Merge 360104ec182f9a9545ed0ea64a77ec7ec749c103 into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

Add very rudimentary extractor for Telegram. Supports only public and
channels which are previawable with browser.

Implements #28748.

docs/supportedsites.md

@@ -934,6 +934,7 @@
  - **TeleBruxelles**
  - **Telecinco**: telecinco.es, cuatro.com and mediaset.es
  - **Telegraaf**
+ - **Telegram**
  - **TeleMB**
  - **TeleQuebec**
  - **TeleQuebecEmission**

youtube_dl/extractor/extractors.py

@@ -1263,6 +1263,7 @@ from .tele13 import Tele13IE
 from .telebruxelles import TeleBruxellesIE
 from .telecinco import TelecincoIE
 from .telegraaf import TelegraafIE
+from .telegram import TelegramIE
 from .telemb import TeleMBIE
 from .telequebec import (
     TeleQuebecIE,

youtube_dl/extractor/telegram.py

@@ -0,0 +1,60 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    ExtractorError,
+    int_or_none,
+)
+
+
+class TelegramIE(InfoExtractor):
+    _VALID_URL = r'https://t\.me/(?P<user>[^/]+)/(?P<id>\d+)'
+    _TEST = {
+        'url': 'https://t.me/telegram/195',
+        'info_dict': {
+            'id': '195',
+            'ext': 'mp4',
+            'title': 'telegram',
+            'description': 'Telegram’s Bot Documentation has been completely overhauled –\xa0adding the latest info, along with detailed screenshots and videos.\n\nNewcomers now have an easy way to learn about all the powerful features, and can build a bot from our step-by-step tutorial with examples for popular programming languages.\n\nExperienced developers can explore recent updates and advanced features, ready for 2022 and beyond.',
+            'duration': 23,
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        m = re.match(r'https://t\.me/(?P<channel>[^/]+)/', url)
+        if m is None:
+            raise ExtractorError('Unable to find channel name')
+        title = m.group('channel')
+        embed_url = url + '?embed=1&mode=tme'
+        html = self._download_webpage(embed_url, video_id)
+
+        video_url = self._search_regex(r'<video src="([^"]+)"', html, 'video_url')
+        formats = [{'url': video_url}]
+
+        duration = self._search_regex(
+            r'<time class="message_video_duration.*?>(\d+:\d+)<', html,
+            'duration', fatal=False)
+        if duration:
+            try:
+                mins, secs = duration.split(':')
+                secs = int_or_none(secs)
+                mins = int_or_none(mins)
+                duration = None if secs is None or mins is None else secs + 60 * mins
+            except ValueError:
+                duration = None
+
+        description = self._html_search_regex(
+            r'<div class="tgme_widget_message_text.*?>(.+?)</div>', html,
+            'description', fatal=False)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'duration': duration,
+            'formats': formats
+        }