import streamlit as st
import tempfile
import os

from inference import predict_from_pdf

# =========================
# PAGE CONFIG
# =========================
st.set_page_config(
    page_title="AI Resume Category Classifier",
    page_icon="📄",
    layout="centered"
)

# =========================
# SIDEBAR
# =========================
st.sidebar.title("📄 Resume Parser – ML App")
st.sidebar.write(
    """
    This app uses a trained **machine learning model** to 
    classify resumes into **job categories** based on their content.
    """
)
st.sidebar.markdown("---")
st.sidebar.caption("Built as an internship project and refined for college submission.")

# =========================
# MAIN CONTENT
# =========================
st.title("📄 AI Resume Category Classifier")
st.write(
    """
    Upload a **PDF resume** and the model will predict its **category/domain**.
    """
)

uploaded_file = st.file_uploader(
    "Upload a resume (PDF file)",
    type=["pdf"],
    accept_multiple_files=False
)

if uploaded_file is not None:
    st.info(f"Uploaded file: `{uploaded_file.name}`")

    if st.button("🔍 Analyze Resume"):
        with st.spinner("Reading and classifying the resume..."):
            # Save uploaded file to a temporary path
            with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp:
                tmp.write(uploaded_file.read())
                temp_pdf_path = tmp.name

            try:
                predicted_label, proba_dict, raw_text = predict_from_pdf(temp_pdf_path)

                # Show predicted category
                st.success(f"Predicted Category: **{predicted_label}**")

                # Show probabilities (if available)
                if proba_dict is not None:
                    st.subheader("Prediction Confidence")
                    sorted_items = sorted(proba_dict.items(), key=lambda x: x[1], reverse=True)
                    for label, prob in sorted_items:
                        st.write(f"- {label}: `{prob:.2%}`")

                # Show extracted text preview
                with st.expander("🔎 View extracted resume text (preview)"):
                    preview = (raw_text or "").strip()
                    if preview:
                        st.text(preview[:2000])  # first 2000 characters
                    else:
                        st.write("No text could be extracted from the PDF.")

            except Exception as e:
                st.error(f"❌ Error while processing the file: {e}")

            finally:
                # Clean up temp file
                if os.path.exists(temp_pdf_path):
                    os.remove(temp_pdf_path)
else:
    st.warning("Please upload a PDF resume to start.")