ask_question_to_pdf.py

# https://github.com/andrewyng/aisuite/blob/main/examples/QnA_with_pdf.ipynb

import requests
import fitz
from io import BytesIO

# Link to paper in pdf format on the cost of avocados.
pdf_path = "https://arxiv.org/pdf/2104.04649"
pdf_text = ""
# Download PDF and load it into memory
response = requests.get(pdf_path)
if response.status_code == 200:
    pdf_data = BytesIO(response.content)  # Load PDF data into BytesIO
    # Open PDF from memory using fitz
    with fitz.open(stream=pdf_data, filetype="pdf") as pdf:
        text = ""
        for page_num in range(pdf.page_count):
            page = pdf[page_num]
            pdf_text += page.get_text("text")  # Extract text
            pdf_text += "\n" + "="*50 + "\n"  # Separator for each page
    print("Downloaded and extracted text from pdf.")
else:
    print(f"Failed to download PDF: {response.status_code}")

question = "Is the price of organic avocados higher than non-organic avocados? What has been the trend?"

import aisuite as ai
client = ai.Client()
messages = [
    {"role": "system", "content": "You are a helpful assistant. Answer the question only based on the below text."},
    {"role": "user", "content": f"Answer the question based on the following text:\n\n{pdf_text}\n\nQuestion: {question}\n"},
]