[extractor/zoom] Fix share URL extraction (#6789)

Authored by: bashonly

yt_dlp/extractor/zoom.py

@@ -13,8 +13,8 @@
 
 class ZoomIE(InfoExtractor):
     IE_NAME = 'zoom'
-    _VALID_URL = r'(?P<base_url>https?://(?:[^.]+\.)?zoom.us/)rec(?:ording)?/(?:play|share)/(?P<id>[A-Za-z0-9_.-]+)'
+    _VALID_URL = r'(?P<base_url>https?://(?:[^.]+\.)?zoom.us/)rec(?:ording)?/(?P<type>play|share)/(?P<id>[A-Za-z0-9_.-]+)'
-    _TEST = {
+    _TESTS = [{
         'url': 'https://economist.zoom.us/rec/play/dUk_CNBETmZ5VA2BwEl-jjakPpJ3M1pcfVYAPRsoIbEByGsLjUZtaa4yCATQuOL3der8BlTwxQePl_j0.EImBkXzTIaPvdZO5',
         'md5': 'ab445e8c911fddc4f9adc842c2c5d434',
         'info_dict': {
@@ -23,17 +23,37 @@ class ZoomIE(InfoExtractor):
             'title': 'China\'s "two sessions" and the new five-year plan',
         },
         'skip': 'Recording requires email authentication to access',
-    }
+    }, {
+        # play URL
+        'url': 'https://ffgolf.zoom.us/rec/play/qhEhXbrxq1Zoucx8CMtHzq1Z_2YZRPVCqWK_K-2FkEGRsSLDeOX8Tu4P6jtjZcRry8QhIbvKZdtr4UNo.QcPn2debFskI9whJ',
+        'md5': '2c4b1c4e5213ebf9db293e88d9385bee',
+        'info_dict': {
+            'id': 'qhEhXbrxq1Zoucx8CMtHzq1Z_2YZRPVCqWK_K-2FkEGRsSLDeOX8Tu4P6jtjZcRry8QhIbvKZdtr4UNo.QcPn2debFskI9whJ',
+            'ext': 'mp4',
+            'title': 'Prépa AF2023 - Séance 5 du 11 avril - R20/VM/GO',
+        },
+    }, {
+        # share URL
+        'url': 'https://us02web.zoom.us/rec/share/hkUk5Zxcga0nkyNGhVCRfzkA2gX_mzgS3LpTxEEWJz9Y_QpIQ4mZFOUx7KZRZDQA.9LGQBdqmDAYgiZ_8',
+        'md5': '90fdc7cfcaee5d52d1c817fc03c43c9b',
+        'info_dict': {
+            'id': 'hkUk5Zxcga0nkyNGhVCRfzkA2gX_mzgS3LpTxEEWJz9Y_QpIQ4mZFOUx7KZRZDQA.9LGQBdqmDAYgiZ_8',
+            'ext': 'mp4',
+            'title': 'Timea Andrea Lelik\'s Personal Meeting Room',
+        },
+    }]
 
-    def _real_extract(self, url):
+    def _get_page_data(self, webpage, video_id):
-        base_url, play_id = self._match_valid_url(url).groups()
+        return self._search_json(
-        webpage = self._download_webpage(url, play_id)
+            r'window\.__data__\s*=', webpage, 'data', video_id, transform_source=js_to_json)
 
+    def _get_real_webpage(self, url, base_url, video_id, url_type):
+        webpage = self._download_webpage(url, video_id, note=f'Downloading {url_type} webpage')
         try:
             form = self._form_hidden_inputs('password_form', webpage)
         except ExtractorError:
-            form = None
+            return webpage
-        if form:
+
         password = self.get_param('videopassword')
         if not password:
             raise ExtractorError(
@@ -41,21 +61,35 @@ def _real_extract(self, url):
         is_meeting = form.get('useWhichPasswd') == 'meeting'
         validation = self._download_json(
             base_url + 'rec/validate%s_passwd' % ('_meet' if is_meeting else ''),
-                play_id, 'Validating passcode', 'Wrong passcode', data=urlencode_postdata({
+            video_id, 'Validating passcode', 'Wrong passcode', data=urlencode_postdata({
                 'id': form[('meet' if is_meeting else 'file') + 'Id'],
                 'passwd': password,
                 'action': form.get('action'),
             }))
         if not validation.get('status'):
             raise ExtractorError(validation['errorMessage'], expected=True)
-            webpage = self._download_webpage(url, play_id)
+        return self._download_webpage(url, video_id, note=f'Re-downloading {url_type} webpage')
 
-        data = self._parse_json(self._search_regex(
+    def _real_extract(self, url):
-            r'(?s)window\.__data__\s*=\s*({.+?});',
+        base_url, url_type, video_id = self._match_valid_url(url).group('base_url', 'type', 'id')
-            webpage, 'data'), play_id, js_to_json)
+
+        if url_type == 'share':
+            webpage = self._get_real_webpage(url, base_url, video_id, 'share')
+            meeting_id = self._get_page_data(webpage, video_id)['meetingId']
+            redirect_path = self._download_json(
+                f'{base_url}nws/recording/1.0/play/share-info/{meeting_id}',
+                video_id, note='Downloading share info JSON')['result']['redirectUrl']
+            url = urljoin(base_url, redirect_path)
+
+        webpage = self._get_real_webpage(url, base_url, video_id, 'play')
+        file_id = self._get_page_data(webpage, video_id)['fileId']
+        if not file_id:
+            # When things go wrong, file_id can be empty string
+            raise ExtractorError('Unable to extract file ID')
 
         data = self._download_json(
-            f'{base_url}nws/recording/1.0/play/info/{data["fileId"]}', play_id)['result']
+            f'{base_url}nws/recording/1.0/play/info/{file_id}', video_id,
+            note='Downloading play info JSON')['result']
 
         subtitles = {}
         for _type in ('transcript', 'cc', 'chapter'):
@@ -91,7 +125,7 @@ def _real_extract(self, url):
             })
 
         return {
-            'id': play_id,
+            'id': video_id,
             'title': str_or_none(traverse_obj(data, ('meet', 'topic'))),
             'subtitles': subtitles,
             'formats': formats,