app.py

import os
import json
import requests
import streamlit as st
from datetime import datetime
from docx import Document
from PyPDF2 import PdfReader

# ------------------------------------------------------------------------------
# 🚀 CONFIGURATION - API & SETTINGS
# ------------------------------------------------------------------------------
GROQ_API_KEY = os.getenv("GROQ_API")
GROQ_ENDPOINT = "https://api.groq.com/openai/v1/chat/completions"
GROQ_MODEL = "llama-3.3-70b-versatile"

# ------------------------------------------------------------------------------
# 🔥 ADVANCED API CALL FUNCTION
# ------------------------------------------------------------------------------
def call_groq_api(messages, temperature=0.7):
    """Handles API calls to Groq for chat-based completions with robust error handling."""
    headers = {
        "Authorization": f"Bearer {GROQ_API_KEY}",
        "Content-Type": "application/json"
    }
    payload = {
        "model": GROQ_MODEL,
        "messages": messages,
        "temperature": temperature,
        "max_tokens": 1024
    }
    try:
        response = requests.post(GROQ_ENDPOINT, headers=headers, json=payload, timeout=15)
        response.raise_for_status()
        return response.json()["choices"][0]["message"]["content"].strip()
    except requests.exceptions.RequestException as e:
        st.error(f"❌ API Request Failed: {e}")
        return None

# ------------------------------------------------------------------------------
# 📄 FILE PROCESSING (PDF / DOCX)
# ------------------------------------------------------------------------------
def extract_resume_text(file_obj):
    """Extracts text from PDF or DOCX resumes."""
    file_bytes = file_obj.read()
    file_obj.seek(0)
    ext = os.path.splitext(file_obj.name)[-1].lower()

    if ext == ".pdf":
        return "\n".join(page.extract_text() for page in PdfReader(file_bytes).pages if page.extract_text())
    elif ext in [".docx", ".doc"]:
        return "\n".join([para.text for para in Document(file_bytes).paragraphs])
    else:
        return file_bytes.decode("utf-8", errors="ignore")

# ------------------------------------------------------------------------------
# 🎯 AI-Powered Resume Parsing
# ------------------------------------------------------------------------------
def parse_resume(resume_text):
    """Extracts structured JSON resume details using Groq's Llama 3.3 model."""
    messages = [
        {"role": "system", "content": "You are a resume parsing expert. Extract structured data."},
        {"role": "user", "content": f"Extract structured data from this resume:\n{resume_text}\n\nOutput as JSON."}
    ]
    return call_groq_api(messages, temperature=0.4)

# ------------------------------------------------------------------------------
# ✍️ AI-Powered Cover Letter Generator
# ------------------------------------------------------------------------------
def generate_cover_letter(candidate_json, job_description):
    """Generates a professional cover letter using structured candidate data."""
    date_str = datetime.today().strftime("%d - %b - %Y")
    messages = [
        {"role": "system", "content": "You are a top-tier career advisor. Write persuasive cover letters."},
        {"role": "user", "content": f"""
        Generate a professional cover letter using:
        - Candidate Profile: {candidate_json}
        - Job Description: {job_description}
        - Date: {date_str}
        
        The cover letter should be engaging, personalized, and formatted professionally.
        """}
    ]
    return call_groq_api(messages, temperature=0.5)

# ------------------------------------------------------------------------------
# 📜 AI-Powered Resume Creator
# ------------------------------------------------------------------------------
def generate_resume(first_name, last_name, location, work_experience, school_experience, skills):
    """Generates a well-formatted professional resume."""
    candidate_data = json.dumps({
        "first_name": first_name,
        "last_name": last_name,
        "location": location,
        "work_experience": work_experience,
        "school_experience": school_experience,
        "skills": skills
    }, indent=2)

    messages = [
        {"role": "system", "content": "You are a professional resume writer."},
        {"role": "user", "content": f"""
        Create a structured, ATS-friendly resume using:
        {candidate_data}

        Ensure:
        - Clear sections (Personal Info, Experience, Education, Skills)
        - Professional formatting
        - A compelling summary if possible.
        """}
    ]
    return call_groq_api(messages, temperature=0.5)

# ------------------------------------------------------------------------------
# 🎨 STREAMLIT UI DESIGN
# ------------------------------------------------------------------------------
st.set_page_config(page_title="AI-Powered Job Assistant", layout="wide")
st.title("🚀 AI-Powered Resume & Cover Letter Generator")
st.markdown("#### Generate a **Resume** or a **Cover Letter** using **Llama 3.3-70B** powered by **Groq AI**.")

tabs = st.tabs(["📄 Cover Letter Generator", "📑 Resume Creator"])

# ----- COVER LETTER GENERATOR -----
with tabs[0]:
    st.header("📄 Cover Letter Generator")
    resume_file = st.file_uploader("Upload Your Resume", type=["pdf", "docx", "txt"])
    job_description = st.text_area("Paste Job Description", height=200)

    if st.button("🔹 Generate Cover Letter"):
        if resume_file and job_description.strip():
            with st.spinner("✨ Processing resume..."):
                resume_text = extract_resume_text(resume_file)
                candidate_json = parse_resume(resume_text)
                cover_letter = generate_cover_letter(candidate_json, job_description)
            st.success("✅ Cover Letter Generated!")
            st.text_area("📜 Your Cover Letter:", cover_letter, height=300)
        else:
            st.warning("⚠️ Please upload a resume and paste the job description.")

# ----- RESUME CREATOR -----
with tabs[1]:
    st.header("📑 Resume Creator")

    with st.form("resume_form"):
        col1, col2 = st.columns(2)
        first_name = col1.text_input("First Name")
        last_name = col2.text_input("Last Name")
        location = st.text_input("Location")
        work_experience = st.text_area("Work Experience", height=150)
        school_experience = st.text_area("Education", height=150)
        skills = st.text_area("Skills", height=100)
        submit = st.form_submit_button("📑 Generate Resume")

    if submit:
        with st.spinner("📝 Creating Resume..."):
            resume_text = generate_resume(first_name, last_name, location, work_experience, school_experience, skills)
        st.success("✅ Resume Generated!")
        st.text_area("📜 Your Resume:", resume_text, height=400)