Skip to content

Building AI agents, atomically

Fix YouTube transcript tool for youtube-transcript-api 1.x instance API

Eigenwised8c64ae83ef3
diff --git a/atomic-forge/tools/youtube_transcript_scraper/tests/test_youtube_transcript_scraper.py b/atomic-forge/tools/youtube_transcript_scraper/tests/test_youtube_transcript_scraper.py
index 0259a07..14d6ca0 100644
--- a/atomic-forge/tools/youtube_transcript_scraper/tests/test_youtube_transcript_scraper.py
+++ b/atomic-forge/tools/youtube_transcript_scraper/tests/test_youtube_transcript_scraper.py
@@ -1,10 +1,10 @@
import os
import sys
import pytest
-from unittest.mock import patch, MagicMock
+from unittest.mock import ANY, patch, MagicMock
from datetime import datetime
-from youtube_transcript_api import NoTranscriptFound, TranscriptsDisabled
+from youtube_transcript_api import FetchedTranscript, FetchedTranscriptSnippet, NoTranscriptFound, TranscriptsDisabled
# Add the parent directory of 'tests' to the Python path
sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), "..")))
@@ -16,15 +16,26 @@ from tool.youtube_transcript_scraper import ( # noqa: E402
YouTubeTranscriptToolConfig,
)
+FETCH_TARGET = "tool.youtube_transcript_scraper.YouTubeTranscriptApi.fetch"
+
+
+def fetched_rick_astley_transcript() -> FetchedTranscript:
+ return FetchedTranscript(
+ snippets=[
+ FetchedTranscriptSnippet(text="Never gonna give you up", start=0.0, duration=5.0),
+ FetchedTranscriptSnippet(text="Never gonna let you down", start=5.0, duration=5.0),
+ ],
+ video_id="dQw4w9WgXcQ",
+ language="English",
+ language_code="en",
+ is_generated=False,
+ )
+
def test_youtube_transcript_tool():
mock_api_key = "mock_api_key"
mock_video_url = "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
mock_video_id = "dQw4w9WgXcQ"
- mock_transcript = [
- {"text": "Never gonna give you up", "duration": 5.0},
- {"text": "Never gonna let you down", "duration": 5.0},
- ]
mock_metadata = {
"title": "Rick Astley - Never Gonna Give You Up",
"channelTitle": "Rick Astley",
@@ -32,10 +43,7 @@ def test_youtube_transcript_tool():
}
with (
- patch(
- "tool.youtube_transcript_scraper.YouTubeTranscriptApi.get_transcript",
- return_value=mock_transcript,
- ),
+ patch(FETCH_TARGET, autospec=True, return_value=fetched_rick_astley_transcript()) as mock_fetch,
patch("tool.youtube_transcript_scraper.build") as mock_build,
):
mock_youtube = MagicMock()
@@ -46,6 +54,7 @@ def test_youtube_transcript_tool():
input_schema = YouTubeTranscriptToolInputSchema(video_url=mock_video_url)
result = youtube_transcript_tool.run(input_schema)
+ mock_fetch.assert_called_once_with(ANY, mock_video_id)
assert isinstance(result, YouTubeTranscriptToolOutputSchema)
assert result.transcript == "Never gonna give you up Never gonna let you down"
assert result.duration == 10.0
@@ -61,10 +70,6 @@ def test_youtube_transcript_tool():
def test_youtube_transcript_tool_with_language():
mock_api_key = "mock_api_key"
mock_video_url = "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
- mock_transcript = [
- {"text": "Never gonna give you up", "duration": 5.0},
- {"text": "Never gonna let you down", "duration": 5.0},
- ]
mock_metadata = {
"title": "Rick Astley - Never Gonna Give You Up",
"channelTitle": "Rick Astley",
@@ -72,10 +77,7 @@ def test_youtube_transcript_tool_with_language():
}
with (
- patch(
- "tool.youtube_transcript_scraper.YouTubeTranscriptApi.get_transcript",
- return_value=mock_transcript,
- ),
+ patch(FETCH_TARGET, autospec=True, return_value=fetched_rick_astley_transcript()) as mock_fetch,
patch("tool.youtube_transcript_scraper.build") as mock_build,
):
mock_youtube = MagicMock()
@@ -83,9 +85,10 @@ def test_youtube_transcript_tool_with_language():
mock_youtube.videos().list().execute.return_value = {"items": [{"snippet": mock_metadata}]}
youtube_transcript_tool = YouTubeTranscriptTool(YouTubeTranscriptToolConfig(api_key=mock_api_key))
- input_schema = YouTubeTranscriptToolInputSchema(video_url=mock_video_url, language="en")
+ input_schema = YouTubeTranscriptToolInputSchema(video_url=mock_video_url, language="de")
result = youtube_transcript_tool.run(input_schema)
+ mock_fetch.assert_called_once_with(ANY, "dQw4w9WgXcQ", languages=["de"])
assert isinstance(result, YouTubeTranscriptToolOutputSchema)
assert result.transcript == "Never gonna give you up Never gonna let you down"
@@ -95,7 +98,8 @@ def test_youtube_transcript_tool_no_transcript():
mock_video_url = "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
with patch(
- "tool.youtube_transcript_scraper.YouTubeTranscriptApi.get_transcript",
+ FETCH_TARGET,
+ autospec=True,
side_effect=NoTranscriptFound(
video_id="dQw4w9WgXcQ",
requested_language_codes=["en"],
@@ -116,8 +120,9 @@ def test_youtube_transcript_tool_transcripts_disabled():
mock_video_url = "https://www.youtube.com/watch?v=dQw4w9WgXcQ"
with patch(
- "tool.youtube_transcript_scraper.YouTubeTranscriptApi.get_transcript",
- side_effect=TranscriptsDisabled("Transcripts are disabled"),
+ FETCH_TARGET,
+ autospec=True,
+ side_effect=TranscriptsDisabled("dQw4w9WgXcQ"),
):
youtube_transcript_tool = YouTubeTranscriptTool(YouTubeTranscriptToolConfig(api_key=mock_api_key))
input_schema = YouTubeTranscriptToolInputSchema(video_url=mock_video_url)
diff --git a/atomic-forge/tools/youtube_transcript_scraper/tool/youtube_transcript_scraper.py b/atomic-forge/tools/youtube_transcript_scraper/tool/youtube_transcript_scraper.py
index 56832f4..e9eab10 100644
--- a/atomic-forge/tools/youtube_transcript_scraper/tool/youtube_transcript_scraper.py
+++ b/atomic-forge/tools/youtube_transcript_scraper/tool/youtube_transcript_scraper.py
@@ -98,14 +98,16 @@ class YouTubeTranscriptTool(BaseTool[YouTubeTranscriptToolInputSchema, YouTubeTr
Exception: If fetching the transcript fails.
"""
video_id = self.extract_video_id(params.video_url)
+ transcript_api = YouTubeTranscriptApi()
try:
if params.language:
- transcripts = YouTubeTranscriptApi.get_transcript(video_id, languages=[params.language])
+ fetched_transcript = transcript_api.fetch(video_id, languages=[params.language])
else:
- transcripts = YouTubeTranscriptApi.get_transcript(video_id)
+ fetched_transcript = transcript_api.fetch(video_id)
except (NoTranscriptFound, TranscriptsDisabled) as e:
raise Exception(f"Failed to fetch transcript for video '{video_id}': {str(e)}")
+ transcripts = fetched_transcript.to_raw_data()
transcript_text = " ".join([transcript["text"] for transcript in transcripts])
total_duration = sum([transcript["duration"] for transcript in transcripts])