Merge 696bf76b8b into f2a4983df7

Closes #11526
Authored by: jshumphrey

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/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

yt_dlp/extractor/radiko.py

@@ -1,9 +1,11 @@
 import base64
+import datetime
 import random
 import re
 import urllib.parse
 
 from .common import InfoExtractor
+from ..networking.exceptions import HTTPError
 from ..utils import (
     ExtractorError,
     clean_html,
@@ -12,12 +14,14 @@ from ..utils import (
     try_call,
     unified_timestamp,
     update_url_query,
+    urlencode_postdata,
 )
 from ..utils.traversal import traverse_obj
 
 
 class RadikoBaseIE(InfoExtractor):
     _GEO_BYPASS = False
+    _NETRC_MACHINE = 'radiko'
     _FULL_KEY = None
     _HOSTS_FOR_TIME_FREE_FFMPEG_UNSUPPORTED = (
         'https://c-rpaa.smartstream.ne.jp',
@@ -37,6 +41,29 @@ class RadikoBaseIE(InfoExtractor):
         'https://c-radiko.smartstream.ne.jp',
     )
 
+    _JST = datetime.timezone(datetime.timedelta(hours=9))
+    _has_tf30 = None
+
+    def _perform_login(self, username, password):
+        try:
+            login_info = self._download_json('https://radiko.jp/ap/member/webapi/member/login', None, note='Logging in',
+                                             data=urlencode_postdata({'mail': username, 'pass': password}))
+            self._has_tf30 = '2' in login_info.get('privileges')
+        except ExtractorError as error:
+            if isinstance(error.cause, HTTPError) and error.cause.status == 401:
+                raise ExtractorError('Invalid username and/or password', expected=True)
+            raise
+
+    def _check_tf30(self):
+        if self._has_tf30 is not None:
+            return self._has_tf30
+        if self._get_cookies('https://radiko.jp').get('radiko_session') is None:
+            return
+        account_info = self._download_json('https://radiko.jp/ap/member/webapi/v2/member/login/check',
+                                           None, note='Checking account status from cookies', expected_status=400)
+        self._has_tf30 = account_info.get('timefreeplus') == '1'
+        return self._has_tf30
+
     def _negotiate_token(self):
         _, auth1_handle = self._download_webpage_handle(
             'https://radiko.jp/v2/api/auth1', None, 'Downloading authentication page',
@@ -99,17 +126,39 @@ class RadikoBaseIE(InfoExtractor):
         self._FULL_KEY = full_key
         return full_key
 
+    def _get_broadcast_day(self, timestring):
+        dt = datetime.datetime.strptime(timestring, '%Y%m%d%H%M%S')
+        if dt.hour < 5:
+            dt -= datetime.timedelta(days=1)
+        return dt
+
+    def _get_broadcast_day_end(self, dt):
+        dt += datetime.timedelta(days=1)
+        return datetime.datetime(dt.year, dt.month, dt.day, 5, 0, 0, tzinfo=self._JST)
+
     def _find_program(self, video_id, station, cursor):
+        broadcast_day = self._get_broadcast_day(cursor)
+        broadcast_day_str = broadcast_day.strftime('%Y%m%d')
+
+        broadcast_day_end = self._get_broadcast_day_end(broadcast_day)
+        now = datetime.datetime.now(tz=self._JST)
+
+        if broadcast_day_end + datetime.timedelta(days=30) < now:
+            self.raise_no_formats('Programme is no longer available.', video_id=video_id, expected=True)
+        elif broadcast_day_end + datetime.timedelta(days=7) < now and not self._check_tf30():
+            self.raise_login_required('Programme is only available with a Timefree 30 subscription',
+                                      metadata_available=True)
+
         station_program = self._download_xml(
-            f'https://radiko.jp/v3/program/station/weekly/{station}.xml', video_id,
-            note=f'Downloading radio program for {station} station')
+            f'https://api.radiko.jp/program/v3/date/{broadcast_day_str}/station/{station}.xml', station,
+            note=f'Downloading programme information for {broadcast_day_str}')
 
         prog = None
         for p in station_program.findall('.//prog'):
             ft_str, to_str = p.attrib['ft'], p.attrib['to']
             ft = unified_timestamp(ft_str, False)
             to = unified_timestamp(to_str, False)
-            if ft <= cursor and cursor < to:
+            if ft_str <= cursor and cursor < to_str:
                 prog = p
                 break
         if not prog:
@@ -187,7 +236,7 @@ class RadikoIE(RadikoBaseIE):
         station, timestring = self._match_valid_url(url).group('station', 'timestring')
         video_id = join_nonempty(station, timestring)
         vid_int = unified_timestamp(timestring, False)
-        prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, vid_int)
+        prog, station_program, ft, radio_begin, radio_end = self._find_program(video_id, station, timestring)
 
         auth_token, area_id = self._auth_client()