app/api/analytics.py

from fastapi import APIRouter, Depends, HTTPException, status, Request, Query, Body
from sqlalchemy.orm import Session
from sqlalchemy import func, desc
from typing import Optional, List, Dict, Any
from datetime import datetime, timedelta, timezone
import json

from app.db.base import get_db
from app.models.analytics import VisitorLog, AnalyticsEvent
from app.models.user import User, UserRole
from app.api.auth import get_current_user, get_optional_current_user
from app.services.metadata_service import parse_user_agent, get_geoip_info

router = APIRouter(prefix="/analytics", tags=["Analytics"])

def check_admin(user_id: int, db: Session):
    user = db.query(User).filter(User.id == user_id).first()
    if not user or user.role != UserRole.ADMIN:
        raise HTTPException(status_code=403, detail="Not authorized")
    return user

@router.post("/track")
async def track_event(
    request: Request,
    payload: Dict[str, Any] = Body(...),
    db: Session = Depends(get_db),
    current_user_id: Optional[int] = Depends(get_optional_current_user)
):
    """
    Track a visitor event. Creates or updates a visitor log and logs the event.
    Payload: { session_id, event_type, page_url, page_title, event_data, client_metadata }
    """
    # 0. Skip recording for Admin users to keep analytics clean
    if current_user_id:
        user = db.query(User).filter(User.id == current_user_id).first()
        if user and user.role == UserRole.ADMIN:
            return {"isSuccess": True, "message": "Admin event skipped"}

    print(f">>> [ANALYTICS] Received tracking event: {payload.get('event_type')} for {payload.get('page_url')}")

    session_id = payload.get("session_id")
    event_type = payload.get("event_type", "PAGE_VIEW")
    
    # 1. Identify or Create Visitor
    visitor = None
    if session_id:
        visitor = db.query(VisitorLog).filter(VisitorLog.session_id == session_id).first()
    
    # Fallback to IP if no session_id (e.g. tracking disabled or first load)
    ip_address = request.client.host if request.client else None
    user_agent = request.headers.get("User-Agent")
    
    if not visitor:
        ua_info = parse_user_agent(user_agent)
        location_data_raw = await get_geoip_info(ip_address)
        location_data = json.loads(location_data_raw) if location_data_raw else None
        
        visitor = VisitorLog(
            session_id=session_id,
            user_id=current_user_id,
            ip_address=ip_address,
            user_agent=user_agent,
            browser=ua_info["browser"],
            os=ua_info["os"],
            device_type=ua_info["device_type"],
            location_data=location_data,
            client_metadata=payload.get("client_metadata")
        )
        db.add(visitor)
        db.flush() # Get visitor.id
    else:
        # Update last seen and potentially user_id if they just logged in
        visitor.last_seen = datetime.now(timezone.utc)
        if current_user_id and not visitor.user_id:
            visitor.user_id = current_user_id
        
        # Update client metadata if provided
        if payload.get("client_metadata"):
            visitor.client_metadata = payload.get("client_metadata")

    # 2. Log Event
    event = AnalyticsEvent(
        visitor_id=visitor.id,
        event_type=event_type,
        page_url=payload.get("page_url"),
        page_title=payload.get("page_title"),
        event_data=payload.get("event_data")
    )
    db.add(event)
    
    try:
        db.commit()
    except Exception as e:
        db.rollback()
        # Log error or silence it to not break frontend
        print(f">>> [ANALYTICS] Insert error: {e}")
        return {"isSuccess": False}

    return {"isSuccess": True}

@router.get("/admin/stats")
async def get_analytics_stats(
    days: int = Query(7, ge=1, le=90),
    start_date: Optional[str] = Query(None),
    end_date: Optional[str] = Query(None),
    category_id: Optional[int] = Query(None),
    current_user_id: int = Depends(get_current_user),
    db: Session = Depends(get_db)
):
    check_admin(current_user_id, db)
    
    # 0. Date Range Logic
    now = datetime.now(timezone.utc)
    if start_date and end_date:
        try:
            since = datetime.fromisoformat(start_date.replace('Z', '+00:00'))
            until = datetime.fromisoformat(end_date.replace('Z', '+00:00'))
            # Calculate effective days for trends
            days = (until - since).days + 1
            if days < 1: days = 1
            if days > 90: days = 90
        except:
            since = now - timedelta(days=days)
            until = now
    else:
        since = now - timedelta(days=days)
        until = now
    
    # 1. Total unique visitors (in period)
    total_visitors = db.query(func.count(VisitorLog.id)).filter(VisitorLog.last_seen >= since, VisitorLog.last_seen <= until).scalar() or 0
    
    # 2. Total events (in period)
    total_events = db.query(func.count(AnalyticsEvent.id)).filter(AnalyticsEvent.created_at >= since, AnalyticsEvent.created_at <= until).scalar() or 0
    
    # 3. Active Now (last 5 minutes)
    active_since = now - timedelta(minutes=5)
    active_now = db.query(func.count(VisitorLog.id)).filter(VisitorLog.last_seen >= active_since).scalar() or 0
    
    # 4. Device breakdown
    devices = db.query(
        VisitorLog.device_type, 
        func.count(VisitorLog.id)
    ).filter(VisitorLog.last_seen >= since, VisitorLog.last_seen <= until).group_by(VisitorLog.device_type).all()
    
    raw_devices = {d[0]: d[1] for d in devices}
    device_stats = {
        "desktop": raw_devices.get("Desktop", 0),
        "mobile": raw_devices.get("Mobile", 0),
        "tablet": raw_devices.get("Tablet", 0)
    }
    
    # 5. Daily Trends
    trends = []
    # We'll calculate for each day in the range
    for i in range(days - 1, -1, -1):
        day_start = (now - timedelta(days=i)).replace(hour=0, minute=0, second=0, microsecond=0)
        day_end = day_start + timedelta(days=1)
        
        day_visitors = db.query(func.count(VisitorLog.id)).filter(
            VisitorLog.last_seen >= day_start,
            VisitorLog.last_seen < day_end
        ).scalar() or 0
        
        day_events = db.query(func.count(AnalyticsEvent.id)).filter(
            AnalyticsEvent.created_at >= day_start,
            AnalyticsEvent.created_at < day_end
        ).scalar() or 0
        
        trends.append({
            "date": day_start.strftime("%Y-%m-%d"),
            "visitors": day_visitors,
            "events": day_events
        })
    
    # 6. Top pages
    top_pages = db.query(
        AnalyticsEvent.page_url, 
        func.count(AnalyticsEvent.id)
    ).filter(
        AnalyticsEvent.created_at >= since,
        AnalyticsEvent.created_at <= until,
        AnalyticsEvent.event_type == "PAGE_VIEW"
    ).group_by(AnalyticsEvent.page_url).order_by(desc(func.count(AnalyticsEvent.id))).limit(10).all()
    
    pages_data = [{"url": p[0], "views": p[1]} for p in top_pages]

    # 7. Top Viewed Products
    # Logic: Join with Product to allow category filtering
    top_prod_query = db.query(
        AnalyticsEvent.page_url,
        AnalyticsEvent.page_title,
        func.count(AnalyticsEvent.id)
    ).filter(
        AnalyticsEvent.created_at >= since,
        AnalyticsEvent.created_at <= until,
        AnalyticsEvent.page_url.like("%/products/%")
    )

    if category_id:
        # Get category and subcategories
        subcats = db.query(Category.id).filter(or_(Category.id == category_id, Category.parent_id == category_id)).all()
        cat_ids = [c[0] for c in subcats]
        
        # We need to extract product ID from URL to join with Product table
        # URL pattern is usually .../products/{id} or .../products/{slug}
        # For simplicity, we join using LIKE since SQLite/others might not have robust regex support
        top_prod_query = top_prod_query.join(
            Product, 
            or_(
                AnalyticsEvent.page_url.like("%/products/" + cast(Product.id, String)),
                AnalyticsEvent.page_url.like("%/products/" + Product.slug)
            )
        ).filter(Product.category_id.in_(cat_ids))

    top_products = top_prod_query.group_by(
        AnalyticsEvent.page_url, AnalyticsEvent.page_title
    ).order_by(desc(func.count(AnalyticsEvent.id))).limit(10).all()

    products_data = [{"url": p[0], "name": p[1], "views": p[2]} for p in top_products]

    # 8. Top Searches
    top_search_query = db.query(AnalyticsEvent.event_data).filter(
        AnalyticsEvent.created_at >= since,
        AnalyticsEvent.created_at <= until,
        AnalyticsEvent.event_type == "SEARCH"
    )
    
    events_with_data = top_search_query.all()

    search_counts = {}
    for (event_data,) in events_with_data:
        if isinstance(event_data, dict) and "query" in event_data:
            q = event_data["query"].strip().lower()
            if q:
                search_counts[q] = search_counts.get(q, 0) + 1
        elif isinstance(event_data, str): # Handle string-encoded JSON if necessary
            try:
                data = json.loads(event_data)
                q = data.get("query", "").strip().lower()
                if q:
                    search_counts[q] = search_counts.get(q, 0) + 1
            except:
                pass

    searches_data = sorted(
        [{"query": q, "count": c} for q, c in search_counts.items()],
        key=lambda x: x["count"],
        reverse=True
    )[:10]
    
    return {
        "isSuccess": True,
        "value": {
            "total_visitors": total_visitors,
            "total_events": total_events,
            "active_now": active_now,
            "device_stats": device_stats,
            "trends": trends,
            "top_pages": pages_data,
            "top_products": products_data,
            "top_searches": searches_data
        }
    }

@router.get("/admin/visitors")
async def get_recent_visitors(
    page: int = Query(1, ge=1),
    page_size: int = Query(20, ge=1, le=100),
    current_user_id: int = Depends(get_current_user),
    db: Session = Depends(get_db)
):
    check_admin(current_user_id, db)
    
    query = db.query(VisitorLog).order_by(desc(VisitorLog.last_seen))
    
    total = query.count()
    visitors = query.offset((page - 1) * page_size).limit(page_size).all()
    
    results = []
    for v in visitors:
        # Get event count
        event_count = db.query(func.count(AnalyticsEvent.id)).filter(AnalyticsEvent.visitor_id == v.id).scalar() or 0
        
        results.append({
            "id": v.id,
            "ip": v.ip_address,
            "browser": v.browser,
            "os": v.os,
            "device": v.device_type,
            "location": v.location_data,
            "last_seen": v.last_seen.isoformat(),
            "event_count": event_count,
            "user": {"name": v.user.name, "email": v.user.email} if v.user else None
        })
    
    return {
        "isSuccess": True,
        "value": {
            "visitors": results,
            "total": total,
            "page": page,
            "page_size": page_size
        }
    }

@router.get("/admin/events")
async def get_analytics_events(
    page: int = Query(1, ge=1),
    page_size: int = Query(30, ge=1, le=100),
    event_type: Optional[str] = Query(None),
    visitor_id: Optional[int] = Query(None),
    start_date: Optional[str] = Query(None),
    end_date: Optional[str] = Query(None),
    current_user_id: int = Depends(get_current_user),
    db: Session = Depends(get_db)
):
    check_admin(current_user_id, db)
    
    query = db.query(AnalyticsEvent).order_by(desc(AnalyticsEvent.created_at))
    
    # Date filtering
    if start_date:
        try:
            since = datetime.fromisoformat(start_date.replace('Z', '+00:00'))
            query = query.filter(AnalyticsEvent.created_at >= since)
        except: pass
        
    if end_date:
        try:
            until = datetime.fromisoformat(end_date.replace('Z', '+00:00'))
            query = query.filter(AnalyticsEvent.created_at <= until)
        except: pass
    
    if event_type:
        query = query.filter(AnalyticsEvent.event_type == event_type)
        
    if visitor_id:
        query = query.filter(AnalyticsEvent.visitor_id == visitor_id)
        
    total = query.count()
    events = query.offset((page - 1) * page_size).limit(page_size).all()
    
    results = []
    for e in events:
        results.append({
            "id": e.id,
            "event_type": e.event_type,
            "page_url": e.page_url,
            "page_title": e.page_title,
            "event_data": e.event_data,
            "created_at": e.created_at.isoformat(),
            "visitor_id": e.visitor_id,
            "visitor_ip": e.visitor.ip_address if e.visitor else None,
            "visitor_location": e.visitor.location_data if e.visitor else None,
            "user": {"name": e.visitor.user.name, "email": e.visitor.user.email} if e.visitor and e.visitor.user else None
        })
        
    return {
        "isSuccess": True,
        "value": {
            "events": results,
            "total": total,
            "page": page,
            "page_size": page_size
        }
    }