feat(cli): accept YouTube URLs in addition to video IDs

youtube_transcript_api/_cli.py

@@ -1,4 +1,5 @@
 import argparse
+import re
 from importlib.metadata import PackageNotFoundError, version
 from typing import List
 
@@ -199,5 +200,22 @@ def _parse_args(self):
         return self._sanitize_video_ids(parser.parse_args(self._args))
 
     def _sanitize_video_ids(self, args):
-        args.video_ids = [video_id.replace("\\", "") for video_id in args.video_ids]
+        sanitized = []
+        for video_id in args.video_ids:
+            video_id = video_id.replace("\\", "")
+            # Detect YouTube URLs and extract the video ID
+            url_patterns = [
+                r"(?:https?://)?(?:www\.)?youtube\.com/watch\?(?:.*&)?v=([\w-]{11})",
+                r"(?:https?://)?youtu\.be/([\w-]{11})",
+                r"(?:https?://)?(?:www\.)?youtube\.com/embed/([\w-]{11})",
+                r"(?:https?://)?(?:www\.)?youtube\.com/v/([\w-]{11})",
+            ]
+            extracted = None
+            for pattern in url_patterns:
+                match = re.search(pattern, video_id)
+                if match:
+                    extracted = match.group(1)
+                    break
+            sanitized.append(extracted if extracted else video_id)
+        args.video_ids = sanitized
         return args

youtube_transcript_api/test/test_cli.py

@@ -139,6 +139,49 @@ def test_argument_parsing__video_ids_starting_with_dash(self):
         self.assertEqual(parsed_args.format, "pretty")
         self.assertEqual(parsed_args.languages, ["en"])
 
+    def test_argument_parsing__youtube_watch_url(self):
+        parsed_args = YouTubeTranscriptCli(
+            ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"]
+        )._parse_args()
+        self.assertEqual(parsed_args.video_ids, ["dQw4w9WgXcQ"])
+
+    def test_argument_parsing__youtu_be_url(self):
+        parsed_args = YouTubeTranscriptCli(
+            ["https://youtu.be/dQw4w9WgXcQ"]
+        )._parse_args()
+        self.assertEqual(parsed_args.video_ids, ["dQw4w9WgXcQ"])
+
+    def test_argument_parsing__youtube_embed_url(self):
+        parsed_args = YouTubeTranscriptCli(
+            ["https://www.youtube.com/embed/dQw4w9WgXcQ"]
+        )._parse_args()
+        self.assertEqual(parsed_args.video_ids, ["dQw4w9WgXcQ"])
+
+    def test_argument_parsing__youtube_v_url(self):
+        parsed_args = YouTubeTranscriptCli(
+            ["https://www.youtube.com/v/dQw4w9WgXcQ"]
+        )._parse_args()
+        self.assertEqual(parsed_args.video_ids, ["dQw4w9WgXcQ"])
+
+    def test_argument_parsing__youtube_url_with_extra_params(self):
+        parsed_args = YouTubeTranscriptCli(
+            ["https://www.youtube.com/watch?v=dQw4w9WgXcQ&t=30s&list=PLtest"]
+        )._parse_args()
+        self.assertEqual(parsed_args.video_ids, ["dQw4w9WgXcQ"])
+
+    def test_argument_parsing__mix_of_ids_and_urls(self):
+        parsed_args = YouTubeTranscriptCli(
+            [
+                "dQw4w9WgXcQ",
+                "https://www.youtube.com/watch?v=82IOSYpY6Qo",
+                "https://youtu.be/abcdefghijk",
+            ]
+        )._parse_args()
+        self.assertEqual(
+            parsed_args.video_ids,
+            ["dQw4w9WgXcQ", "82IOSYpY6Qo", "abcdefghijk"],
+        )
+
     def test_argument_parsing__fail_without_video_ids(self):
         with self.assertRaises(SystemExit):
             YouTubeTranscriptCli("--format json".split())._parse_args()