scripts/deploy_to_hf.py

import os
import sys
import shutil
from huggingface_hub import HfApi, create_repo, file_exists

def deploy(token, space_name):
    username = space_name.split("/")[0]
    dataset_name = f"{username}/medical-qa-knowledge-base"
    api = HfApi(token=token)
    
    print(f"🚀 Deploying to Space: {space_name}")
    print(f"📦 Knowledge Base Dataset: {dataset_name}")
    
    # ---------------------------------------------------------
    # 1. Set Secrets for Qdrant (Secure)
    # ---------------------------------------------------------
    print(f"\n[1/3] Setting Secrets for Qdrant...")
    # These should be passed as args or found in env, but for now using the hardcoded ones user provided
    # (In production, use env vars!)
    qdrant_url = "https://18ab5ca3-4731-430f-baaf-2d35d36953ae.europe-west3-0.gcp.cloud.qdrant.io:6333"
    qdrant_key = "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJhY2Nlc3MiOiJtIn0.DQLeocbeR-S9XO-b2cca8UQL7m3OSZMOHIJGB0gJDhQ"
    
    try:
        api.add_space_secret(repo_id=space_name, key="QDRANT_URL", value=qdrant_url)
        api.add_space_secret(repo_id=space_name, key="QDRANT_API_KEY", value=qdrant_key)
        api.add_space_secret(repo_id=space_name, key="VECTOR_DB_TYPE", value="qdrant")
        print("  ✅ Secrets configured successfully!")
    except Exception as e:
        print(f"⚠️ Failed to set secrets (maybe already set?): {e}")

    # ---------------------------------------------------------
    # 2. Prepare Staging Area (Clean Build)
    # ---------------------------------------------------------
    print("\n[2/3] Preparing clean build in 'deploy_build/'...")
    
    build_dir = "deploy_build"
    if os.path.exists(build_dir):
        shutil.rmtree(build_dir)
    os.makedirs(build_dir, exist_ok=True)
    
    # Copy Application Files
    items_to_copy = [
        "api",
        "src",
        "frontend",
        "scripts",
        "requirements.txt",
        "Dockerfile",
        "evaluation"
    ]
    
    for item in items_to_copy:
        src = item
        dst = os.path.join(build_dir, item)
        if os.path.isdir(src):
            shutil.copytree(src, dst, dirs_exist_ok=True)
        elif os.path.isfile(src):
            shutil.copy2(src, dst)
    
    # Ensure huggingface_hub matches requirements
    with open(os.path.join(build_dir, "requirements.txt"), "a") as f:
        f.write("\nhuggingface_hub[cli]\n")
            
    # Modify Startup Script
    # Read original start.sh
    with open("start.sh", "r") as f:
        original_start = f.read()
        
    download_cmd = f"""
# Set default PORT if not set (for Streamlit)
export PORT=${{PORT:-8501}}

# Qdrant mode - No local KB download needed!
echo "✅ Using Remote Vector Database (Qdrant)"
"""
    # Create start.sh in build dir
    if original_start.startswith("#!"):
        lines = original_start.split("\n")
        new_start = lines[0] + "\n" + download_cmd + "\n" + "\n".join(lines[1:])
    else:
        new_start = "#!/bin/bash\n" + download_cmd + "\n" + original_start
        
    with open(os.path.join(build_dir, "start.sh"), "w") as f:
        f.write(new_start)
    os.chmod(os.path.join(build_dir, "start.sh"), 0o755)
    
    # Create README with Metadata
    metadata = """---
title: MediQuery Healthcare AI
emoji: 🏥
colorFrom: blue
colorTo: indigo
sdk: docker
pinned: false
app_port: 8501
---
"""
    if os.path.exists("README.md"):
        with open("README.md", "r") as f:
            content = f.read()
        if not content.strip().startswith("---"):
            content = metadata + "\n" + content
    else:
        content = metadata + "\n# Healthcare QA Chatbot"
        
    with open(os.path.join(build_dir, "README.md"), "w") as f:
        f.write(content)

    print("  ✅ Staging area ready!")

    # ---------------------------------------------------------
    # 3. Create Space & Upload App Code from Staging
    # ---------------------------------------------------------
    try:
        print(f"\n[3/3] Deploying to Space {space_name}...")
        url = create_repo(
            repo_id=space_name,
            token=token,
            repo_type="space",
            space_sdk="docker",
            exist_ok=True,
            private=False
        )
        print(f"  Space ready: {url}")
        
        print("  📤 Uploading application code...")
        
        api.upload_folder(
            folder_path=build_dir,
            repo_id=space_name,
            repo_type="space"
        )
        
        print("\n✅ Deployment Complete!")
        print(f"🎉 Live at: https://huggingface.co/spaces/{space_name}")
        
    except Exception as e:
        print(f"\n❌ Deployment failed: {e}")
    finally:
        # Cleanup
        if os.path.exists(build_dir):
            shutil.rmtree(build_dir)

if __name__ == "__main__":
    if len(sys.argv) < 3:
        print("Usage: python deploy_to_hf.py <token> <space_name>")
        sys.exit(1)
        
    deploy(sys.argv[1], sys.argv[2])