app.py

import streamlit as st
import tempfile
import os
import json

from rag_pipeline import run_pipeline, load_document, split_documents
from tools import summarize_document_tool, generate_mcqs_tool, topic_explanation_tool

# ── Page Config ─────────────────────────────────────────────────────────────
st.set_page_config(
    page_title="Smart Academic Assistant",
    page_icon="🎓",
    layout="wide"
)

st.title("🎓 Smart Academic Assistant")
st.markdown("Upload your study material and ask questions — powered by RAG + Groq LLM")

# ── Sidebar ──────────────────────────────────────────────────────────────────
with st.sidebar:
    st.header("⚙️ Settings")
    mode = st.radio(
        "Choose Mode",
        ["📖 Q&A (RAG)", "🤖 Agentic Tools"]
    )
    st.markdown("---")
    st.markdown("**Models used:**")
    st.markdown("- LLM: `llama-3.1-8b-instant` via Groq")
    st.markdown("- Embeddings: `all-MiniLM-L6-v2`")
    st.markdown("- Vector Store: `FAISS`")

# ── File Upload ───────────────────────────────────────────────────────────────
st.header("📂 Upload Document")
uploaded_file = st.file_uploader(
    "Upload a PDF, DOCX, or TXT file",
    type=["pdf", "docx", "txt"]
)

tmp_path = None
doc_text = ""

if uploaded_file:
    suffix = os.path.splitext(uploaded_file.name)[-1]
    with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp:
        tmp.write(uploaded_file.read())
        tmp_path = tmp.name

    st.success(f"✅ File uploaded: **{uploaded_file.name}**")

    try:
        docs = load_document(tmp_path)
        doc_text = "\n\n".join([d.page_content for d in docs])
    except Exception as e:
        st.error(f"Error reading file: {e}")

if mode == "📖 Q&A (RAG)":
    st.header("❓ Ask a Question")

    question = st.text_input(
        "Enter your question about the document:",
        placeholder="e.g. What is the main topic of this document?"
    )

    if st.button("🔍 Get Answer", type="primary"):
        if not uploaded_file:
            st.warning("Please upload a document first.")
        elif not question.strip():
            st.warning("Please enter a question.")
        else:
            with st.spinner("Thinking... 🤔"):
                try:
                    result = run_pipeline(tmp_path, question)

                    st.header("📊 Structured Response")

                    col1, col2 = st.columns(2)
                    with col1:
                        st.metric("Confidence Score", result.get("confidence_score", "N/A"))
                    with col2:
                        st.metric("Source", result.get("source_document", "N/A")[:40])

                    st.subheader("Question")
                    st.info(result.get("question", question))

                    st.subheader("Answer")
                    st.success(result.get("answer", "No answer found."))

                    st.subheader("Raw JSON Output")
                    st.json(result)

                except Exception as e:
                    st.error(f"Pipeline error: {e}")

elif mode == "🤖 Agentic Tools":
    st.header("🤖 Agentic Tools")

    if not uploaded_file:
        st.warning("Upload a document to use agentic tools.")
    else:
        tool_choice = st.selectbox(
            "Select a tool:",
            [
                "📝 Summarize Document",
                "❓ Generate MCQs",
                "💡 Explain a Topic"
            ]
        )

        if tool_choice == "📝 Summarize Document":
            if st.button("Summarize", type="primary"):
                with st.spinner("Summarizing..."):
                    result = summarize_document_tool.invoke(doc_text)
                    st.subheader("📝 Summary")
                    st.write(result)

        elif tool_choice == "❓ Generate MCQs":
            if st.button("Generate MCQs", type="primary"):
                with st.spinner("Generating questions..."):
                    result = generate_mcqs_tool.invoke(doc_text)
                    st.subheader("❓ MCQs")
                    st.write(result)

        elif tool_choice == "💡 Explain a Topic":
            topic = st.text_input(
                "Enter topic to explain:",
                placeholder="e.g. Binary Trees, Gradient Descent, French Revolution"
            )
            if st.button("Explain", type="primary"):
                if not topic.strip():
                    st.warning("Enter a topic.")
                else:
                    with st.spinner(f"Explaining '{topic}'..."):
                        result = topic_explanation_tool.invoke(topic)
                        st.subheader(f"💡 {topic}")
                        st.write(result)

st.markdown("---")
st.markdown(
    "Built with LangChain · Groq · FAISS · Streamlit | "
    "Smart Academic Assistant v1.0"
)