upload.py

import requests
import pandas as pd
import io
from fpdf import FPDF
import json

def convert_xlsx_to_pdf(file):
    """Converts an XLSX file to a PDF and returns a BytesIO object with a filename."""
    excel_data = pd.ExcelFile(file)
    pdf = FPDF()
    pdf.set_auto_page_break(auto=True, margin=15)
    pdf.add_page()
    pdf.set_font("Arial", size=12)
    
    for sheet_name in excel_data.sheet_names:
        pdf.cell(200, 10, txt=f"Sheet: {sheet_name}", ln=True, align='C')
        pdf.ln(10)
        df = excel_data.parse(sheet_name)
        
        for i in range(min(10, len(df))):  # Limiting rows for readability
            row_data = " | ".join(str(x) for x in df.iloc[i])
            pdf.multi_cell(0, 10, row_data)
            
        pdf.ln(5)
    
    pdf_output = io.BytesIO()
    pdf_output.write(pdf.output(dest='S').encode('latin1'))  # Convert to bytes
    pdf_output.seek(0)
    
    # Manually add a 'name' attribute to mimic a file
    pdf_output.name = file.name.replace(".xlsx", ".pdf")
    
    return pdf_output

def upload_file_to_vectara(file, customer_id, api_key, corpus_key):
    """Uploads a file to Vectara API v2."""
    url = f"https://api.vectara.io/v2/corpora/{corpus_key}/upload_file"
    headers = {
        "customer-id": customer_id,
        "x-api-key": api_key,
        "Accept": "application/json"
    }
    
    metadata = {"type_file": "excel"} if file.name.endswith('.xlsx') else {}
    
    if file.name.endswith('.xlsx'):
        file = convert_xlsx_to_pdf(file)  # Convert XLSX to PDF
    
    files = {
        'metadata': (None, json.dumps(metadata), 'application/json'),
        "file": (file.name, file.getvalue())}  # Now file.name exists
    
    
    response = requests.post(url, headers=headers, files=files)
    
    return response.json()