app.py

# inference.py

from pptx import Presentation
import re
from transformers import pipeline

def extract_text_from_pptx(file_path):
    presentation = Presentation(file_path)

    text = []
    for slide_number, slide in enumerate(presentation.slides, start=1):
        for shape in slide.shapes:
            if hasattr(shape, "text"):
                text.append(shape.text)

    return "\n".join(text)

def main():
    file_path = "path/to/your/powerpoint.pptx"  # Specify the path to your PowerPoint file

    extracted_text = extract_text_from_pptx(file_path)
    cleaned_text = re.sub(r'\s+', ' ', extracted_text)

    print(cleaned_text)

    classifier = pipeline("text-classification", model="Ahmed235/roberta_classification")
    summarizer = pipeline("summarization", model="Falconsai/text_summarization")

    result = classifier(cleaned_text)[0]
    predicted_label = result['label']
    predicted_probability = result['score']

    print("Predicted Label:", predicted_label)
    print(f"Evaluate the topic according to {predicted_label} is: {predicted_probability}")
    print(summarizer(cleaned_text, max_length=80, min_length=30, do_sample=False))

if __name__ == "__main__":
    main()