Merge a571bbaab10c8ba9892c269d9bdb3fe58f004a90 into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

* thx seproDev, yt-dlp/yt-dlp/pull/12694/commits/ae69e3c
* also add documenting comment in `InfoExtractor`

docs/supportedsites.md

@@ -613,6 +613,7 @@
  - **niconico**: ニコニコ動画
  - **NiconicoPlaylist**
  - **Nintendo**
+ - **Nitter**
  - **njoy**: N-JOY
  - **njoy:embed**
  - **NJPWWorld**: 新日本プロレスワールド

youtube_dl/extractor/bokecc.py

@@ -32,7 +32,7 @@ class BokeCCBaseIE(InfoExtractor):
 
 
 class BokeCCIE(BokeCCBaseIE):
-    _IE_DESC = 'CC视频'
+    IE_DESC = 'CC视频'
     _VALID_URL = r'https?://union\.bokecc\.com/playvideo\.bo\?(?P<query>.*)'
 
     _TESTS = [{

youtube_dl/extractor/cloudy.py

@@ -9,7 +9,7 @@ from ..utils import (
 
 
 class CloudyIE(InfoExtractor):
-    _IE_DESC = 'cloudy.ec'
+    IE_DESC = 'cloudy.ec'
     _VALID_URL = r'https?://(?:www\.)?cloudy\.ec/(?:v/|embed\.php\?.*?\bid=)(?P<id>[A-Za-z0-9]+)'
     _TESTS = [{
         'url': 'https://www.cloudy.ec/v/af511e2527aac',

youtube_dl/extractor/common.py

@@ -422,6 +422,8 @@ class InfoExtractor(object):
     _GEO_COUNTRIES = None
     _GEO_IP_BLOCKS = None
     _WORKING = True
+    # supply this in public subclasses: used in supported sites list, etc
+    # IE_DESC = 'short description of IE'
 
     def __init__(self, downloader=None):
         """Constructor. Receives an optional downloader."""

youtube_dl/extractor/extractors.py

@@ -830,6 +830,7 @@ from .ninecninemedia import NineCNineMediaIE
 from .ninegag import NineGagIE
 from .ninenow import NineNowIE
 from .nintendo import NintendoIE
+from .nitter import NitterIE
 from .njpwworld import NJPWWorldIE
 from .nobelprize import NobelPrizeIE
 from .nonktube import NonkTubeIE

youtube_dl/extractor/itv.py

@@ -35,15 +35,6 @@ from ..utils import (
 
 class ITVBaseIE(InfoExtractor):
 
-    def _search_nextjs_data(self, webpage, video_id, **kw):
-        transform_source = kw.pop('transform_source', None)
-        fatal = kw.pop('fatal', True)
-        return self._parse_json(
-            self._search_regex(
-                r'''<script\b[^>]+\bid=('|")__NEXT_DATA__\1[^>]*>(?P<js>[^<]+)</script>''',
-                webpage, 'next.js data', group='js', fatal=fatal, **kw),
-            video_id, transform_source=transform_source, fatal=fatal)
-
     def __handle_request_webpage_error(self, err, video_id=None, errnote=None, fatal=True):
         if errnote is False:
             return False
@@ -109,7 +100,9 @@ class ITVBaseIE(InfoExtractor):
 
 class ITVIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:(?P<w>watch)|hub)/[^/]+/(?(w)[\w-]+/)(?P<id>\w+)'
-    _IE_DESC = 'ITVX'
+    IE_DESC = 'ITVX'
+    _WORKING = False
+
     _TESTS = [{
         'note': 'Hub URLs redirect to ITVX',
         'url': 'https://www.itv.com/hub/liar/2a4547a0012',
@@ -270,7 +263,7 @@ class ITVIE(ITVBaseIE):
                 'ext': determine_ext(href, 'vtt'),
             })
 
-        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default='{}')
+        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default={})
         video_data.update(traverse_obj(next_data, ('props', 'pageProps', ('title', 'episode')), expected_type=dict)[0] or {})
         title = traverse_obj(video_data, 'headerTitle', 'episodeTitle')
         info = self._og_extract(webpage, require_title=not title)
@@ -323,7 +316,7 @@ class ITVIE(ITVBaseIE):
 
 class ITVBTCCIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?!(?:watch|hub)/)(?:[^/]+/)+(?P<id>[^/?#&]+)'
-    _IE_DESC = 'ITV articles: News, British Touring Car Championship'
+    IE_DESC = 'ITV articles: News, British Touring Car Championship'
     _TESTS = [{
         'note': 'British Touring Car Championship',
         'url': 'https://www.itv.com/btcc/articles/btcc-2018-all-the-action-from-brands-hatch',

youtube_dl/extractor/nitter.py

@@ -0,0 +1,173 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_urlparse
+from ..utils import (
+    parse_count,
+    unified_strdate,
+    unified_timestamp,
+    remove_end,
+    determine_ext,
+)
+import re
+
+
+class NitterIE(InfoExtractor):
+    # Taken from https://github.com/zedeus/nitter/wiki/Instances
+    INSTANCES = ('nitter.net',
+                 'nitter.snopyta.org',
+                 'nitter.42l.fr',
+                 'nitter.nixnet.services',
+                 'nitter.13ad.de',
+                 'nitter.pussthecat.org',
+                 'nitter.mastodont.cat',
+                 'nitter.dark.fail',
+                 'nitter.tedomum.net',
+                 'nitter.cattube.org',
+                 'nitter.fdn.fr',
+                 'nitter.1d4.us',
+                 'nitter.kavin.rocks',
+                 'tweet.lambda.dance',
+                 'nitter.cc',
+                 'nitter.weaponizedhumiliation.com',
+                 'nitter.vxempire.xyz',
+                 'nitter.unixfox.eu',
+                 'nitter.domain.glass',
+                 'nitter.himiko.cloud',
+                 'nitter.eu',
+                 'nitter.ethibox.fr',
+                 '3nzoldnxplag42gqjs23xvghtzf6t6yzssrtytnntc6ppc7xxuoneoad.onion',
+                 'nitter.l4qlywnpwqsluw65ts7md3khrivpirse744un3x7mlskqauz5pyuzgqd.onion',
+                 'npf37k3mtzwxreiw52ccs5ay4e6qt2fkcs2ndieurdyn2cuzzsfyfvid.onion')
+
+    _INSTANCES_RE = '(?:' + '|'.join([re.escape(instance) for instance in INSTANCES]) + ')'
+    _VALID_URL = r'https?://%(instance)s/(?P<uploader_id>.+)/status/(?P<id>[0-9]+)(#.)?' % {'instance': _INSTANCES_RE}
+    current_instance = INSTANCES[0]  # the test and official instance
+    _TESTS = [
+        {
+            # GIF (wrapped in mp4)
+            'url': 'https://' + current_instance + '/firefox/status/1314279897502629888#m',
+            'info_dict': {
+                'id': '1314279897502629888',
+                'ext': 'mp4',
+                'title': 'Firefox 🔥 - You know the old saying, if you see something say something. Now you actually can with the YouTube regrets extension.   Report harmful YouTube recommendations so others can avoid watching them. ➡️ https://mzl.la/3iFIiyg  #UnfckTheInternet',
+                'description': 'You know the old saying, if you see something say something. Now you actually can with the YouTube regrets extension.   Report harmful YouTube recommendations so others can avoid watching them. ➡️ https://mzl.la/3iFIiyg  #UnfckTheInternet',
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Firefox 🔥',
+                'uploader_id': 'firefox',
+                'uploader_url': 'https://' + current_instance + '/firefox',
+                'upload_date': '20201008',
+                'timestamp': 1602183720,
+            },
+        }, {  # normal video
+            'url': 'https://' + current_instance + '/Le___Doc/status/1299715685392756737#m',
+            'info_dict': {
+                'id': '1299715685392756737',
+                'ext': 'mp4',
+                'title': 'Le Doc - "Je ne prédis jamais rien" D Raoult, Août 2020...',
+                'description': '"Je ne prédis jamais rien" D Raoult, Août 2020...',
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Le Doc',
+                'uploader_id': 'Le___Doc',
+                'uploader_url': 'https://' + current_instance + '/Le___Doc',
+                'upload_date': '20200829',
+                'timestamp': 1598711341,
+                'view_count': int,
+                'like_count': int,
+                'repost_count': int,
+                'comment_count': int,
+            },
+        }, {  # video embed in a "Streaming Political Ads" box
+            'url': 'https://' + current_instance + '/mozilla/status/1321147074491092994#m',
+            'info_dict': {
+                'id': '1321147074491092994',
+                'ext': 'mp4',
+                'title': "Mozilla - Are you being targeted with weird, ominous or just plain annoying political ads while streaming your favorite shows?  This isn't a real political ad, but if you're watching streaming TV in the U.S., chances are you've seen quite a few.   Learn more ➡️ https://mzl.la/StreamingAds",
+                'description': "Are you being targeted with weird, ominous or just plain annoying political ads while streaming your favorite shows?  This isn't a real political ad, but if you're watching streaming TV in the U.S., chances are you've seen quite a few.   Learn more ➡️ https://mzl.la/StreamingAds",
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'uploader': 'Mozilla',
+                'uploader_id': 'mozilla',
+                'uploader_url': 'https://' + current_instance + '/mozilla',
+                'upload_date': '20201027',
+                'timestamp': 1603820982
+            },
+        },
+    ]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        parsed_url = compat_urlparse.urlparse(url)
+        base_url = parsed_url.scheme + '://' + parsed_url.netloc
+
+        self._set_cookie(parsed_url.netloc, 'hlsPlayback', 'on')
+        webpage = self._download_webpage(url, video_id)
+
+        video_url = base_url + self._html_search_regex(r'(?:<video[^>]+data-url|<source[^>]+src)="([^"]+)"', webpage, 'video url')
+        ext = determine_ext(video_url)
+
+        if ext == 'unknown_video':
+            formats = self._extract_m3u8_formats(video_url, video_id, ext='mp4')
+        else:
+            formats = [{
+                'url': video_url,
+                'ext': ext
+            }]
+
+        title = (
+            self._og_search_description(webpage).replace('\n', ' ')
+            or self._html_search_regex(r'<div class="tweet-content[^>]+>([^<]+)</div>', webpage, 'title'))
+        description = title
+
+        mobj = re.match(self._VALID_URL, url)
+        uploader_id = (
+            mobj.group('uploader_id')
+            or self._html_search_regex(r'<a class="fullname"[^>]+title="([^"]+)"', webpage, 'uploader name', fatal=False))
+
+        if uploader_id:
+            uploader_url = base_url + '/' + uploader_id
+
+        uploader = self._html_search_regex(r'<a class="fullname"[^>]+title="([^"]+)"', webpage, 'uploader name', fatal=False)
+
+        if uploader:
+            title = uploader + ' - ' + title
+
+        view_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-play[^>]*></span>\s([^<]+)</div>', webpage, 'view count', fatal=False))
+        like_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-heart[^>]*></span>\s([^<]+)</div>', webpage, 'like count', fatal=False))
+        repost_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-retweet[^>]*></span>\s([^<]+)</div>', webpage, 'repost count', fatal=False))
+        comment_count = parse_count(self._html_search_regex(r'<span[^>]+class="icon-comment[^>]*></span>\s([^<]+)</div>', webpage, 'repost count', fatal=False))
+
+        thumbnail = base_url + (self._html_search_meta('og:image', webpage, 'thumbnail url')
+                                or self._html_search_regex(r'<video[^>]+poster="([^"]+)"', webpage, 'thumbnail url', fatal=False))
+
+        thumbnail = remove_end(thumbnail, '%3Asmall')  # if parsed with regex, it should contain this
+
+        thumbnails = []
+        thumbnail_ids = ('thumb', 'small', 'large', 'medium', 'orig')
+        for id in thumbnail_ids:
+            thumbnails.append({
+                'id': id,
+                'url': thumbnail + '%3A' + id,
+            })
+
+        date = self._html_search_regex(r'<span[^>]+class="tweet-date"[^>]*><a[^>]+title="([^"]+)"', webpage, 'upload date', fatal=False)
+        upload_date = unified_strdate(date)
+        timestamp = unified_timestamp(date)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'uploader': uploader,
+            'timestamp': timestamp,
+            'uploader_id': uploader_id,
+            'uploader_url': uploader_url,
+            'view_count': view_count,
+            'like_count': like_count,
+            'repost_count': repost_count,
+            'comment_count': comment_count,
+            'formats': formats,
+            'thumbnails': thumbnails,
+            'thumbnail': thumbnail,
+            'upload_date': upload_date,
+        }

youtube_dl/extractor/senateisvp.py

@@ -47,7 +47,7 @@ class SenateISVPIE(InfoExtractor):
         ['vetaff', '76462', 'http://vetaff-f.akamaihd.net'],
         ['arch', '', 'http://ussenate-f.akamaihd.net/']
     ]
-    _IE_NAME = 'senate.gov'
+    IE_NAME = 'senate.gov'
     _VALID_URL = r'https?://(?:www\.)?senate\.gov/isvp/?\?(?P<qs>.+)'
     _TESTS = [{
         'url': 'http://www.senate.gov/isvp/?comm=judiciary&type=live&stt=&filename=judiciary031715&auto_play=false&wmode=transparent&poster=http%3A%2F%2Fwww.judiciary.senate.gov%2Fthemes%2Fjudiciary%2Fimages%2Fvideo-poster-flash-fit.png',

youtube_dl/jsinterp.py

@@ -686,6 +686,8 @@ class JSInterpreter(object):
             raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
+        if obj is JS_Undefined:
+            return 'undefined'
         try:
             return json.dumps(obj)
         except TypeError: