Use different downloader for youtube

docs/run_patches.sh

@@ -15,12 +15,12 @@ sed -i  's/with HiddenPrints():/if True:/g' $sp/langchain_community/utilities/se
 # find "$sp" -type f -name "*.py" -exec sed -i -E 's/(sys\.stdout\s*=\s*.*)/pass # \1/; s/(sys\.stderr\s*=\s*.*)/pass # \1/' {} +
 
 # use pytubefix instead, pytube too old and various issues
-sed -i 's/Pytube/PytubeFix/g'  $sp/fiftyone/utils/youtube.py
-sed -i 's/pytube>=15/pytube>=6/g' $sp/fiftyone/utils/youtube.py
-sed -i 's/pytube/pytubefix/g' $sp/fiftyone/utils/youtube.py
+#sed -i 's/Pytube/PytubeFix/g'  $sp/fiftyone/utils/youtube.py
+#sed -i 's/pytube>=15/pytube>=6/g' $sp/fiftyone/utils/youtube.py
+#sed -i 's/pytube/pytubefix/g' $sp/fiftyone/utils/youtube.py
 
 # diff -Naru /home/jon/miniconda3/envs/h2ogpt/lib/python3.10/site-packages/pytubefix/extract.py ~/extract.py > docs/pytubefix.patch
-patch $sp/pytubefix/extract.py docs/pytubefix.patch
+#patch $sp/pytubefix/extract.py docs/pytubefix.patch
 
 # fix asyncio same way websockets was fixed, else keep hitting errors in async calls
 # https://github.com/python-websockets/websockets/commit/f9fd2cebcd42633ed917cd64e805bea17879c2d7

reqs_optional/requirements_optional_image.txt

@@ -7,4 +7,4 @@ yt-dlp>=2024.8.6
 # moviepy>=0.5.1
 
 # for fiftyone with patches
-pytubefix==6.15.4
+pytubefix==8.1.1

src/version.py

@@ -1 +1 @@
-__version__ = "028f433b69021436a7ffd5f26970eb7ec0bfbfa7"
+__version__ = "f28e728d5e5a3f67c2642641e2743d2c7fbe7aea"

src/vision/extract_movie.py

@@ -1,4 +1,5 @@
 import os
+import sys
 import uuid
 
 from src.utils import makedirs, sanitize_filename, get_gradio_tmp
@@ -14,8 +15,13 @@ def extract_unique_frames(urls=None, file=None, download_dir=None, export_dir=No
         makedirs(download_dir, exist_ok=True)
     # os.environ['FIFTYONE_DISABLE_SERVICES'] = 'True'
     if urls:
-        import fiftyone.utils.youtube as fouy
-        fouy.download_youtube_videos(urls, download_dir=download_dir)
+        if 'openai_server' not in sys.path:
+            sys.path.append('openai_server')
+        from openai_server.agent_tools.download_web_video import download_web_video
+        for url in urls:
+            download_web_video(video_url=url, base_url="https://www.youtube.com", output_dir=download_dir)
+        #import fiftyone.utils.youtube as fouy
+        #fouy.download_youtube_videos(urls, download_dir=download_dir)
 
     # Create a FiftyOne Dataset
     import fiftyone as fo