nordabiz/scripts/generate_audit_report.py

#!/usr/bin/env python3
"""
Unified Audit Report Generator for NordaBiz
=============================================

Generates comprehensive audit reports combining:
- Social Media audit data
- Google Business Profile audit data
- SEO audit data
- Competitor monitoring data (if available)

Usage:
    python generate_audit_report.py --company-id 26
    python generate_audit_report.py --all
    python generate_audit_report.py --company-id 26 --type social

Author: NordaBiz Development Team
Created: 2026-02-06
"""

import os
import sys
import json
import argparse
import logging
from datetime import datetime, date, timedelta
from typing import Optional, Dict, List, Any
from pathlib import Path

# Load .env file
try:
    from dotenv import load_dotenv
    script_dir = Path(__file__).resolve().parent
    project_root = script_dir.parent
    env_path = project_root / '.env'
    if env_path.exists():
        load_dotenv(env_path)
except ImportError:
    pass

sys.path.insert(0, str(Path(__file__).resolve().parent.parent))

from sqlalchemy import create_engine, text
from sqlalchemy.orm import sessionmaker
from database import (
    Company, CompanySocialMedia, GBPAudit, CompanyWebsiteAnalysis,
    CompanyCompetitor, CompetitorSnapshot, AuditReport, SessionLocal
)

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

DATABASE_URL = os.getenv(
    'DATABASE_URL',
    'postgresql://nordabiz_app:CHANGE_ME@127.0.0.1:5432/nordabiz'
)


class AuditReportGenerator:
    """Generates unified audit reports for companies."""

    def __init__(self, database_url: str = DATABASE_URL):
        self.engine = create_engine(database_url)
        self.Session = sessionmaker(bind=self.engine)

    def generate_report(self, company_id: int, report_type: str = 'full') -> Dict[str, Any]:
        """Generate a comprehensive audit report for a company."""
        with self.Session() as session:
            company = session.query(Company).filter(Company.id == company_id).first()
            if not company:
                raise ValueError(f"Company {company_id} not found")

            logger.info(f"Generating {report_type} report for: {company.name} (ID: {company_id})")

            report_data = {
                'company': {
                    'id': company.id,
                    'name': company.name,
                    'slug': company.slug,
                    'website': company.website,
                    'city': company.address_city,
                    'category': company.category.name if company.category else None,
                },
                'generated_at': datetime.now().isoformat(),
                'report_type': report_type,
                'sections': {},
                'scores': {},
            }

            sections_included = {'social': False, 'gbp': False, 'seo': False, 'competitors': False}

            # Social Media section
            if report_type in ('full', 'social'):
                social_data = self._get_social_data(session, company_id)
                if social_data:
                    report_data['sections']['social'] = social_data
                    report_data['scores']['social'] = social_data.get('average_completeness', 0)
                    sections_included['social'] = True

            # GBP section
            if report_type in ('full', 'gbp'):
                gbp_data = self._get_gbp_data(session, company_id)
                if gbp_data:
                    report_data['sections']['gbp'] = gbp_data
                    report_data['scores']['gbp'] = gbp_data.get('completeness_score', 0)
                    sections_included['gbp'] = True

            # SEO section
            if report_type in ('full', 'seo'):
                seo_data = self._get_seo_data(session, company_id)
                if seo_data:
                    report_data['sections']['seo'] = seo_data
                    report_data['scores']['seo'] = seo_data.get('overall_score', 0)
                    sections_included['seo'] = True

            # Competitors section
            if report_type == 'full':
                competitor_data = self._get_competitor_data(session, company_id)
                if competitor_data:
                    report_data['sections']['competitors'] = competitor_data
                    sections_included['competitors'] = True

            # Calculate overall score
            scores = [v for v in report_data['scores'].values() if v and v > 0]
            overall = int(sum(scores) / len(scores)) if scores else 0
            report_data['scores']['overall'] = overall

            # Save report
            report = AuditReport(
                company_id=company_id,
                report_type=report_type,
                period_start=date.today() - timedelta(days=30),
                period_end=date.today(),
                overall_score=overall,
                social_score=report_data['scores'].get('social'),
                gbp_score=report_data['scores'].get('gbp'),
                seo_score=report_data['scores'].get('seo'),
                sections=sections_included,
                data=report_data,
                generated_by='system',
                status='draft',
            )
            session.add(report)
            session.commit()
            session.refresh(report)

            report_data['report_id'] = report.id
            logger.info(f"Report #{report.id} generated. Overall score: {overall}/100")

            return report_data

    def _get_social_data(self, session, company_id: int) -> Optional[Dict]:
        """Get social media audit data."""
        profiles = session.query(CompanySocialMedia).filter(
            CompanySocialMedia.company_id == company_id,
            CompanySocialMedia.is_valid == True
        ).all()

        if not profiles:
            return None

        platforms = []
        total_completeness = 0

        for p in profiles:
            platform_data = {
                'platform': p.platform,
                'url': p.url,
                'page_name': p.page_name,
                'followers_count': p.followers_count,
                'has_bio': p.has_bio,
                'has_profile_photo': p.has_profile_photo,
                'completeness_score': p.profile_completeness_score or 0,
                'last_checked': p.last_checked_at.isoformat() if p.last_checked_at else None,
            }
            platforms.append(platform_data)
            total_completeness += (p.profile_completeness_score or 0)

        average = int(total_completeness / len(platforms)) if platforms else 0

        return {
            'platforms_found': len(platforms),
            'platforms': platforms,
            'average_completeness': average,
            'missing_platforms': self._find_missing_platforms(profiles),
        }

    @staticmethod
    def _find_missing_platforms(profiles) -> List[str]:
        """Find platforms without profiles."""
        all_platforms = {'facebook', 'instagram', 'linkedin', 'youtube', 'twitter', 'tiktok'}
        found = {p.platform for p in profiles}
        return sorted(all_platforms - found)

    def _get_gbp_data(self, session, company_id: int) -> Optional[Dict]:
        """Get GBP audit data."""
        audit = session.query(GBPAudit).filter(
            GBPAudit.company_id == company_id
        ).order_by(GBPAudit.audit_date.desc()).first()

        if not audit:
            return None

        return {
            'completeness_score': audit.completeness_score,
            'score_category': audit.score_category,
            'audit_date': audit.audit_date.isoformat() if audit.audit_date else None,
            'fields_status': audit.fields_status,
            'review_count': audit.review_count,
            'average_rating': float(audit.average_rating) if audit.average_rating else None,
            'photo_count': audit.photo_count,
            'nap_consistent': audit.nap_consistent,
            'review_response_rate': float(audit.review_response_rate) if audit.review_response_rate else None,
            'review_sentiment': audit.review_sentiment,
            'recommendations': audit.recommendations or [],
        }

    def _get_seo_data(self, session, company_id: int) -> Optional[Dict]:
        """Get SEO audit data."""
        analysis = session.query(CompanyWebsiteAnalysis).filter(
            CompanyWebsiteAnalysis.company_id == company_id
        ).first()

        if not analysis:
            return None

        return {
            'overall_score': analysis.seo_overall_score,
            'pagespeed_seo': analysis.pagespeed_seo_score,
            'pagespeed_performance': analysis.pagespeed_performance_score,
            'pagespeed_accessibility': analysis.pagespeed_accessibility_score,
            'meta_title': analysis.meta_title,
            'has_structured_data': analysis.has_structured_data,
            'has_sitemap': analysis.has_sitemap,
            'has_robots_txt': analysis.has_robots_txt,
            'is_mobile_friendly': analysis.is_mobile_friendly,
            'local_seo_score': analysis.local_seo_score,
            'has_local_business_schema': analysis.has_local_business_schema,
            'citations_count': analysis.citations_count,
            'content_freshness_score': analysis.content_freshness_score,
            'core_web_vitals': {
                'lcp_ms': analysis.largest_contentful_paint_ms,
                'fid_ms': analysis.first_input_delay_ms,
                'cls': float(analysis.cumulative_layout_shift) if analysis.cumulative_layout_shift else None,
            },
            'seo_issues': analysis.seo_issues,
        }

    def _get_competitor_data(self, session, company_id: int) -> Optional[Dict]:
        """Get competitor monitoring data."""
        competitors = session.query(CompanyCompetitor).filter(
            CompanyCompetitor.company_id == company_id,
            CompanyCompetitor.is_active == True
        ).all()

        if not competitors:
            return None

        competitor_list = []
        for comp in competitors:
            # Get latest snapshot
            latest = session.query(CompetitorSnapshot).filter(
                CompetitorSnapshot.competitor_id == comp.id
            ).order_by(CompetitorSnapshot.snapshot_date.desc()).first()

            competitor_list.append({
                'name': comp.competitor_name,
                'rating': float(comp.competitor_rating) if comp.competitor_rating else None,
                'review_count': comp.competitor_review_count,
                'category': comp.competitor_category,
                'latest_changes': latest.changes if latest else None,
            })

        return {
            'total_tracked': len(competitors),
            'competitors': competitor_list,
        }


def main():
    parser = argparse.ArgumentParser(description='Generate Unified Audit Report')
    parser.add_argument('--company-id', type=int, help='Generate report for specific company')
    parser.add_argument('--all', action='store_true', help='Generate for all active companies')
    parser.add_argument('--type', choices=['full', 'social', 'gbp', 'seo'], default='full')
    parser.add_argument('--json', action='store_true', help='Output JSON to stdout')
    parser.add_argument('--verbose', '-v', action='store_true')
    args = parser.parse_args()

    if args.verbose:
        logging.getLogger().setLevel(logging.DEBUG)

    generator = AuditReportGenerator()

    if args.company_id:
        report = generator.generate_report(args.company_id, args.type)
        if args.json:
            print(json.dumps(report, default=str, indent=2, ensure_ascii=False))
        else:
            print(f"\nReport #{report.get('report_id')} generated")
            print(f"Overall score: {report['scores'].get('overall', 0)}/100")
            for section, data in report.get('sections', {}).items():
                print(f"  {section}: included")

    elif args.all:
        engine = create_engine(DATABASE_URL)
        Session = sessionmaker(bind=engine)
        with Session() as session:
            companies = session.query(Company).filter(Company.status == 'active').all()
            company_ids = [c.id for c in companies]

        for cid in company_ids:
            try:
                report = generator.generate_report(cid, args.type)
                print(f"Company {cid}: score={report['scores'].get('overall', 0)}")
            except Exception as e:
                logger.error(f"Company {cid} failed: {e}")
    else:
        parser.print_help()
        sys.exit(1)


if __name__ == '__main__':
    main()