Merge 5888bba9e659795587555e56552e38ec8713211f into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

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

test/test_YoutubeDL.py

@@ -411,6 +411,15 @@ class TestFormatSelection(unittest.TestCase):
         downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts]
         self.assertEqual(downloaded_ids, ['248+141'])
 
+        for f in ['248+141', '141+248']:
+            info_dict = _make_result(list(formats_order), extractor='youtube')
+            ydl = YDL({'format': f})
+            yie = YoutubeIE(ydl)
+            yie._sort_formats(info_dict['formats'])
+            ydl.process_ie_result(info_dict)
+            downloaded_ids = [info['format_id'] for info in ydl.downloaded_info_dicts]
+            self.assertEqual(downloaded_ids, ['248+141'])
+
         for f1, f2 in zip(formats_order, formats_order[1:]):
             info_dict = _make_result([f1, f2], extractor='youtube')
             ydl = YDL({'format': 'best/bestvideo'})

youtube_dl/YoutubeDL.py

@@ -1489,39 +1489,48 @@ class YoutubeDL(object):
                             yield matches[-1]
             elif selector.type == MERGE:
                 def _merge(formats_info):
-                    format_1, format_2 = [f['format_id'] for f in formats_info]
+                    format_1, format_2 = formats_info[0], formats_info[1]
+                    format_1_id, format_2_id = format_1['format_id'], format_2['format_id']
                     # The first format must contain the video and the
                     # second the audio
-                    if formats_info[0].get('vcodec') == 'none':
+
+                    # If the user swapped the two inputs, try swapping it for
+                    # them
+                    if format_1.get('acodec') != 'none' and format_2.get('vcodec') != 'none':
+                        temp = format_1
+                        format_1 = format_2
+                        format_2 = temp
+
+                    if format_1.get('vcodec') == 'none':
                         self.report_error('The first format must '
                                           'contain the video, try using '
-                                          '"-f %s+%s"' % (format_2, format_1))
-                        return
+                                          '"-f %s+%s"' % (format_1_id, format_2_id))
+
                     # Formats must be opposite (video+audio)
-                    if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none':
+                    if format_1.get('acodec') == 'none' and format_2.get('acodec') == 'none':
                         self.report_error(
                             'Both formats %s and %s are video-only, you must specify "-f video+audio"'
-                            % (format_1, format_2))
+                            % (format_1_id, format_2_id))
                         return
                     output_ext = (
-                        formats_info[0]['ext']
+                        format_1['ext']
                         if self.params.get('merge_output_format') is None
                         else self.params['merge_output_format'])
                     return {
                         'requested_formats': formats_info,
-                        'format': '%s+%s' % (formats_info[0].get('format'),
-                                             formats_info[1].get('format')),
-                        'format_id': '%s+%s' % (formats_info[0].get('format_id'),
-                                                formats_info[1].get('format_id')),
-                        'width': formats_info[0].get('width'),
-                        'height': formats_info[0].get('height'),
-                        'resolution': formats_info[0].get('resolution'),
-                        'fps': formats_info[0].get('fps'),
-                        'vcodec': formats_info[0].get('vcodec'),
-                        'vbr': formats_info[0].get('vbr'),
-                        'stretched_ratio': formats_info[0].get('stretched_ratio'),
-                        'acodec': formats_info[1].get('acodec'),
-                        'abr': formats_info[1].get('abr'),
+                        'format': '%s+%s' % (format_1.get('format'),
+                                             format_2.get('format')),
+                        'format_id': '%s+%s' % (format_1.get('format_id'),
+                                                format_2.get('format_id')),
+                        'width': format_1.get('width'),
+                        'height': format_1.get('height'),
+                        'resolution': format_1.get('resolution'),
+                        'fps': format_1.get('fps'),
+                        'vcodec': format_1.get('vcodec'),
+                        'vbr': format_1.get('vbr'),
+                        'stretched_ratio': format_1.get('stretched_ratio'),
+                        'acodec': format_2.get('acodec'),
+                        'abr': format_2.get('abr'),
                         'ext': output_ext,
                     }
 

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/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/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: