app.py

import os
import sys
import traceback
import logging
from pathlib import Path

# Setup logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
)
logger = logging.getLogger(__name__)

logger.info("=" * 80)
logger.info("🚀 Job Application Agent - Startup")
logger.info("=" * 80)

# Add src to Python path
current_dir = os.path.dirname(os.path.abspath(__file__))
src_path = os.path.join(current_dir, 'src')
if src_path not in sys.path:
    sys.path.insert(0, src_path)

logger.info(f"📁 Current directory: {current_dir}")
logger.info(f"📁 Source directory: {src_path}")
logger.info(f"✓ Python path updated")

# Import Gradio
try:
    import gradio as gr
    gradio_version = gr.__version__
    logger.info(f"✓ Gradio imported successfully (version: {gradio_version})")
except ImportError as e:
    logger.error(f"✗ Failed to import Gradio: {e}")
    logger.error("Install with: pip install gradio")
    sys.exit(1)

# Import PyTorch
try:
    import torch
    logger.info(f"✓ PyTorch imported successfully (GPU available: {torch.cuda.is_available()})")
except ImportError as e:
    logger.error(f"✗ Failed to import PyTorch: {e}")
    logger.error("Install with: pip install torch")
    sys.exit(1)

# Import application components
logger.info("\n📦 Attempting to import application modules...")

mayini_model = None
vocab = None
customizer = None
classifier = None
scraper = None
agent = None

# Try to import actual modules
try:
    logger.info("  - Importing MAYINI model...")
    from mayini_model import MAYINIModel, MAYINIVocabulary
    logger.info("    ✓ MAYINI model imported")
except ImportError as e:
    logger.warning(f"  ⚠️  Could not import MAYINI: {e}")
    logger.warning("     Creating mock MAYINI classes...")
    
    class MAYINIVocabulary:
        def __init__(self, vocab_size=5000):
            self.vocab_size = vocab_size
        def encode(self, text, max_len=512):
            return torch.zeros(1, max_len, dtype=torch.long)
        def get_embeddings(self):
            return torch.randn(5000, 256)
    
    class MAYINIModel:
        def __init__(self, **kwargs):
            self.config = kwargs
        def eval(self):
            return self
        def get_embeddings(self, input_ids):
            return torch.randn(1, 512, 256)
        def forward(self, input_ids):
            return torch.randn(1, 5000)
        def count_parameters(self):
            return 3500000
    
    logger.info("    ✓ Mock MAYINI classes created")

try:
    logger.info("  - Importing Job Scraper...")
    from scraper import JobScraper
    logger.info("    ✓ Scraper imported")
except ImportError as e:
    logger.warning(f"  ⚠️  Could not import Scraper: {e}")
    logger.warning("     Creating mock Scraper class...")
    
    class JobScraper:
        def __init__(self):
            self.jobs = [
                {
                    'title': 'Senior Python Developer',
                    'company': 'Tech Giants Inc',
                    'location': 'Remote',
                    'description': 'We are looking for a Senior Python Developer',
                    'requirements': ['Python', 'Docker', 'AWS'],
                    'salary_range': '$120k - $160k',
                    'experience_required': 5
                },
                {
                    'title': 'ML Engineer',
                    'company': 'AI Solutions',
                    'location': 'San Francisco',
                    'description': 'Machine Learning Engineer role',
                    'requirements': ['Python', 'PyTorch', 'TensorFlow'],
                    'salary_range': '$150k - $180k',
                    'experience_required': 4
                }
            ]
        def get_all_jobs(self):
            return self.jobs
        def search_jobs(self, keywords=None, location=None, limit=10):
            return self.jobs[:limit]
    
    logger.info("    ✓ Mock Scraper class created")

try:
    logger.info("  - Importing Resume Customizer...")
    from customizer import ResumeCustomizer
    logger.info("    ✓ Customizer imported")
except ImportError as e:
    logger.warning(f"  ⚠️  Could not import Customizer: {e}")
    logger.warning("     Creating mock Customizer class...")
    
    class ResumeCustomizer:
        def __init__(self, model, vocab):
            self.model = model
            self.vocab = vocab
        def customize_for_job(self, job):
            return {
                'summary': f"Experienced professional ready for {job.get('title', 'N/A')} role",
                'skills': ['Python', 'Docker', 'AWS', 'Git', 'REST API'],
                'customized_for': {'match_score': 0.85}
            }
    
    logger.info("    ✓ Mock Customizer class created")

try:
    logger.info("  - Importing Job Classifier...")
    from classifier import JobRelevanceClassifier
    logger.info("    ✓ Classifier imported")
except ImportError as e:
    logger.warning(f"  ⚠️  Could not import Classifier: {e}")
    logger.warning("     Creating mock Classifier class...")
    
    class JobRelevanceClassifier:
        def __init__(self, **kwargs):
            pass
        def classify_job(self, job, skills):
            return 0.75
        def get_match_details(self, job, skills):
            return {
                'relevance_score': 0.75,
                'recommendation': 'Consider applying',
                'matching_skills': ['Python', 'Docker'],
                'missing_skills': ['Kubernetes']
            }
        def rank_jobs(self, jobs, skills):
            return [(j, 0.7 + i*0.05) for i, j in enumerate(jobs)]
    
    logger.info("    ✓ Mock Classifier class created")

try:
    logger.info("  - Importing Application Agent...")
    from agent import JobApplicationAgent
    logger.info("    ✓ Agent imported")
except ImportError as e:
    logger.warning(f"  ⚠️  Could not import Agent: {e}")
    logger.warning("     Creating mock Agent class...")
    
    class JobApplicationAgent:
        def __init__(self, scraper, customizer, classifier):
            self.scraper = scraper
            self.customizer = customizer
            self.classifier = classifier
        def search_and_apply(self, keywords=None, location=None, num_jobs=5):
            jobs = self.scraper.search_jobs(limit=num_jobs)
            return {
                'total_jobs_found': len(jobs),
                'relevant_jobs': len(jobs),
                'pass_rate': 0.85,
                'applications': [
                    {'job': j, 'relevance_score': 0.8, 'match_details': {
                        'matching_skills': ['Python', 'Docker'],
                        'missing_skills': ['Kubernetes']
                    }} for j in jobs
                ]
            }
    
    logger.info("    ✓ Mock Agent class created")

logger.info("\n" + "=" * 80)
logger.info("✅ All components loaded")
logger.info("=" * 80)

# ============================================================================
# INITIALIZATION
# ============================================================================

logger.info("\n🔧 Initializing components...")

try:
    vocab = MAYINIVocabulary(vocab_size=5000)
    logger.info("✓ MAYINI Vocabulary initialized")
    mayini_model = MAYINIModel(vocab_size=5000, hidden_dim=256, num_heads=8, num_layers=4)
    mayini_model.eval()
    logger.info("✓ MAYINI Model initialized")
    scraper = JobScraper()
    logger.info("✓ Job Scraper initialized")
    customizer = ResumeCustomizer(mayini_model, vocab)
    logger.info("✓ Resume Customizer initialized")
    classifier = JobRelevanceClassifier()
    logger.info("✓ Job Classifier initialized")
    agent = JobApplicationAgent(scraper, customizer, classifier)
    logger.info("✓ Application Agent initialized")
except Exception as e:
    logger.error(f"Error during initialization: {e}")
    logger.error(traceback.format_exc())

logger.info("\n" + "=" * 80)
logger.info("✅ INITIALIZATION COMPLETE - Ready to serve!")
logger.info("=" * 80 + "\n")

# ============================================================================
# INTERFACE FUNCTIONS
# ============================================================================

def search_jobs_interface(keywords: str, location: str, num_jobs: int) -> str:
    """Search and rank jobs"""
    try:
        if not keywords or not keywords.strip():
            return "❌ **Error:** Please enter keywords"
        
        results = agent.search_and_apply(
            keywords=keywords.strip(),
            location=location.strip() if location else "Remote",
            num_jobs=int(num_jobs)
        )
        
        output = f"✅ **Search Results**\n\n"
        output += f"- Found: {results.get('total_jobs_found', 0)} jobs\n"
        output += f"- Relevant: {results.get('relevant_jobs', 0)} jobs\n"
        output += f"- Pass Rate: {results.get('pass_rate', 0):.1%}\n\n"
        output += "---\n\n"
        
        for i, app in enumerate(results.get('applications', [])[:5], 1):
            job = app.get('job', {})
            score = app.get('relevance_score', 0)
            
            output += f"**{i}. {job.get('title', 'N/A')}**\n"
            output += f"- Company: {job.get('company', 'N/A')}\n"
            output += f"- Location: {job.get('location', 'N/A')}\n"
            output += f"- 🎯 Relevance: **{score:.0%}**\n"
            output += f"- 💰 Salary: {job.get('salary_range', 'Not specified')}\n"
            output += f"- 📊 Experience: {job.get('experience_required', 'N/A')} years\n\n"
        
        return output
    except Exception as e:
        logger.error(f"Error in search: {e}\n{traceback.format_exc()}")
        return f"❌ **Error:** {str(e)}\n\nPlease try again."

def customize_resume_interface(job_title: str, company: str, requirements: str) -> str:
    """Customize resume for job"""
    try:
        if not job_title or not job_title.strip():
            return "❌ **Error:** Please enter job title"
        
        job = {
            'title': job_title.strip(),
            'company': company.strip(),
            'requirements': [r.strip() for r in requirements.split(',') if r.strip()] if requirements else []
        }
        
        customized = customizer.customize_for_job(job)
        
        output = f"✅ **Customized Resume**\n\n"
        output += f"**Job:** {job_title} @ {company}\n\n"
        output += f"**Summary:**\n{customized.get('summary', 'N/A')}\n\n"
        output += f"**Top Skills:**\n"
        
        for skill in customized.get('skills', [])[:10]:
            output += f"• {skill}\n"
        
        return output
    except Exception as e:
        logger.error(f"Error in customization: {e}\n{traceback.format_exc()}")
        return f"❌ **Error:** {str(e)}\n\nPlease check your inputs."

def classify_job_interface(job_title: str, requirements: str) -> str:
    """Classify job relevance"""
    try:
        if not job_title or not job_title.strip():
            return "❌ **Error:** Please enter job title"
        
        job = {
            'title': job_title.strip(),
            'requirements': [r.strip() for r in requirements.split(',') if r.strip()] if requirements else [],
            'location': 'Remote',
            'company': 'Unknown',
            'description': job_title.strip(),
            'experience_required': 5,
            'salary_range': 'Unknown'
        }
        
        resume_skills = [
            "Python", "Docker", "AWS", "PostgreSQL", "REST API",
            "Microservices", "Git", "Kubernetes", "Machine Learning"
        ]
        
        score = classifier.classify_job(job, resume_skills)
        details = classifier.get_match_details(job, resume_skills)
        
        output = f"✅ **Job Classification**\n\n"
        output += f"**Job:** {job_title}\n\n"
        
        score_percent = score * 100
        if score >= 0.8:
            emoji = "🟢"
            level = "EXCELLENT"
        elif score >= 0.6:
            emoji = "🟡"
            level = "GOOD"
        elif score >= 0.4:
            emoji = "🟠"
            level = "FAIR"
        else:
            emoji = "🔴"
            level = "POOR"
        
        output += f"{emoji} **Score:** {score_percent:.1f}% ({level})\n\n"
        output += f"✓ **Matching:** {', '.join(details.get('matching_skills', []))}\n"
        output += f"✗ **Missing:** {', '.join(details.get('missing_skills', []))}\n"
        
        return output
    except Exception as e:
        logger.error(f"Error in classification: {e}\n{traceback.format_exc()}")
        return f"❌ **Error:** {str(e)}\n\nPlease check your inputs."

# ============================================================================
# GRADIO INTERFACE
# ============================================================================

logger.info("🎨 Building Gradio interface...")

with gr.Blocks(title="Job Application Agent", theme=gr.themes.Soft()) as demo:
    
    gr.Markdown("""
    # 🤖 Job Application Agent
    ### AI-Powered Job Search & Resume Customization
    **Powered by MAYINI Framework - Custom Transformer ML Model**
    """)
    
    with gr.Tab("🔍 Search & Match Jobs"):
        gr.Markdown("Find jobs matching your skills using AI-powered matching.")
        with gr.Row():
            with gr.Column():
                search_keywords = gr.Textbox(
                    label="Keywords", 
                    placeholder="python docker aws", 
                    value="python"
                )
                search_location = gr.Textbox(
                    label="Location", 
                    placeholder="Remote", 
                    value="Remote"
                )
                search_num = gr.Slider(
                    minimum=1, 
                    maximum=20, 
                    value=5, 
                    step=1, 
                    label="Number of Jobs"
                )
                search_btn = gr.Button("🔍 Search", variant="primary")
            with gr.Column():
                search_output = gr.Markdown(value="### Results will appear here...")
        
        search_btn.click(
            fn=search_jobs_interface, 
            inputs=[search_keywords, search_location, search_num], 
            outputs=search_output
        )
    
    with gr.Tab("📄 Customize Resume"):
        gr.Markdown("Tailor your resume for specific job opportunities.")
        with gr.Row():
            with gr.Column():
                customize_title = gr.Textbox(
                    label="Job Title", 
                    placeholder="Senior Python Developer"
                )
                customize_company = gr.Textbox(
                    label="Company", 
                    placeholder="Tech Inc"
                )
                customize_req = gr.Textbox(
                    label="Requirements", 
                    placeholder="Python, Docker, AWS", 
                    lines=2
                )
                customize_btn = gr.Button("✨ Customize", variant="primary")
            with gr.Column():
                customize_output = gr.Markdown(value="### Customized resume will appear here...")
        
        customize_btn.click(
            fn=customize_resume_interface, 
            inputs=[customize_title, customize_company, customize_req], 
            outputs=customize_output
        )
    
    with gr.Tab("🎯 Classify Job"):
        gr.Markdown("Check job relevance to your skills.")
        with gr.Row():
            with gr.Column():
                classify_title = gr.Textbox(
                    label="Job Title", 
                    placeholder="ML Engineer"
                )
                classify_req = gr.Textbox(
                    label="Requirements", 
                    placeholder="Python, PyTorch", 
                    lines=2
                )
                classify_btn = gr.Button("🎯 Classify", variant="primary")
            with gr.Column():
                classify_output = gr.Markdown(value="### Results will appear here...")
        
        classify_btn.click(
            fn=classify_job_interface, 
            inputs=[classify_title, classify_req], 
            outputs=classify_output
        )
    
    with gr.Tab("ℹ️ About"):
        gr.Markdown("""
        ## Job Application Agent
        Powered by **MAYINI Framework** - a custom Transformer-based ML model.
        
        **Features:**
        - 🔍 AI-powered job search
        - 📄 Resume customization
        - 🎯 Job relevance scoring
        
        **MAYINI Specs:**
        - Vocabulary: 5,000 tokens
        - Hidden Dims: 256
        - Heads: 8
        - Layers: 4
        - Parameters: ~3.5M
        
        **Repository:**
        [GitHub](https://github.com/907-bot/Job-Application-Agent)
        """)

logger.info("✓ Gradio interface built successfully\n")

# ============================================================================
# LAUNCH - FIXED FOR GRADIO COMPATIBILITY
# ============================================================================

if __name__ == "__main__":
    logger.info("🚀 LAUNCHING APPLICATION")
    logger.info("Access at: http://0.0.0.0:7860\n")
    
    try:
        # FIXED: Removed invalid 'concurrency_count' parameter
        # Using only valid parameters for Gradio 4.0+
        demo.launch(
            server_name="0.0.0.0",
            server_port=7860,
            show_error=True,
            share=False
        )
    except Exception as e:
        logger.error(f"Launch failed: {e}\n{traceback.format_exc()}")
        sys.exit(1)