main.py

from fastapi import FastAPI, File, UploadFile, Form
import pandas as pd
from google.cloud import storage
import io
import os

app = FastAPI()

# GCS credentials
gcs_credentials_file = "ow-stu-us-ce1-dev.json"
gcs_bucket_name = "ow-stu-us-ce1-ai-platform"

# Ensure the GCS bucket exists
gcs_client = storage.Client.from_service_account_json(gcs_credentials_file)
gcs_bucket = gcs_client.bucket(gcs_bucket_name)

# File path in GCS bucket
gcs_file_path = "deepak_6593/db.csv"

def append_to_gcs_csv(new_data, gcs_file_path):
    # Standardize column names for new data
    new_data.columns = ['category', 'score']

    # Check if the file exists in GCS bucket
    blob = gcs_bucket.blob(gcs_file_path)
    exists = blob.exists()

    # If file exists, read it into a DataFrame
    if exists:
        existing_data = pd.read_csv(io.BytesIO(blob.download_as_bytes()))
        # Ensure existing data has the right columns
        existing_data = existing_data[['category', 'score']]
        # Append new data to existing data
        combined_data = pd.concat([existing_data, new_data], ignore_index=True)
    else:
        combined_data = new_data

    # Convert combined DataFrame to CSV and upload it
    csv_data = combined_data.to_csv(index=False).encode('utf-8')
    blob.upload_from_string(csv_data, content_type='text/csv')

def read_from_gcs_csv(gcs_file_path):
    blob = gcs_bucket.blob(gcs_file_path)
    return pd.read_csv(io.BytesIO(blob.download_as_text()))

@app.post("/upload-file/")
async def upload_file(file: UploadFile = File(...)):
    df = pd.read_csv(io.StringIO((await file.read()).decode('utf-8')))
    append_to_gcs_csv(df, gcs_file_path)
    return {"message": "File uploaded successfully"}

@app.post("/upload-data/")
async def upload_data(category: str = Form(...), score: float = Form(...)):
    df = pd.DataFrame([[category, score]], columns=['category', 'score'])
    append_to_gcs_csv(df, gcs_file_path)
    return {"message": "Data uploaded successfully"}

@app.post("/clear-data/")
async def clear_data():
    # Create an empty DataFrame with the same columns
    empty_df = pd.DataFrame(columns=['category', 'score'])
    # Convert the empty DataFrame to CSV
    csv_data = empty_df.to_csv(index=False).encode('utf-8')
    # Overwrite the existing file in GCS with the empty CSV data
    gcs_bucket.blob(gcs_file_path).upload_from_string(csv_data, content_type='text/csv')
    return {"message": "Data cleared successfully"}