app.py

import streamlit as st
from pptx import Presentation
import io

def read_pptx(file):
    """
    Reads a PowerPoint file and extracts text from slides.
    """
    prs = Presentation(file)
    text_content = []
    for i, slide in enumerate(prs.slides):
        slide_text = f"--- Slide {i+1} ---\n"
        for shape in slide.shapes:
            if hasattr(shape, "text"):
                slide_text += shape.text + "\n"
        text_content.append(slide_text)
    return "\n".join(text_content)

st.title("PowerPoint Slide Reader")

uploaded_file = st.file_uploader("Choose a PowerPoint file (.pptx)", type=["pptx"])

if uploaded_file is not None:
    st.success("File uploaded successfully!")

    # To read the file as bytes, which python-pptx expects
    bytes_data = uploaded_file.getvalue()

    # Create a BytesIO object from the bytes data
    # This allows python-pptx to read the file from memory
    pptx_file_like = io.BytesIO(bytes_data)

    try:
        slide_text = read_pptx(pptx_file_like)
        st.subheader("Extracted Text Content:")
        st.text_area("Slide Content", slide_text, height=500)

    except Exception as e:
        st.error(f"Error processing PowerPoint file: {e}")
        st.info("Please ensure it's a valid .pptx file.")