import streamlit as st
from extractor import ocr_document, extract_fields

st.set_page_config(page_title="InvoiceExtract", layout="centered")
st.title("🧾 InvoiceExtract — OCR + Extraction")

file = st.file_uploader("Upload une facture (PDF ou image)", type=["pdf", "png", "jpg", "jpeg"])

if file:
    file_bytes = file.read()
    is_pdf = file.type == "application/pdf"

    with st.spinner("Analyse OCR en cours..."):
        text, images_to_show = ocr_document(file_bytes, is_pdf)

    if images_to_show:
        st.image(images_to_show[0], caption="Aperçu du document", use_container_width=True)

    st.subheader("Données Extraites (JSON)")
    data = extract_fields(text)
    st.json(data) # [cite: 15, 26]

    with st.expander("Voir le texte brut (OCR)"):
        st.text(text) # [cite: 17]