luciano/markitdown
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

FIX YouTube transcript errors

Browse source
This commit is contained in:
Josh 2025-05-08 14:11:09 -07:00
parent 041be54471
commit 2d92e94608
1 changed files with 19 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

18
packages/markitdown/src/markitdown/converters/_youtube_converter.py
View file

@ -153,9 +153,14 @@ class YouTubeConverter(DocumentConverter):
params = parse_qs(parsed_url.query) # type: ignore params = parse_qs(parsed_url.query) # type: ignore
if "v" in params and params["v"][0]: if "v" in params and params["v"][0]:
video_id = str(params["v"][0]) video_id = str(params["v"][0])
transcript_list = ytt_api.list(video_id)
languages = ['en']
for transcript in transcript_list:
languages.append(transcript.language_code)
break
try: try:
youtube_transcript_languages = kwargs.get( youtube_transcript_languages = kwargs.get(
"youtube_transcript_languages", ("en",) "youtube_transcript_languages", languages
) )
# Retry the transcript fetching operation # Retry the transcript fetching operation
transcript = self._retry_operation( transcript = self._retry_operation(
@ -165,12 +170,21 @@ class YouTubeConverter(DocumentConverter):
retries=3, # Retry 3 times retries=3, # Retry 3 times
delay=2, # 2 seconds delay between retries delay=2, # 2 seconds delay between retries
) )
if transcript: if transcript:
transcript_text = " ".join( transcript_text = " ".join(
[part.text for part in transcript] [part.text for part in transcript]
) # type: ignore ) # type: ignore
except Exception as e: except Exception as e:
# No transcript available
if len(languages) == 1:
print(f"Error fetching transcript: {e}") print(f"Error fetching transcript: {e}")
else:
# Translate transcript into first kwarg
transcript = transcript_list.find_transcript(languages).translate(youtube_transcript_languages[0]).fetch()
transcript_text = " ".join(
[part.text for part in transcript]
)
if transcript_text: if transcript_text:
webpage_text += f"\n### Transcript\n{transcript_text}\n" webpage_text += f"\n### Transcript\n{transcript_text}\n"
@ -222,3 +236,5 @@ class YouTubeConverter(DocumentConverter):
attempt += 1 attempt += 1
# If all attempts fail, raise the last exception # If all attempts fail, raise the last exception
raise Exception(f"Operation failed after {retries} attempts.") raise Exception(f"Operation failed after {retries} attempts.")