urn_id(video_data) or '%s/%s' % (course_slug, video_slug)
aa7e974a	89
aa7e974a	90
402cd603	91	class LinkedInIE(LinkedInBaseIE):
	92	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/posts/.+?(?P<id>\d+)'
	93	_TESTS = [{
	94	'url': 'https://www.linkedin.com/posts/mishalkhawaja_sendinblueviews-toronto-digitalmarketing-ugcPost-6850898786781339649-mM20',
	95	'info_dict': {
	96	'id': '6850898786781339649',
	97	'ext': 'mp4',
	98	'title': 'Mishal K. on LinkedIn: #sendinblueviews #toronto #digitalmarketing',
	99	'description': 'md5:be125430bab1c574f16aeb186a4d5b19',
	100	'creator': 'Mishal K.'
	101	},
	102	}]
	103
	104	def _real_extract(self, url):
	105	video_id = self._match_id(url)
	106	webpage = self._download_webpage(url, video_id)
	107
	108	title = self._html_search_regex(r'<title>([^<]+)</title>', webpage, 'title')
	109	description = clean_html(get_element_by_class('share-update-card__update-text', webpage))
	110	like_count = int_or_none(get_element_by_class('social-counts-reactions__social-counts-numRections', webpage))
	111	creator = strip_or_none(clean_html(get_element_by_class('comment__actor-name', webpage)))
9222c381	112
402cd603	113	sources = self._parse_json(extract_attributes(self._search_regex(r'(<video[^>]+>)', webpage, 'video'))['data-sources'], video_id)
	114	formats = [{
	115	'url': source['src'],
	116	'ext': mimetype2ext(source.get('type')),
	117	'tbr': float_or_none(source.get('data-bitrate'), scale=1000),
	118	} for source in sources]
	119
	120	self._sort_formats(formats)
	121
	122	return {
	123	'id': video_id,
	124	'formats': formats,
	125	'title': title,
	126	'like_count': like_count,
	127	'creator': creator,
	128	'thumbnail': self._og_search_thumbnail(webpage),
	129	'description': description,
	130	}
aa7e974a RA	131
	132
	133	class LinkedInLearningIE(LinkedInLearningBaseIE):
	134	IE_NAME = 'linkedin:learning'
	135	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<course_slug>[^/]+)/(?P<id>[^/?#]+)'
	136	_TEST = {
	137	'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals/welcome?autoplay=true',
	138	'md5': 'a1d74422ff0d5e66a792deb996693167',
	139	'info_dict': {
	140	'id': '90426',
	141	'ext': 'mp4',
	142	'title': 'Welcome',
	143	'timestamp': 1430396150.82,
	144	'upload_date': '20150430',
	145	},
	146	}
	147
8dc831f7 AG	148	def json2srt(self, transcript_lines, duration=None):
	149	srt_data = ''
	150	for line, (line_dict, next_dict) in enumerate(zip_longest(transcript_lines, transcript_lines[1:])):
	151	start_time, caption = line_dict['transcriptStartAt'] / 1000, line_dict['caption']
	152	end_time = next_dict['transcriptStartAt'] / 1000 if next_dict else duration or start_time + 1
80c360d7 AG	153	srt_data += '%d\n%s --> %s\n%s\n\n' % (line + 1, srt_subtitles_timecode(start_time),
	154	srt_subtitles_timecode(end_time),
	155	caption)
8dc831f7 AG	156	return srt_data
8dc831f7 AG	157
aa7e974a	158	def _real_extract(self, url):
5ad28e7f	159	course_slug, video_slug = self._match_valid_url(url).groups()
aa7e974a	160
aa7e974a RA	161	formats = []
	162	for width, height in ((640, 360), (960, 540), (1280, 720)):
	163	video_data = self._call_api(
	164	course_slug, 'selectedVideo', video_slug, height)['selectedVideo']
	165
	166	video_url_data = video_data.get('url') or {}
	167	progressive_url = video_url_data.get('progressiveUrl')
	168	if progressive_url:
	169	formats.append({
	170	'format_id': 'progressive-%dp' % height,
	171	'url': progressive_url,
8dc831f7	172	'ext': 'mp4',
aa7e974a RA	173	'height': height,
	174	'width': width,
	175	'source_preference': 1,
	176	})
	177
	178	title = video_data['title']
	179
	180	audio_url = video_data.get('audio', {}).get('progressiveUrl')
	181	if audio_url:
	182	formats.append({
	183	'abr': 64,
	184	'ext': 'm4a',
	185	'format_id': 'audio',
	186	'url': audio_url,
	187	'vcodec': 'none',
	188	})
	189
	190	streaming_url = video_url_data.get('streamingUrl')
	191	if streaming_url:
	192	formats.extend(self._extract_m3u8_formats(
	193	streaming_url, video_slug, 'mp4',
	194	'm3u8_native', m3u8_id='hls', fatal=False))
	195
54f37eea	196	# It seems like this would be correctly handled by default
	197	# However, unless someone can confirm this, the old
	198	# behaviour is being kept as-is
	199	self._sort_formats(formats, ('res', 'source_preference'))
8dc831f7 AG	200	subtitles = {}
	201	duration = int_or_none(video_data.get('durationInSeconds'))
	202	transcript_lines = try_get(video_data, lambda x: x['transcript']['lines'], expected_type=list)
	203	if transcript_lines:
	204	subtitles['en'] = [{
	205	'ext': 'srt',
	206	'data': self.json2srt(transcript_lines, duration)
	207	}]
aa7e974a RA	208
aa7e974a RA	209	return {
91effe22	210	'id': self._get_video_id(video_data, course_slug, video_slug),
aa7e974a RA	211	'title': title,
	212	'formats': formats,
	213	'thumbnail': video_data.get('defaultThumbnail'),
	214	'timestamp': float_or_none(video_data.get('publishedOn'), 1000),
8dc831f7 AG	215	'duration': duration,
8dc831f7 AG	216	'subtitles': subtitles,
aa7e974a RA	217	}
	218
	219
	220	class LinkedInLearningCourseIE(LinkedInLearningBaseIE):
	221	IE_NAME = 'linkedin:learning:course'
	222	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<id>[^/?#]+)'
	223	_TEST = {
	224	'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals',
	225	'info_dict': {
	226	'id': 'programming-foundations-fundamentals',
	227	'title': 'Programming Foundations: Fundamentals',
	228	'description': 'md5:76e580b017694eb89dc8e8923fff5c86',
	229	},
	230	'playlist_mincount': 61,
	231	}
	232
	233	@classmethod
	234	def suitable(cls, url):
	235	return False if LinkedInLearningIE.suitable(url) else super(LinkedInLearningCourseIE, cls).suitable(url)
	236
	237	def _real_extract(self, url):
	238	course_slug = self._match_id(url)
	239	course_data = self._call_api(course_slug, 'chapters,description,title')
	240
	241	entries = []
91effe22	242	for chapter_number, chapter in enumerate(course_data.get('chapters', []), 1):
aa7e974a	243	chapter_title = chapter.get('title')
91effe22	244	chapter_id = self._get_urn_id(chapter)
aa7e974a RA	245	for video in chapter.get('videos', []):
	246	video_slug = video.get('slug')
	247	if not video_slug:
	248	continue
	249	entries.append({
b1447572	250	'_type': 'url_transparent',
91effe22	251	'id': self._get_video_id(video, course_slug, video_slug),
aa7e974a RA	252	'title': video.get('title'),
	253	'url': 'https://www.linkedin.com/learning/%s/%s' % (course_slug, video_slug),
	254	'chapter': chapter_title,
91effe22 RA	255	'chapter_number': chapter_number,
91effe22 RA	256	'chapter_id': chapter_id,
aa7e974a RA	257	'ie_key': LinkedInLearningIE.ie_key(),
	258	})
	259
	260	return self.playlist_result(
	261	entries, course_slug,
	262	course_data.get('title'),
	263	course_data.get('description'))