app.py

import gradio as gr
from extractor import extract_text
from cleaner import clean_text
from formatter import format_text

def process_taxdoc(file):
    raw_text = extract_text(file)
    cleaned = clean_text(raw_text)
    formatted = format_text(cleaned)

    output_path = "processed_" + file.name.split("/")[-1].replace(".pdf", ".txt").replace(".docx", ".txt")
    with open(output_path, "w", encoding="utf-8") as f:
        f.write(formatted)

    return formatted, output_path

iface = gr.Interface(
    fn=process_taxdoc,
    inputs=gr.File(file_types=[".pdf", ".txt"]),
    outputs=[gr.Textbox(label="Formatted Text", lines=20), gr.File(label="Download File")],
    title="📄 TaxDoc Preprocessor for RAG",
    description="Upload Indian tax PDFs or TXT files to clean and structure them for better RAG retrieval.",
)

if __name__ == "__main__":
    iface.launch()