app.py

import streamlit as st
import pandas as pd

from src.extractor import extract_skills, extract_experience
from src.scorer import compute_similarity, calculate_score
from src.utils import extract_text, clean_text

# =========================
# PAGE CONFIG
# =========================
st.set_page_config(page_title="AI Resume Screener", layout="wide")

st.markdown(
    "<h1 style='text-align:center;color:#4CAF50;'>🚀 AI Resume Screening Dashboard</h1>",
    unsafe_allow_html=True
)

# =========================
# INPUT SECTION
# =========================
col1, col2 = st.columns(2)

with col1:
    job_desc = st.text_area("📌 Job Description", height=200)

with col2:
    uploaded_files = st.file_uploader("📂 Upload Resumes", accept_multiple_files=True)

# =========================
# MAIN BUTTON
# =========================
if st.button("🔍 Analyze Candidates"):

    if not job_desc:
        st.warning("⚠️ Enter Job Description")
        st.stop()

    if not uploaded_files:
        st.warning("⚠️ Upload at least one resume")
        st.stop()

    resumes = []
    names = []
    skills_list = []
    exp_list = []

    # =========================
    # PROCESS FILES
    # =========================
    for file in uploaded_files:
        text = extract_text(file)
        clean = clean_text(text)

        resumes.append(clean)
        names.append(file.name)

        skills_list.append(extract_skills(clean))
        exp_list.append(extract_experience(clean))

    # =========================
    # JOB DESCRIPTION
    # =========================
    jd_clean = clean_text(job_desc)
    jd_skills = extract_skills(jd_clean)

    # =========================
    # SIMILARITY
    # =========================
    similarities = compute_similarity(resumes, jd_clean)

    results = []

    # =========================
    # SCORING
    # =========================
    for i in range(len(names)):
        matched = len(set(skills_list[i]) & set(jd_skills))
        missing = list(set(jd_skills) - set(skills_list[i]))

        skill_score = matched / len(jd_skills) if jd_skills else 0

        final_score = calculate_score(
            similarities[i],
            skill_score,
            exp_list[i]
        )
            # 🔍 DEBUG
        st.write("Resume:", names[i])
        st.write("Extracted Skills:", skills_list[i])
        st.write("Similarity:", similarities[i])
        st.write("Experience:", exp_list[i])
        st.write("Final Score:", final_score)
        st.write("---")

        results.append({
            "Resume": names[i],
            "Score": float(round(final_score, 3)),
            "Matched Skills": ", ".join(set(skills_list[i]) & set(jd_skills)),
            "Missing Skills": ", ".join(missing),
            "Experience": exp_list[i]
        })

    # =========================
    # CREATE DATAFRAME
    # =========================
    df = pd.DataFrame(results)
    df = df.sort_values(by="Score", ascending=False)

    # =========================
    # ENSURE NUMERIC SCORE (FIX 🔥)
    # =========================
    df["Score"] = pd.to_numeric(df["Score"], errors="coerce")

    # =========================
    # TOP CANDIDATE SHORTLIST
    # =========================
    top_n = max(1, int(len(df) * 0.4))

    df["Status"] = "Rejected"
    df.iloc[:top_n, df.columns.get_loc("Status")] = "Shortlisted"

    # =========================
    # KPI METRICS
    # =========================
    st.subheader("📊 Overview")

    total = len(df)
    shortlisted = len(df[df["Status"] == "Shortlisted"])
    rejected = len(df[df["Status"] == "Rejected"])

    c1, c2, c3 = st.columns(3)
    c1.metric("Total Candidates", total)
    c2.metric("Shortlisted", shortlisted)
    c3.metric("Rejected", rejected)

    # =========================
    # TOP CANDIDATE
    # =========================
    st.subheader("🏆 Top Candidate")

    top = df.iloc[0]
    st.success(f"{top['Resume']} | Score: {top['Score']}")

    # =========================
    # STATUS ICONS
    # =========================
    df["Status"] = df["Status"].apply(
        lambda x: "🟢 Shortlisted" if x == "Shortlisted" else "🔴 Rejected"
    )

    # =========================
    # TABLE
    # =========================
    st.subheader("📋 Candidate Ranking")
    st.dataframe(df, use_container_width=True)

    # =========================
    # CHART (FIXED 🔥)
    # =========================
    st.subheader("📈 Score Distribution")

    chart_data = df[["Resume", "Score"]].set_index("Resume")

    if chart_data["Score"].notnull().sum() > 0:
        st.bar_chart(chart_data)
    else:
        st.warning("No valid score data to display")

    # =========================
    # DOWNLOAD
    # =========================
    csv = df.to_csv(index=False).encode("utf-8")

    st.download_button(
        "📥 Download Results",
        data=csv,
        file_name="results.csv",
        mime="text/csv"
    )