Merge b76f13229d into f2a4983df7

Closes #11526
Authored by: jshumphrey

yt_dlp/extractor/_extractors.py

@@ -582,6 +582,7 @@ from .egghead import (
 )
 from .eighttracks import EightTracksIE
 from .eitb import EitbIE
+from .elemental_tv import ElementalTVIE
 from .elementorembed import ElementorEmbedIE
 from .elonet import ElonetIE
 from .elpais import ElPaisIE

yt_dlp/extractor/archiveorg.py

@@ -205,6 +205,26 @@ class ArchiveOrgIE(InfoExtractor):
                 },
             },
         ],
+    }, {
+        # The reviewbody is None for one of the reviews; just need to extract data without crashing
+        'url': 'https://archive.org/details/gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
+        'info_dict': {
+            'id': 'gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
+            'ext': 'mp3',
+            'title': 'Stuck Inside of Mobile with the Memphis Blues Again',
+            'creators': ['Grateful Dead'],
+            'duration': 338.31,
+            'track': 'Stuck Inside of Mobile with the Memphis Blues Again',
+            'description': 'md5:764348a470b986f1217ffd38d6ac7b72',
+            'display_id': 'gd95-04-02d1t04.shn',
+            'location': 'Pyramid Arena',
+            'uploader': 'jon@archive.org',
+            'album': '1995-04-02 - Pyramid Arena',
+            'upload_date': '20040519',
+            'track_number': 4,
+            'release_date': '19950402',
+            'timestamp': 1084927901,
+        },
     }]
 
     @staticmethod
@@ -335,7 +355,7 @@ class ArchiveOrgIE(InfoExtractor):
                 info['comments'].append({
                     'id': review.get('review_id'),
                     'author': review.get('reviewer'),
-                    'text': str_or_none(review.get('reviewtitle'), '') + '\n\n' + review.get('reviewbody'),
+                    'text': join_nonempty('reviewtitle', 'reviewbody', from_dict=review, delim='\n\n'),
                     'timestamp': unified_timestamp(review.get('createdate')),
                     'parent': 'root'})
 

yt_dlp/extractor/elemental_tv.py

@@ -0,0 +1,107 @@
+import json
+import re
+import time
+
+from ..extractor.common import InfoExtractor
+from ..networking.exceptions import HTTPError
+from ..utils import ExtractorError
+
+
+class ElementalTVIE(InfoExtractor):
+    _LOGIN_REQUIRED = True
+    _NETRC_MACHINE = 'elemental_tv'
+    _VALID_URL = r'https?://play\.elemental\.tv/channel/[0-9a-f]{24}'
+    _TESTS = [{
+        'url': 'https://play.elemental.tv/channel/573f5a14761973ec1d502507',
+        'info_dict': {
+            'id': '573f5a14761973ec1d502507',
+            'ext': 'mp4',
+            'title': 'БНТ 1 HD',
+            'thumbnail': 'https://play.elemental.tv/v1/tumblrs/573f5a14761973ec1d502507',
+            'age_limit': 0,
+        },
+    }]
+
+    access_token = ''
+    channel_id = ''
+
+    def _get_channel_id(self, url):
+        url_parts = re.search('(?<=channel/)[0-9a-f]{24}', url)
+
+        if not url_parts or not url_parts.group(0):
+            return None
+
+        self.channel_id = url_parts.group(0)
+        self.write_debug(f'Channel ID: {self.channel_id}')
+
+        if not self.channel_id:
+            raise ExtractorError('Channel ID not found')
+
+    def _get_stream_metadata(self):
+        try:
+            headers = {
+                'Authorization': 'Bearer ' + self.access_token,
+            }
+            res_api = self._download_json(
+                'https://play.elemental.tv/v1/channels', self.channel_id, headers=headers)
+            data = res_api.get('data').get(self.channel_id)
+
+            if not data:
+                self.write_debug('Getting metadata failed')
+                return {}
+
+            return {
+                'title': data.get('name'),
+                'age_limit': data.get('age'),
+                'thumbnail': data.get('tumblrurl'),
+            }
+        except Exception:
+            self.write_debug('Getting metadata failed')
+            return {}
+
+    def _get_stream_url(self):
+        # Stream URL needs current epoch time rounded to 10000s
+        begin = int((time.time() - 60) / 10000) * 10000
+        stream_url = 'https://play.elemental.tv/v1/playlists/%s/playlist.m3u8?begin=%d&access_token=%s' % (self.channel_id, begin, self.access_token)
+
+        if not stream_url or '.m3u8' not in stream_url:
+            raise ExtractorError('Unable to get stream URL')
+
+        return stream_url
+
+    def _perform_login(self, username, password):
+        post_data = {
+            'email': str(username),
+            'grant_type': 'client_credentials',
+            'password': str(password),
+            'rememberme': 'true',
+        }
+
+        try:
+            res_api = self._download_json(
+                'https://play.elemental.tv/v1/users/login', self.channel_id, data=json.dumps(post_data).encode()).get('data')
+        except ExtractorError as e:
+            if isinstance(e.cause, HTTPError) and e.cause.status == 400:
+                error_message = self._parse_json(e.cause.response.read().decode(), self.channel_id).get('error_info').get('description')
+                raise ExtractorError(error_message, expected=True)
+
+        if not res_api or not res_api.get('access_token'):
+            raise ExtractorError('Accessing login token failed')
+
+        self.access_token = res_api.get('access_token')
+
+        if res_api.get('token_type') != 'Bearer':
+            raise ExtractorError('Unknown login token type')
+
+    def _real_extract(self, url):
+        self._get_channel_id(url)
+        stream_url = self._get_stream_url()
+        formats, subtitles = self._extract_m3u8_formats_and_subtitles(stream_url, self.channel_id, ext='mp4')
+
+        return {
+            'id': self.channel_id,
+            'is_live': True,
+            'formats': formats,
+            'subtitles': subtitles,
+            **self._get_stream_metadata(),
+        }

yt_dlp/extractor/facebook.py

@@ -563,13 +563,13 @@ class FacebookIE(InfoExtractor):
                 return extract_video_data(try_get(
                     js_data, lambda x: x['jsmods']['instances'], list) or [])
 
-        def extract_dash_manifest(video, formats):
+        def extract_dash_manifest(vid_data, formats, mpd_url=None):
             dash_manifest = traverse_obj(
-                video, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', expected_type=str)
+                vid_data, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', 'manifest_xml', expected_type=str)
             if dash_manifest:
                 formats.extend(self._parse_mpd_formats(
                     compat_etree_fromstring(urllib.parse.unquote_plus(dash_manifest)),
-                    mpd_url=url_or_none(video.get('dash_manifest_url'))))
+                    mpd_url=url_or_none(video.get('dash_manifest_url')) or mpd_url))
 
         def process_formats(info):
             # Downloads with browser's User-Agent are rate limited. Working around
@@ -619,9 +619,12 @@ class FacebookIE(InfoExtractor):
                         video = video['creation_story']
                         video['owner'] = traverse_obj(video, ('short_form_video_context', 'video_owner'))
                         video.update(reel_info)
-                    fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
+
                     formats = []
                     q = qualities(['sd', 'hd'])
+
+                    # Legacy formats extraction
+                    fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
                     for key, format_id in (('playable_url', 'sd'), ('playable_url_quality_hd', 'hd'),
                                            ('playable_url_dash', ''), ('browser_native_hd_url', 'hd'),
                                            ('browser_native_sd_url', 'sd')):
@@ -629,7 +632,7 @@ class FacebookIE(InfoExtractor):
                         if not playable_url:
                             continue
                         if determine_ext(playable_url) == 'mpd':
-                            formats.extend(self._extract_mpd_formats(playable_url, video_id))
+                            formats.extend(self._extract_mpd_formats(playable_url, video_id, fatal=False))
                         else:
                             formats.append({
                                 'format_id': format_id,
@@ -638,6 +641,28 @@ class FacebookIE(InfoExtractor):
                                 'url': playable_url,
                             })
                     extract_dash_manifest(fmt_data, formats)
+
+                    # New videoDeliveryResponse formats extraction
+                    fmt_data = traverse_obj(video, ('videoDeliveryResponseFragment', 'videoDeliveryResponseResult'))
+                    mpd_urls = traverse_obj(fmt_data, ('dash_manifest_urls', ..., 'manifest_url', {url_or_none}))
+                    dash_manifests = traverse_obj(fmt_data, ('dash_manifests', lambda _, v: v['manifest_xml']))
+                    for idx, dash_manifest in enumerate(dash_manifests):
+                        extract_dash_manifest(dash_manifest, formats, mpd_url=traverse_obj(mpd_urls, idx))
+                    if not dash_manifests:
+                        # Only extract from MPD URLs if the manifests are not already provided
+                        for mpd_url in mpd_urls:
+                            formats.extend(self._extract_mpd_formats(mpd_url, video_id, fatal=False))
+                    for prog_fmt in traverse_obj(fmt_data, ('progressive_urls', lambda _, v: v['progressive_url'])):
+                        format_id = traverse_obj(prog_fmt, ('metadata', 'quality', {str.lower}))
+                        formats.append({
+                            'format_id': format_id,
+                            # sd, hd formats w/o resolution info should be deprioritized below DASH
+                            'quality': q(format_id) - 3,
+                            'url': prog_fmt['progressive_url'],
+                        })
+                    for m3u8_url in traverse_obj(fmt_data, ('hls_playlist_urls', ..., 'hls_playlist_url', {url_or_none})):
+                        formats.extend(self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', fatal=False, m3u8_id='hls'))
+
                     if not formats:
                         # Do not append false positive entry w/o any formats
                         return