Upload folder using huggingface_hub

app.py

@@ -4,19 +4,17 @@ import gradio as gr
 from gradio_modal import Modal
 import chromadb
 from modules.collector import fetch_channel_videos_from_url
-from modules.db import delete_channel_from_collection, get_indexed_channels
+from modules.db import (
+    delete_channel_from_collection,
+    get_collection,
+    get_indexed_channels,
+)
 from modules.indexer import index_videos
 from modules.answerer import answer_query, LLMAnswer, VideoItem, build_video_html
 from dotenv import load_dotenv
 
 load_dotenv()
 
-# -------------------------------
-# Setup Chroma
-# -------------------------------
-client = chromadb.PersistentClient(path="./youtube_db")
-collection = client.get_or_create_collection("yt_metadata", embedding_function=None)
-
 
 # -------------------------------
 # Utility functions
@@ -67,7 +65,7 @@ def refresh_channel(api_key, channel_url: str):
     videos = fetch_channel_videos_from_url(api_key, channel_url)
     for v in videos:
         v["channel_url"] = channel_url
-    index_videos(videos, collection, channel_url=channel_url)
+    index_videos(videos, get_collection(), channel_url=channel_url)
     return len(videos)
 
 
@@ -76,13 +74,13 @@ def index_channels(channel_urls: str):
     yt_api_key = os.environ["YOUTUBE_API_KEY"]
     urls = [u.strip() for u in re.split(r"[\n,]+", channel_urls) if u.strip()]
     total_videos = sum(refresh_channel(yt_api_key, url) for url in urls)
-    yield f"✅ Indexed {total_videos} videos from {len(urls)} channels.", list_channels_radio()
+    yield f"✅ Indexed {total_videos} videos from {len(urls)} channels.", gr.update(choices=list_channels_radio())
     return
 
 
 def refresh_all_channels():
     yt_api_key = os.environ["YOUTUBE_API_KEY"]
-    channels = get_indexed_channels(collection)
+    channels = get_indexed_channels(get_collection())
     if not channels:
         return "⚠️ No channels available to refresh.", list_channels_radio()
     total_videos = 0
@@ -100,7 +98,7 @@ def refresh_all_channels():
 # Channel selection as radio
 # -------------------------------
 def list_channels_radio():
-    channels = get_indexed_channels(collection)
+    channels = get_indexed_channels(get_collection())
     choices = []
     for key, val in channels.items():
         if isinstance(val, dict):
@@ -142,14 +140,14 @@ def fetch_channel_html(channel_url: str):
 def delete_channel(channel_url: str):
     delete_channel_from_collection(channel_url)
     # Return updated radio choices
-    return list_channels_radio()
+    return gr.update(choices=list_channels_radio())
 
 
 # -------------------------------
 # LLM query
 # -------------------------------
 def handle_query(query: str):
-    answer_text, video_html = answer_query(query, collection)
+    answer_text, video_html = answer_query(query, get_collection())
     return answer_text, video_html
 
 
@@ -198,7 +196,7 @@ with gr.Blocks() as demo:
                 )
 
                 delete_channel_btn = gr.Button(
-                    "🗑️ Delete", size="sm", scale=0, variant="stop", visible=False
+                    "🗑️ Delete", size="sm", scale=0, variant="stop"
                 )
 
             refresh_status = gr.Markdown(label="Refresh Status", container=False)