helper_functions.py

import PyPDF2
import tiktoken
# Ensure NLTK resources are available


def extract_text_from_pdf(file):
    """
    Extracts text from a PDF file and tracks text by page.

    :param file: Uploaded PDF file object.
    :return: Tuple (text, page_texts), where:
             - text is the combined text of the entire PDF.
             - page_texts is a list of tuples [(page_number, page_text), ...].
    """
    pdf_reader = PyPDF2.PdfReader(file)
    text = ""
    page_texts = []
    for i, page in enumerate(pdf_reader.pages):
        page_content = page.extract_text()
        text += page_content
        page_texts.append((i + 1, page_content))  # Track page numbers (1-indexed)
    return text, page_texts

def count_tokens(string: str) -> int:
    """Returns the number of tokens in a text string."""
    encoding = tiktoken.get_encoding("o200k_base")
    return len(encoding.encode(string))