import streamlit as st
import pandas as pd
from llm_services.agenthub import recommend_talent_agent
from llm_services.tools import recommend_talent_tool

st.set_page_config(
    page_title="Talent Recommender",
    page_icon="🎯",
    layout="wide"
)

st.markdown("""
<style>
    .profile-card {
        background-color: #f8f9fa;
        border-radius: 10px;
        padding: 20px;
        margin-bottom: 20px;
        box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
    }
    .metrics-container {
        display: flex;
        justify-content: space-between;
        margin-top: 15px;
    }
    .metric-item {
        text-align: center;
        padding: 10px;
        border-radius: 5px;
        background-color: #e9ecef;
    }
    .header-container {
        padding: 1.5rem;
        background: linear-gradient(90deg, #4b6cb7 0%, #182848 100%);
        color: white;
        border-radius: 10px;
        margin-bottom: 2rem;
    }
</style>
""", unsafe_allow_html=True)

st.markdown("""
<div class="header-container">
    <h1 style="text-align: center;">Talent Recommender</h1>
    <p style="text-align: center; font-size: 1.2rem;">Find the perfect influencer match for your brand</p>
</div>
""", unsafe_allow_html=True)

if 'search_history' not in st.session_state:
    st.session_state.search_history = []

st.markdown("### What kind of talent are you looking for?")
brand_request = st.text_area(
    "Describe your needs in natural language",
    placeholder="e.g., We need financial advisors with high engagement to promote our investment app to professionals aged 30-50",
    height=120
)

search_button = st.button("Find Talent", type="primary")

if search_button and brand_request:
    if brand_request not in st.session_state.search_history:
        st.session_state.search_history.append(brand_request)
    
    with st.spinner("Finding the perfect talent matches..."):
        try:
            search_args = recommend_talent_agent(brand_request=brand_request)
            
            with st.expander("Search Parameters", expanded=False):
                st.json(search_args)
            
            profiles = recommend_talent_tool(**search_args)
            
            st.subheader(f"Top 10 K Results")
            
            tab1, tab2 = st.tabs(["Cards View", "Table View"])
            
            with tab1:
                for i, profile in enumerate(profiles):
                    with st.container():
                        st.markdown(f"""
                        <div class="profile-card">
                            <h3>{profile['name']}</h3>
                            <p><strong>Age:</strong> {profile['age']} | <strong>Gender:</strong> {profile['gender']}</p>
                            <p><strong>Verticals:</strong> {', '.join(profile['verticals'])}</p>
                            <p><strong>Bio:</strong> {profile['bio']}</p>
                            <div class="metrics-container">
                                <div class="metric-item">
                                    <p style="margin:0; font-weight:bold;">{profile['follower_count']:,}</p>
                                    <p style="margin:0; font-size:0.8rem;">Followers</p>
                                </div>
                                <div class="metric-item">
                                    <p style="margin:0; font-weight:bold;">{profile['overall_engagement']:.1%}</p>
                                    <p style="margin:0; font-size:0.8rem;">Engagement</p>
                                </div>
                            </div>
                        </div>
                        """, unsafe_allow_html=True)
                        
            
            with tab2:
                table_data = []
                for profile in profiles:
                    table_data.append({
                        "Name": profile['name'],
                        "Age": profile['age'],
                        "Gender": profile['gender'],
                        "Verticals": ", ".join(profile['verticals']),
                        "Followers": profile['follower_count'],
                        "Engagement": f"{profile['overall_engagement']:.1%}"
                    })
                
                df = pd.DataFrame(table_data)
                st.dataframe(
                    df, 
                    use_container_width=True,
                    hide_index=True
                )
                
                
        except Exception as e:
            st.error(f"An error occurred: {str(e)}")
            st.info("Please try refining your request or check your connection.")

else:
    if st.session_state.search_history:
        st.markdown("### Recent Searches")
        for idx, search in enumerate(st.session_state.search_history[-3:]):
            if st.button(f"{search}", key=f"history_{idx}"):
                brand_request = search
                st.experimental_rerun()
    
    st.markdown("""
    ### How to use this tool:
    
    Simply describe what kind of talent you're looking for in natural language. Our AI will analyze your request and find the most suitable matches from our database.
    
    **Example:** "We need financial advisors with high engagement rates to promote our new investment app targeting professionals aged 35-55."
    """)

st.markdown("---")
st.markdown("""
<p style="text-align: center; color: #6c757d; font-size: 0.8rem;">
    Talent Recommender v1.0 | Powered by AI | © 2025
</p>
""", unsafe_allow_html=True)