Update app.py

app.py

@@ -2,11 +2,13 @@ import os
 import re
 import json
 import time
-from typing import Dict, Any, List
+from typing import Dict, Any, List, Optional
 from urllib.parse import urlparse, parse_qs
 
+import gradio as gr
 from fastapi import FastAPI, Request, HTTPException
 from fastapi.responses import JSONResponse
+from fastapi.middleware.cors import CORSMiddleware
 
 try:
     from huggingface_hub import HfApi
@@ -22,6 +24,15 @@ os.makedirs(UPLOAD_DIR, exist_ok=True)
 
 app = FastAPI(title="Data Collection Server", description="Receives text/URLs from captioning/image servers, groups by course, compiles JSON and optionally uploads to HuggingFace.")
 
+# Enable CORS for Gradio
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
 # In-memory store for course data
 courses: Dict[str, Dict[str, Any]] = {}
 
@@ -227,7 +238,100 @@ async def debug_course(course: str):
     return courses[course]
 
 
+# Gradio Interface
+def get_course_summary():
+    """Get a formatted summary of all courses and their items"""
+    if not courses:
+        return "No courses available"
+    
+    summary = []
+    for course, data in courses.items():
+        items = data["items"]
+        last_updated = data["last_updated"]
+        last_updated_str = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(last_updated)) if last_updated else "never"
+        
+        summary.append(f"Course: {course}")
+        summary.append(f"Items: {len(items)}")
+        summary.append(f"Last Updated: {last_updated_str}")
+        if items:
+            summary.append("\nLast 3 captions:")
+            for item in items[-3:]:
+                text = item.get("text", "").strip()
+                urls = item.get("urls", [])
+                summary.append(f"- {text[:200]}... ({len(urls)} URLs)")
+        summary.append("\n" + "-"*40 + "\n")
+    
+    return "\n".join(summary)
+
+def compile_course_ui(course_name: str) -> str:
+    """Compile a course from the Gradio UI"""
+    if not course_name:
+        return "Please enter a course name"
+    try:
+        path = compile_course(course_name)
+        return f"Successfully compiled course to {path}"
+    except Exception as e:
+        return f"Error compiling course: {str(e)}"
+
+def add_test_caption(course_name: str, caption_text: str) -> str:
+    """Add a test caption through the Gradio UI"""
+    if not course_name or not caption_text:
+        return "Please enter both course name and caption text"
+    
+    entry = {
+        "timestamp": now_ts(),
+        "text": caption_text,
+        "urls": extract_urls(caption_text),
+        "json": None,
+        "form": {},
+        "headers": {"user-agent": "Gradio UI"},
+    }
+    
+    add_entry(course_name, entry)
+    return f"Added caption to course '{course_name}'. Total items: {len(courses[course_name]['items'])}"
+
+def create_gradio_interface():
+    """Create the Gradio interface tabs and components"""
+    with gr.Blocks(title="Data Collection Server UI") as interface:
+        gr.Markdown("# Data Collection Server Monitor")
+        
+        with gr.Tab("Course Status"):
+            status_output = gr.Textbox(label="Current Status", value="Loading...", lines=15)
+            refresh_btn = gr.Button("Refresh Status")
+            refresh_btn.click(get_course_summary, outputs=status_output)
+        
+        with gr.Tab("Manual Controls"):
+            gr.Markdown("### Compile Course")
+            with gr.Row():
+                compile_course_input = gr.Textbox(label="Course Name")
+                compile_btn = gr.Button("Compile to JSON")
+            compile_output = gr.Textbox(label="Compile Result", lines=3)
+            compile_btn.click(compile_course_ui, inputs=compile_course_input, outputs=compile_output)
+            
+            gr.Markdown("### Add Test Caption")
+            with gr.Row():
+                test_course_input = gr.Textbox(label="Course Name")
+                test_caption_input = gr.Textbox(label="Caption Text", lines=3)
+            test_btn = gr.Button("Add Caption")
+            test_output = gr.Textbox(label="Result", lines=2)
+            test_btn.click(
+                add_test_caption,
+                inputs=[test_course_input, test_caption_input],
+                outputs=test_output
+            )
+    
+    # Update status on load
+    interface.load(get_course_summary, outputs=status_output)
+    return interface
+
+# Mount Gradio app to FastAPI
+interface = create_gradio_interface()
+app = gr.mount_gradio_app(app, interface, path="/ui")
+
 if __name__ == "__main__":
     import uvicorn
     port = int(os.getenv("PORT", "8000"))
+    print(f"[INFO] Server starting on port {port}")
+    print(f"[INFO] FastAPI docs: http://localhost:{port}/docs")
+    print(f"[INFO] Gradio UI: http://localhost:{port}/ui")
     uvicorn.run(app, host="0.0.0.0", port=port)