Skip to content

Python tool for converting files and office documents to Markdown.

Support short YouTube URLs in converter (#1882)

Mukunda Rao Katta7de6ce4943b9
diff --git a/packages/markitdown/src/markitdown/converters/_youtube_converter.py b/packages/markitdown/src/markitdown/converters/_youtube_converter.py
index 9ca814c6..c3779743 100644
--- a/packages/markitdown/src/markitdown/converters/_youtube_converter.py
+++ b/packages/markitdown/src/markitdown/converters/_youtube_converter.py
@@ -37,6 +37,24 @@ ACCEPTED_FILE_EXTENSIONS = [
class YouTubeConverter(DocumentConverter):
"""Handle YouTube specially, focusing on the video title, description, and transcript."""
+ def _get_video_id(self, url: str) -> Union[str, None]:
+ """Extract a YouTube video ID from supported URL formats."""
+ parsed_url = urlparse(url)
+ hostname = parsed_url.netloc.lower()
+ path_parts = [part for part in parsed_url.path.split("/") if part]
+
+ if hostname in {"youtu.be", "www.youtu.be"}:
+ return path_parts[0] if path_parts else None
+
+ if hostname in {"youtube.com", "www.youtube.com", "m.youtube.com"}:
+ if path_parts[:1] == ["watch"]:
+ params = parse_qs(parsed_url.query)
+ return params.get("v", [None])[0]
+ if path_parts[:1] in (["shorts"], ["embed"]):
+ return path_parts[1] if len(path_parts) > 1 else None
+
+ return None
+
def accepts(
self,
file_stream: BinaryIO,
@@ -53,7 +71,7 @@ class YouTubeConverter(DocumentConverter):
url = unquote(url)
url = url.replace(r"\?", "?").replace(r"\=", "=")
- if not url.startswith("https://www.youtube.com/watch?"):
+ if not self._get_video_id(url):
# Not a YouTube URL
return False
@@ -146,10 +164,8 @@ class YouTubeConverter(DocumentConverter):
if IS_YOUTUBE_TRANSCRIPT_CAPABLE:
ytt_api = YouTubeTranscriptApi()
transcript_text = ""
- parsed_url = urlparse(stream_info.url) # type: ignore
- params = parse_qs(parsed_url.query) # type: ignore
- if "v" in params and params["v"][0]:
- video_id = str(params["v"][0])
+ video_id = self._get_video_id(stream_info.url or "")
+ if video_id:
transcript_list = ytt_api.list(video_id)
languages = ["en"]
for transcript in transcript_list:
diff --git a/packages/markitdown/tests/test_module_misc.py b/packages/markitdown/tests/test_module_misc.py
index 2fa52f71..1effa4cd 100644
--- a/packages/markitdown/tests/test_module_misc.py
+++ b/packages/markitdown/tests/test_module_misc.py
@@ -23,6 +23,7 @@ from markitdown import (
FileConversionException,
StreamInfo,
)
+from markitdown.converters import YouTubeConverter
# This file contains module tests that are not directly tested by the FileTestVectors.
# This includes things like helper functions and runtime conversion options
@@ -214,6 +215,38 @@ def test_stream_info_operations() -> None:
assert updated_stream_info.url == "url.1"
+ ("url", "video_id"),
+ [
+ ("https://www.youtube.com/watch?v=V2qZ_lgxTzg", "V2qZ_lgxTzg"),
+ ("https://youtu.be/V2qZ_lgxTzg", "V2qZ_lgxTzg"),
+ ("https://www.youtube.com/shorts/V2qZ_lgxTzg", "V2qZ_lgxTzg"),
+ ("https://www.youtube.com/embed/V2qZ_lgxTzg", "V2qZ_lgxTzg"),
+ ],
+)
+def test_youtube_converter_extracts_supported_video_ids(
+ url: str, video_id: str
+) -> None:
+ converter = YouTubeConverter()
+ assert converter._get_video_id(url) == video_id
+
+
+ "url",
+ [
+ "https://www.youtube.com/watch?v=V2qZ_lgxTzg",
+ "https://youtu.be/V2qZ_lgxTzg",
+ "https://www.youtube.com/shorts/V2qZ_lgxTzg",
+ ],
+)
+def test_youtube_converter_accepts_supported_url_formats(url: str) -> None:
+ converter = YouTubeConverter()
+ assert converter.accepts(
+ io.BytesIO(b"<html></html>"),
+ StreamInfo(url=url, extension=".html"),
+ )
+
+
def test_data_uris() -> None:
# Test basic parsing of data URIs
data_uri = "data:text/plain;base64,SGVsbG8sIFdvcmxkIQ=="