app.py

import gradio as gr
import fitz  # PyMuPDF
import openai
import os

# Load API key from Hugging Face secret
openai.api_key = os.getenv("OPENAI_API_KEY")

# Extract text from PDF
def extract_text_from_pdf(file):
    text = ""
    with fitz.open(stream=file.read(), filetype="pdf") as doc:
        for page in doc:
            text += page.get_text()
    return text

# Split text into chunks
def split_text(text, max_tokens=1500):
    import textwrap
    return textwrap.wrap(text, max_tokens)

# Generate summary
def generate_summary(text):
    chunks = split_text(text)
    summaries = []
    for i, chunk in enumerate(chunks):
        print(f"⏳ Summarizing chunk {i + 1}/{len(chunks)}...")
        response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=[{"role": "user", "content": f"Summarize the following:\n{chunk}"}],
            temperature=0.5,
        )
        summaries.append(response['choices'][0]['message']['content'].strip())
    return "\n\n".join(summaries)

# Generate all formats
def summarize_paper(pdf_file):
    try:
        raw_text = extract_text_from_pdf(pdf_file)
        summary = generate_summary(raw_text)

        response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=[
                {"role": "user", "content": f"Make an ELI5 version of this:\n{summary}"}
            ],
            temperature=0.7,
        )
        eli5 = response['choices'][0]['message']['content'].strip()

        response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=[
                {"role": "user", "content": f"Write a TL;DR of this:\n{summary}"}
            ],
            temperature=0.7,
        )
        tldr = response['choices'][0]['message']['content'].strip()

        response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=[
                {"role": "user", "content": f"Why does this research matter? {summary}"}
            ],
            temperature=0.7,
        )
        why_it_matters = response['choices'][0]['message']['content'].strip()

        return summary, eli5, tldr, why_it_matters
    except Exception as e:
        return f"❌ Error: {str(e)}", "", "", ""

# Gradio Interface
with gr.Blocks() as demo:
    gr.Markdown("📄 **Paper News Summarizer**\nUpload a research paper PDF and get a human-friendly summary, ELI5, TL;DR, and why it matters.")
    with gr.Row():
        pdf_file = gr.File(label="Upload Research Paper (PDF)", file_types=[".pdf"])
    submit_btn = gr.Button("Submit")
    clear_btn = gr.Button("Clear")

    with gr.Row():
        full_summary = gr.Textbox(label="📘 Full Summary", lines=10, interactive=False)
        eli5_summary = gr.Textbox(label="🧒 ELI5", lines=5, interactive=False)
    with gr.Row():
        tldr_summary = gr.Textbox(label="⚡ TL;DR", lines=2, interactive=False)
        importance = gr.Textbox(label="🎯 Why It Matters", lines=5, interactive=False)

    submit_btn.click(summarize_paper, inputs=pdf_file, outputs=[full_summary, eli5_summary, tldr_summary, importance])
    clear_btn.click(lambda: ("", "", "", ""), outputs=[full_summary, eli5_summary, tldr_summary, importance])

demo.launch()