kmtnflower/backend/app/api/stats.py

"""Stats API: trigger log parsing and query aggregated visitor data."""

import asyncio
from datetime import date, timedelta
from typing import AsyncGenerator, Optional

from fastapi import APIRouter, Depends, Query, HTTPException, status
from fastapi.responses import StreamingResponse
from sqlalchemy import desc, func, select
from sqlalchemy.ext.asyncio import AsyncSession

from app.auth import decode_token, get_current_admin_user
from app.database import get_session
from app.log_parser import process_unparsed_logs
from app.models import LogParseState, VisitorStatDaily, VisitorStatGeo
from app.user_models import User

router = APIRouter()


# ── Helpers ────────────────────────────────────────────────────

def _default_date_range() -> tuple[date, date]:
    """Return (start_date, end_date) for last 30 days."""
    end = date.today() - timedelta(days=1)  # today isn't parsed yet
    start = end - timedelta(days=29)
    return start, end


async def _verify_admin_from_token(token: str) -> User:
    """Verify admin from a raw JWT token string (for SSE query-param auth)."""
    from app.database import async_session
    from sqlalchemy import select as sa_select

    payload = decode_token(token)
    user_id = int(payload["sub"])
    if not payload.get("is_admin"):
        raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail="Admin access required")

    async with async_session() as session:
        result = await session.execute(sa_select(User).where(User.id == user_id))
        user = result.scalar_one_or_none()
        if user is None:
            raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="User not found")
        return user


# ── SSE parse endpoint ───────────────────────────────────────

@router.get("/parse")
async def trigger_parse(token: Optional[str] = Query(None)):
    """Trigger log file parsing. Streams progress via Server-Sent Events.

    Accepts JWT via `token` query parameter (EventSource can't send headers).
    """
    # Authenticate — SSE can't send Authorization headers, so use query param
    if token:
        await _verify_admin_from_token(token)
    else:
        raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="Token required")

    async def event_generator() -> AsyncGenerator[str, None]:
        # Collect progress events, then yield them
        events: list[tuple[str, int, int]] = []
        result = await process_unparsed_logs(lambda fd, c, t: events.append((fd, c, t)))

        for file_date, current, total in events:
            yield f'event: progress\ndata: {{"file": "{file_date}", "current": {current}, "total": {total}}}\n\n'

        yield f'event: done\ndata: {{"files_processed": {result["files_processed"]}, "rows_inserted": {result["rows_inserted"]}}}\n\n'

    return StreamingResponse(
        event_generator(),
        media_type="text/event-stream",
        headers={
            "Cache-Control": "no-cache",
            "Connection": "keep-alive",
            "X-Accel-Buffering": "no",  # Disable nginx buffering for SSE
        },
    )


# ── Summary ────────────────────────────────────────────────────

@router.get("/summary")
async def get_summary(
    start_date: Optional[date] = Query(None),
    end_date: Optional[date] = Query(None),
    session: AsyncSession = Depends(get_session),
    _: User = Depends(get_current_admin_user),
):
    """Get total unique visitors for a date range."""
    if start_date is None or end_date is None:
        start_date, end_date = _default_date_range()

    result = await session.execute(
        select(func.coalesce(func.sum(VisitorStatDaily.unique_visitors), 0)).where(
            VisitorStatDaily.date >= start_date,
            VisitorStatDaily.date <= end_date,
        )
    )
    total_unique = result.scalar()

    return {"total_unique_visitors": total_unique}


# ── Time series ────────────────────────────────────────────────

@router.get("/timeseries")
async def get_timeseries(
    start_date: Optional[date] = Query(None),
    end_date: Optional[date] = Query(None),
    session: AsyncSession = Depends(get_session),
    _: User = Depends(get_current_admin_user),
):
    """Get daily unique visitor counts for a date range."""
    if start_date is None or end_date is None:
        start_date, end_date = _default_date_range()

    stmt = (
        select(
            VisitorStatDaily.date,
            VisitorStatDaily.unique_visitors,
        )
        .where(
            VisitorStatDaily.date >= start_date,
            VisitorStatDaily.date <= end_date,
        )
        .order_by(VisitorStatDaily.date)
    )
    result = await session.execute(stmt)
    rows = result.all()
    return [
        {"date": row.date.isoformat(), "unique_visitors": row.unique_visitors}
        for row in rows
    ]


# ── Geo ────────────────────────────────────────────────────────

@router.get("/geo")
async def get_geo_stats(
    start_date: Optional[date] = Query(None),
    end_date: Optional[date] = Query(None),
    session: AsyncSession = Depends(get_session),
    _: User = Depends(get_current_admin_user),
):
    """Get geographic distribution of visitors for a date range.

    Aggregates by country (country, country_code, lat, lon), summing visitor_count.
    Returns one row per country/region for the map to render.
    """
    if start_date is None or end_date is None:
        start_date, end_date = _default_date_range()

    stmt = (
        select(
            VisitorStatGeo.country,
            VisitorStatGeo.country_code,
            VisitorStatGeo.latitude,
            VisitorStatGeo.longitude,
            func.sum(VisitorStatGeo.visitor_count).label("total_visitors"),
        )
        .where(
            VisitorStatGeo.date >= start_date,
            VisitorStatGeo.date <= end_date,
        )
        .group_by(
            VisitorStatGeo.country,
            VisitorStatGeo.country_code,
            VisitorStatGeo.latitude,
            VisitorStatGeo.longitude,
        )
        .order_by(desc(func.sum(VisitorStatGeo.visitor_count)))
    )
    result = await session.execute(stmt)
    rows = result.all()
    return [
        {
            "country": row.country or "Unknown",
            "country_code": row.country_code or "",
            "latitude": row.latitude,
            "longitude": row.longitude,
            "visitors": row.total_visitors,
        }
        for row in rows
    ]