app.py

import gradio as gr
import tempfile
import re
from PyPDF2 import PdfReader, PdfFileReader
import os 
import spacy
import pytesseract
import pdf2image
import subprocess
from pdf2image.exceptions import (
    PDFInfoNotInstalledError,
    PDFPageCountError,
    PDFSyntaxError
)

def clean_text(text):
    nlp = spacy.load("en_core_web_sm", disable=["tagger", "parser", "ner", "textcat"])
    text = re.sub(r'\n+', '\n', text) 
    text = re.sub(r'\s+', ' ', text)
    return text.strip()

def image_to_latex(image):
    image_path = "/tmp/equation.png"  # Modify as needed
    image.save(image_path)
    result = subprocess.run(["pix2tex", image_path], capture_output=True, text=True)
    return result.stdout



def pdf_to_text(file):
    with open(file.name, 'rb') as f:
        reader = PdfReader(f)
        full_text = ''
        for i, page in enumerate(reader.pages):
            page_text = page.extract_text()
            if page_text is None:
                images = pdf2image.convert_from_path(file.name, first_page=i+1, last_page=i+2)
                for image in images:
                    page_text = image_to_latex(image)
            page_text = clean_text(page_text)
            if len(page_text.split()) > 5: 
                page_number = i + 1
                page_text = "## Metadata: Page Number " + str(page_number) + "\n" + page_text
                full_text += page_text + "\n\n"
    base_name = os.path.splitext(os.path.basename(file.name))[0]
    output_file_name = base_name + ".txt"
    with open(output_file_name, 'w') as f:
        f.write(full_text)
    return output_file_name, page_number



iface = gr.Interface(fn=pdf_to_text,
                     inputs=gr.inputs.File(label="Your PDF"),
                     outputs=gr.outputs.File(label="Download TXT"),
                     title="PDF to TXT",
                     description="Convert your PDF files to clean text")
iface.launch()