app.py

from fastapi import FastAPI, HTTPException
from typing import Optional, List
from pydantic import BaseModel
import pandas as pd
from datasets import Dataset, load_dataset
from huggingface_hub import upload_file
import os

# Configuration - MUST come first
DATASET_NAME = "Ritesh1035/Customer_Data"  # Your HuggingFace dataset
HF_TOKEN = os.getenv("HF_TOKEN")  # Set this in HF Space secrets

app = FastAPI(title="Customer API", description="Customer management API with HuggingFace Dataset storage")

@app.on_event("startup")
async def startup_event():
    """Test connectivity on startup"""
    print(f"🚀 Starting Customer API...")
    print(f"📊 Dataset: {DATASET_NAME}")
    print(f"🔑 HF Token: {'✅ Set' if HF_TOKEN else '❌ Not set'}")

    # Test dataset connectivity
    try:
        customers = load_customers_from_dataset()
        print(f"✅ Successfully connected to dataset with {len(customers)} customers")
    except Exception as e:
        print(f"⚠️  Dataset connection warning: {e}")

    print("🎉 API is ready!")

@app.get("/")
def read_root():
    hf_token_status = "✅ Set" if HF_TOKEN else "❌ Not set"
    return {
        "message": "Customer API is running! Visit /docs for API documentation.",
        "dataset": DATASET_NAME,
        "hf_token_status": hf_token_status
    }

@app.get("/health")
def health_check():
    """Health check endpoint"""
    return {"status": "healthy", "service": "Customer API"}

# 1 Define blueprint
class Customer(BaseModel):
    id:int
    name: str
    email: str
    phone: Optional[str]=None
    address: Optional[str]=None

# 2 HuggingFace Dataset Functions

def load_customers_from_dataset() -> List[Customer]:
    """Load customers from HF Dataset"""
    try:
        print(f"📊 Loading dataset: {DATASET_NAME}")

        # Load dataset with minimal options
        dataset = load_dataset(
            DATASET_NAME,
            split="train",
            token=HF_TOKEN
        )

        print(f"✅ Dataset loaded: {len(dataset)} rows")
        customers = []

        for row in dataset:
            try:
                customers.append(Customer(
                    id=int(row['id']),
                    name=str(row['name']),
                    email=str(row['email']),
                    phone=str(row.get('phone', '')) if row.get('phone') and row.get('phone') != '' else None,
                    address=str(row.get('address', '')) if row.get('address') and row.get('address') != '' else None
                ))
            except Exception as row_error:
                print(f"⚠️ Skipping invalid row: {row_error}")
                continue

        print(f"📋 Loaded {len(customers)} customers")
        return customers

    except Exception as e:
        print(f"⚠️ Load error: {e}")
        return []

def save_customers_to_dataset(customers: List[Customer]):
    """Save customers to HF Dataset"""
    try:
        # Convert to pandas DataFrame
        data = []
        for customer in customers:
            data.append({
                'id': customer.id,
                'name': customer.name,
                'email': customer.email,
                'phone': customer.phone or "",
                'address': customer.address or ""
            })

        df = pd.DataFrame(data)
        print(f"💾 Saving {len(customers)} customers to dataset...")

        # Save locally first as backup
        df.to_csv("customers.csv", index=False)
        print("📁 Local backup saved")

        # Upload to HF Dataset if token is available
        if HF_TOKEN:
            try:
                upload_file(
                    path_or_fileobj="customers.csv",
                    path_in_repo="customers.csv",
                    repo_id=DATASET_NAME,
                    repo_type="dataset",
                    token=HF_TOKEN
                )
                print(f"✅ Successfully saved {len(customers)} customers to dataset")

                # Small delay for sync
                import time
                time.sleep(1)

            except Exception as e:
                print(f"⚠️ Upload error: {e}")
                # Don't fail the operation, just log the error
        else:
            print("⚠️ HF_TOKEN not set - data saved locally only")

    except Exception as e:
        print(f"❌ Error in save_customers_to_dataset: {e}")
        # Don't raise exception to avoid breaking the API
# Create
@app.post("/Customer", response_model=Customer)
def create_customer(customer: Customer):
    print(f"➕ Creating customer ID: {customer.id}")
    customers = load_customers_from_dataset()
    print(f"📊 Current customers: {len(customers)}")

    # Check for unique ID
    existing_ids = [c.id for c in customers]
    if customer.id in existing_ids:
        print(f"❌ ID {customer.id} already exists")
        raise HTTPException(
            status_code=400,
            detail=f"Customer ID {customer.id} already exists"
        )

    customers.append(customer)
    save_customers_to_dataset(customers)
    print(f"✅ Created customer {customer.id}")
    return customer

# Read
@app.get("/Customer", response_model=List[Customer])
def get_customer():
    print("📋 Getting all customers")
    customers = load_customers_from_dataset()
    print(f"📊 Found {len(customers)} customers")
    return customers

# Update
@app.put("/Customer/{id}", response_model=Customer)
def update_customer(id: int, customer: Customer):
    customers = load_customers_from_dataset()
    for i, existing_customer in enumerate(customers):
        if existing_customer.id == id:
            customers[i] = customer
            save_customers_to_dataset(customers)
            return customer
    raise HTTPException(status_code=404, detail="customer not found")

# Delete
@app.delete("/Customer/{id}", response_model=Customer)
def delete_customer(id: int):
    customers = load_customers_from_dataset()
    for i, customer_obj in enumerate(customers):
        if customer_obj.id == id:
            deleted_customer = customers.pop(i)
            save_customers_to_dataset(customers)
            return deleted_customer
    raise HTTPException(status_code=404, detail="customer not found")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)