app.py

from flask import Flask, render_template, request
import os
from werkzeug.utils import secure_filename
import fitz
from anthropic import Anthropic
from dotenv import load_dotenv
import markdown

load_dotenv()

app = Flask(__name__)
UPLOAD_FOLDER = 'uploads'
ALLOWED_EXTENSIONS = {'pdf'}

app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER

### Helper Functions ###
def allowed_file(filename):
    #   Splits the filename using the furthest right '.' as the delimiter.
    #   Files extensions are always at the end. So [1] will always return the file extension
    return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS

def extract_text(filepath):
    #   Stores the pdf in a document object which can be looped through per page
    pdf_document = fitz.open(filepath)
    extracted_text = ""
    for page in pdf_document:
        extracted_text += page.get_text()
    return extracted_text

#   Summarizing function utilizing Anthropic API
def summarize(text):
    client = Anthropic(
        api_key=os.environ.get("ANTHROPIC_API_KEY"),
    )

    #   Building the message to deliver along with some system assumptions
    message = client.messages.create(
        max_tokens = 1024,
        system = "You are a study assistant. Given the following text from a student's notes or textbook, produce: 1. A concise summary (3-5 sentences) 2. A bullet-point list of the 5-8 most important key points",
        messages=[
            {
                "role": "user",
                "content" : f"Here is the text to summarize: \n\n {text}",
            }
            ],
        model="claude-opus-4-6"
    )

    return message.content[0].text

@app.route("/")
def index():
    return render_template('index.html')

@app.route("/upload", methods=["POST"])
def upload_post():
    #   Guard 1: Check if a file was uploaded or a file with a valid name.
    #   The string argument represents the name attribute of the HTML file input. 
    #   Thet must match to establish a connection
    file = request.files.get('pdf_upload')
    if not file or file.filename == '':
        return render_template('index.html', error="No file submitted")

    #   Guard 2: Ensure the uploaded file is of the correct extension
    if not allowed_file(file.filename):
        return render_template('index.html', error="Only PDF files are allowed")
    
    #   Builds the full path to save the file. Independent of OS
    #   secure_filename cleans potentially malicious file naming. Removes spaces, special characters, etc
    filepath = os.path.join(app.config['UPLOAD_FOLDER'], secure_filename(file.filename))
    file.save(filepath)

    pdf_text = extract_text(filepath)
    print(pdf_text)

    ai_summary = summarize(pdf_text)
    #   Converts Markdown into HTML (since Anthropic reponse is in markdown)
    ai_summary_html = markdown.markdown(ai_summary)
    print(ai_summary)
    #   render_template is a Flask function tha takes an HTML file from templates/
    #   and sends it to the browser as a reponse
    return render_template('index.html', summary = ai_summary_html)