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

import itertools
import re

from .common import InfoExtractor
from ..utils import (
    ExtractorError,
    extract_attributes,
    float_or_none,
    int_or_none,
    mimetype2ext,
    srt_subtitles_timecode,
    traverse_obj,
    try_get,
    url_or_none,
    urlencode_postdata,
    urljoin,
)


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

    def _perform_login(self, username, password):
        if self._logged_in:
            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': username,
            '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': f'_{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, f'Downloading{sub} JSON metadata', 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 f'{course_slug}/{video_slug}'


class LinkedInIE(LinkedInBaseIE):
    _VALID_URL = r'https?://(?:www\.)?linkedin\.com/posts/[^/?#]+-(?P<id>\d+)-\w{4}/?(?:[?#]|$)'
    _TESTS = [{
        'url': 'https://www.linkedin.com/posts/mishalkhawaja_sendinblueviews-toronto-digitalmarketing-ugcPost-6850898786781339649-mM20',
        'info_dict': {
            'id': '6850898786781339649',
            'ext': 'mp4',
            'title': 'Mishal K. on LinkedIn: #sendinblueviews #toronto #digitalmarketing #nowhiring #sendinblue…',
            'description': 'md5:2998a31f6f479376dd62831f53a80f71',
            'uploader': 'Mishal K.',
            'thumbnail': 're:^https?://media.licdn.com/dms/image/.*$',
            'like_count': int,
        },
    }, {
        'url': 'https://www.linkedin.com/posts/the-mathworks_2_what-is-mathworks-cloud-center-activity-7151241570371948544-4Gu7',
        'info_dict': {
            'id': '7151241570371948544',
            'ext': 'mp4',
            'title': 'MathWorks on LinkedIn: What Is MathWorks Cloud Center?',
            'description': 'md5:95f9d4eeb6337882fb47eefe13d7a40c',
            'uploader': 'MathWorks',
            'thumbnail': 're:^https?://media.licdn.com/dms/image/.*$',
            'like_count': int,
            'subtitles': 'mincount:1',
        },
    }]

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

        video_attrs = extract_attributes(self._search_regex(r'(<video[^>]+>)', webpage, 'video'))
        sources = self._parse_json(video_attrs['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]
        subtitles = {'en': [{
            'url': video_attrs['data-captions-url'],
            'ext': 'vtt',
        }]} if url_or_none(video_attrs.get('data-captions-url')) else {}

        return {
            'id': video_id,
            'formats': formats,
            'title': self._og_search_title(webpage, default=None) or self._html_extract_title(webpage),
            'like_count': int_or_none(self._search_regex(
                r'\bdata-num-reactions="(\d+)"', webpage, 'reactions', default=None)),
            'uploader': traverse_obj(
                self._yield_json_ld(webpage, video_id),
                (lambda _, v: v['@type'] == 'SocialMediaPosting', 'author', 'name', {str}), get_all=False),
            'thumbnail': self._og_search_thumbnail(webpage),
            'description': self._og_search_description(webpage, default=None),
            'subtitles': subtitles,
        }


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(itertools.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 += (
                f'{line + 1}\n'
                f'{srt_subtitles_timecode(start_time)} --> {srt_subtitles_timecode(end_time)}\n'
                f'{caption}\n\n')
        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': f'progressive-{height}p',
                    '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))

        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,
            # It seems like this would be correctly handled by default
            # However, unless someone can confirm this, the old
            # behaviour is being kept as-is
            '_format_sort_fields': ('res', 'source_preference'),
        }


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().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': f'https://www.linkedin.com/learning/{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
e3a3ed8a	1	import itertools
aa7e974a RA	2	import re
	3
	4	from .common import InfoExtractor
	5	from ..utils import (
	6	ExtractorError,
017adb28	7	extract_attributes,
aa7e974a RA	8	float_or_none,
aa7e974a RA	9	int_or_none,
402cd603	10	mimetype2ext,
e897bd82	11	srt_subtitles_timecode,
017adb28	12	traverse_obj,
8dc831f7	13	try_get,
017adb28	14	url_or_none,
aa7e974a	15	urlencode_postdata,
c9120294	16	urljoin,
aa7e974a RA	17	)
	18
	19
402cd603	20	class LinkedInBaseIE(InfoExtractor):
aa7e974a	21	_NETRC_MACHINE = 'linkedin'
da483200	22	_logged_in = False
aa7e974a	23
52efa4b3	24	def _perform_login(self, username, password):
402cd603	25	if self._logged_in:
402cd603	26	return
402cd603	27
	28	login_page = self._download_webpage(
	29	self._LOGIN_URL, None, 'Downloading login page')
	30	action_url = urljoin(self._LOGIN_URL, self._search_regex(
	31	r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page, 'post url',
	32	default='https://www.linkedin.com/uas/login-submit', group='url'))
	33	data = self._hidden_inputs(login_page)
	34	data.update({
52efa4b3	35	'session_key': username,
402cd603	36	'session_password': password,
	37	})
	38	login_submit_page = self._download_webpage(
	39	action_url, None, 'Logging in',
	40	data=urlencode_postdata(data))
	41	error = self._search_regex(
	42	r'<span[^>]+class="error"[^>]>\s(.+?)\s*</span>',
	43	login_submit_page, 'error', default=None)
	44	if error:
	45	raise ExtractorError(error, expected=True)
	46	LinkedInBaseIE._logged_in = True
	47
	48
	49	class LinkedInLearningBaseIE(LinkedInBaseIE):
	50	_LOGIN_URL = 'https://www.linkedin.com/uas/login?trk=learning'
	51
aa7e974a RA	52	def _call_api(self, course_slug, fields, video_slug=None, resolution=None):
	53	query = {
	54	'courseSlug': course_slug,
	55	'fields': fields,
	56	'q': 'slugs',
	57	}
	58	sub = ''
	59	if video_slug:
	60	query.update({
	61	'videoSlug': video_slug,
add96eb9	62	'resolution': f'_{resolution}',
aa7e974a RA	63	})
	64	sub = ' %dp' % resolution
	65	api_url = 'https://www.linkedin.com/learning-api/detailedCourses'
da483200	66	if not self._get_cookies(api_url).get('JSESSIONID'):
da483200	67	self.raise_login_required()
aa7e974a	68	return self._download_json(
add96eb9	69	api_url, video_slug, f'Downloading{sub} JSON metadata', headers={
aa7e974a RA	70	'Csrf-Token': self._get_cookies(api_url)['JSESSIONID'].value,
	71	}, query=query)['elements'][0]
	72
91effe22 RA	73	def _get_urn_id(self, video_data):
91effe22 RA	74	urn = video_data.get('urn')
aa7e974a RA	75	if urn:
	76	mobj = re.search(r'urn:li:lyndaCourse:\d+,(\d+)', urn)
	77	if mobj:
	78	return mobj.group(1)
91effe22 RA	79
91effe22 RA	80	def _get_video_id(self, video_data, course_slug, video_slug):
add96eb9	81	return self._get_urn_id(video_data) or f'{course_slug}/{video_slug}'
aa7e974a	82
aa7e974a	83
402cd603	84	class LinkedInIE(LinkedInBaseIE):
017adb28	85	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/posts/[^/?#]+-(?P<id>\d+)-\w{4}/?(?:[?#]\|$)'
402cd603	86	_TESTS = [{
	87	'url': 'https://www.linkedin.com/posts/mishalkhawaja_sendinblueviews-toronto-digitalmarketing-ugcPost-6850898786781339649-mM20',
	88	'info_dict': {
	89	'id': '6850898786781339649',
	90	'ext': 'mp4',
017adb28	91	'title': 'Mishal K. on LinkedIn: #sendinblueviews #toronto #digitalmarketing #nowhiring #sendinblue…',
	92	'description': 'md5:2998a31f6f479376dd62831f53a80f71',
	93	'uploader': 'Mishal K.',
	94	'thumbnail': 're:^https?://media.licdn.com/dms/image/.*$',
add96eb9	95	'like_count': int,
017adb28	96	},
	97	}, {
	98	'url': 'https://www.linkedin.com/posts/the-mathworks_2_what-is-mathworks-cloud-center-activity-7151241570371948544-4Gu7',
	99	'info_dict': {
	100	'id': '7151241570371948544',
	101	'ext': 'mp4',
	102	'title': 'MathWorks on LinkedIn: What Is MathWorks Cloud Center?',
	103	'description': 'md5:95f9d4eeb6337882fb47eefe13d7a40c',
	104	'uploader': 'MathWorks',
	105	'thumbnail': 're:^https?://media.licdn.com/dms/image/.*$',
	106	'like_count': int,
add96eb9	107	'subtitles': 'mincount:1',
402cd603	108	},
	109	}]
	110
	111	def _real_extract(self, url):
	112	video_id = self._match_id(url)
	113	webpage = self._download_webpage(url, video_id)
	114
017adb28	115	video_attrs = extract_attributes(self._search_regex(r'(<video[^>]+>)', webpage, 'video'))
017adb28	116	sources = self._parse_json(video_attrs['data-sources'], video_id)
402cd603	117	formats = [{
	118	'url': source['src'],
	119	'ext': mimetype2ext(source.get('type')),
	120	'tbr': float_or_none(source.get('data-bitrate'), scale=1000),
	121	} for source in sources]
017adb28	122	subtitles = {'en': [{
	123	'url': video_attrs['data-captions-url'],
	124	'ext': 'vtt',
	125	}]} if url_or_none(video_attrs.get('data-captions-url')) else {}
402cd603	126
402cd603	127	return {
	128	'id': video_id,
	129	'formats': formats,
017adb28	130	'title': self._og_search_title(webpage, default=None) or self._html_extract_title(webpage),
	131	'like_count': int_or_none(self._search_regex(
	132	r'\bdata-num-reactions="(\d+)"', webpage, 'reactions', default=None)),
	133	'uploader': traverse_obj(
	134	self._yield_json_ld(webpage, video_id),
	135	(lambda _, v: v['@type'] == 'SocialMediaPosting', 'author', 'name', {str}), get_all=False),
402cd603	136	'thumbnail': self._og_search_thumbnail(webpage),
017adb28	137	'description': self._og_search_description(webpage, default=None),
017adb28	138	'subtitles': subtitles,
402cd603	139	}
aa7e974a RA	140
	141
	142	class LinkedInLearningIE(LinkedInLearningBaseIE):
	143	IE_NAME = 'linkedin:learning'
	144	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<course_slug>[^/]+)/(?P<id>[^/?#]+)'
	145	_TEST = {
	146	'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals/welcome?autoplay=true',
	147	'md5': 'a1d74422ff0d5e66a792deb996693167',
	148	'info_dict': {
	149	'id': '90426',
	150	'ext': 'mp4',
	151	'title': 'Welcome',
	152	'timestamp': 1430396150.82,
	153	'upload_date': '20150430',
	154	},
	155	}
	156
8dc831f7 AG	157	def json2srt(self, transcript_lines, duration=None):
8dc831f7 AG	158	srt_data = ''
e3a3ed8a	159	for line, (line_dict, next_dict) in enumerate(itertools.zip_longest(transcript_lines, transcript_lines[1:])):
8dc831f7 AG	160	start_time, caption = line_dict['transcriptStartAt'] / 1000, line_dict['caption']
8dc831f7 AG	161	end_time = next_dict['transcriptStartAt'] / 1000 if next_dict else duration or start_time + 1
add96eb9	162	srt_data += (
	163	f'{line + 1}\n'
	164	f'{srt_subtitles_timecode(start_time)} --> {srt_subtitles_timecode(end_time)}\n'
	165	f'{caption}\n\n')
8dc831f7 AG	166	return srt_data
8dc831f7 AG	167
aa7e974a	168	def _real_extract(self, url):
5ad28e7f	169	course_slug, video_slug = self._match_valid_url(url).groups()
aa7e974a	170
aa7e974a RA	171	formats = []
	172	for width, height in ((640, 360), (960, 540), (1280, 720)):
	173	video_data = self._call_api(
	174	course_slug, 'selectedVideo', video_slug, height)['selectedVideo']
	175
	176	video_url_data = video_data.get('url') or {}
	177	progressive_url = video_url_data.get('progressiveUrl')
	178	if progressive_url:
	179	formats.append({
add96eb9	180	'format_id': f'progressive-{height}p',
aa7e974a	181	'url': progressive_url,
8dc831f7	182	'ext': 'mp4',
aa7e974a RA	183	'height': height,
	184	'width': width,
	185	'source_preference': 1,
	186	})
	187
	188	title = video_data['title']
	189
	190	audio_url = video_data.get('audio', {}).get('progressiveUrl')
	191	if audio_url:
	192	formats.append({
	193	'abr': 64,
	194	'ext': 'm4a',
	195	'format_id': 'audio',
	196	'url': audio_url,
	197	'vcodec': 'none',
	198	})
	199
	200	streaming_url = video_url_data.get('streamingUrl')
	201	if streaming_url:
	202	formats.extend(self._extract_m3u8_formats(
	203	streaming_url, video_slug, 'mp4',
	204	'm3u8_native', m3u8_id='hls', fatal=False))
	205
8dc831f7 AG	206	subtitles = {}
	207	duration = int_or_none(video_data.get('durationInSeconds'))
	208	transcript_lines = try_get(video_data, lambda x: x['transcript']['lines'], expected_type=list)
	209	if transcript_lines:
	210	subtitles['en'] = [{
	211	'ext': 'srt',
add96eb9	212	'data': self.json2srt(transcript_lines, duration),
8dc831f7	213	}]
aa7e974a RA	214
aa7e974a RA	215	return {
91effe22	216	'id': self._get_video_id(video_data, course_slug, video_slug),
aa7e974a RA	217	'title': title,
	218	'formats': formats,
	219	'thumbnail': video_data.get('defaultThumbnail'),
	220	'timestamp': float_or_none(video_data.get('publishedOn'), 1000),
8dc831f7 AG	221	'duration': duration,
8dc831f7 AG	222	'subtitles': subtitles,
9f14daf2	223	# It seems like this would be correctly handled by default
	224	# However, unless someone can confirm this, the old
	225	# behaviour is being kept as-is
add96eb9	226	'_format_sort_fields': ('res', 'source_preference'),
aa7e974a RA	227	}
	228
	229
	230	class LinkedInLearningCourseIE(LinkedInLearningBaseIE):
	231	IE_NAME = 'linkedin:learning:course'
	232	_VALID_URL = r'https?://(?:www\.)?linkedin\.com/learning/(?P<id>[^/?#]+)'
	233	_TEST = {
	234	'url': 'https://www.linkedin.com/learning/programming-foundations-fundamentals',
	235	'info_dict': {
	236	'id': 'programming-foundations-fundamentals',
	237	'title': 'Programming Foundations: Fundamentals',
	238	'description': 'md5:76e580b017694eb89dc8e8923fff5c86',
	239	},
	240	'playlist_mincount': 61,
	241	}
	242
	243	@classmethod
	244	def suitable(cls, url):
add96eb9	245	return False if LinkedInLearningIE.suitable(url) else super().suitable(url)
aa7e974a RA	246
	247	def _real_extract(self, url):
	248	course_slug = self._match_id(url)
	249	course_data = self._call_api(course_slug, 'chapters,description,title')
	250
	251	entries = []
91effe22	252	for chapter_number, chapter in enumerate(course_data.get('chapters', []), 1):
aa7e974a	253	chapter_title = chapter.get('title')
91effe22	254	chapter_id = self._get_urn_id(chapter)
aa7e974a RA	255	for video in chapter.get('videos', []):
	256	video_slug = video.get('slug')
	257	if not video_slug:
	258	continue
	259	entries.append({
b1447572	260	'_type': 'url_transparent',
91effe22	261	'id': self._get_video_id(video, course_slug, video_slug),
aa7e974a	262	'title': video.get('title'),
add96eb9	263	'url': f'https://www.linkedin.com/learning/{course_slug}/{video_slug}',
aa7e974a	264	'chapter': chapter_title,
91effe22 RA	265	'chapter_number': chapter_number,
91effe22 RA	266	'chapter_id': chapter_id,
aa7e974a RA	267	'ie_key': LinkedInLearningIE.ie_key(),
	268	})
	269
	270	return self.playlist_result(
	271	entries, course_slug,
	272	course_data.get('title'),
	273	course_data.get('description'))