TieMeasureFlow/src/backend/api/routers/statistics.py

"""Statistics router - SPC endpoints for Metrologist dashboard."""
from datetime import datetime
from src.backend.models.api.tipi import UtcNaiveInput

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

from src.backend.database import get_db
from src.backend.api.middleware.api_key import require_metrologist
from src.backend.models.orm.measurement import Measurement
from src.backend.models.orm.production import ProductionRun
from src.backend.models.orm.recipe import RecipeVersion
from src.backend.models.orm.task import RecipeSubtask, RecipeTask
from src.backend.models.orm.task_result import TaskResult  # #58: esiti dei task della sessione
from src.backend.models.orm.user import User
from src.backend.models.api.statistics import (
    CapabilityData,
    ControlChartData,
    HistogramData,
    SummaryData,
)
from src.backend.services.spc_service import (
    compute_capability,
    compute_control_chart,
    compute_histogram,
    compute_summary,
)

router = APIRouter(prefix="/api/statistics", tags=["statistics"])


async def _query_measurements(
    db: AsyncSession,
    recipe_id: int,
    version_id: int | None = None,
    subtask_id: int | None = None,
    date_from: datetime | None = None,
    date_to: datetime | None = None,
    operator_id: int | None = None,
    lot_number: str | None = None,
    serial_number: str | None = None,
    production_run_id: int | None = None,  # #58: solo le misure di una sessione
) -> list[Measurement]:
    """Query measurements with common filters.

    Returns list of Measurement ORM objects ordered by measured_at.
    """
    filters = []

    # recipe_id filter via RecipeVersion subquery
    if version_id is not None:
        filters.append(Measurement.version_id == version_id)
    else:
        version_ids = select(RecipeVersion.id).where(
            RecipeVersion.recipe_id == recipe_id
        )
        filters.append(Measurement.version_id.in_(version_ids))

    if subtask_id is not None:
        filters.append(Measurement.subtask_id == subtask_id)
    if date_from is not None:
        filters.append(Measurement.measured_at >= date_from)
    if date_to is not None:
        filters.append(Measurement.measured_at <= date_to)
    if operator_id is not None:
        filters.append(Measurement.measured_by == operator_id)
    if lot_number is not None:
        filters.append(Measurement.lot_number == lot_number)
    if serial_number is not None:
        filters.append(Measurement.serial_number == serial_number)

    if production_run_id is not None:
        filters.append(Measurement.production_run_id == production_run_id)

    query = (
        select(Measurement)
        .where(and_(*filters) if filters else True)
        .order_by(Measurement.measured_at.asc())
    )
    result = await db.execute(query)
    return list(result.scalars().all())


async def _get_subtask_tolerances(
    db: AsyncSession,
    subtask_id: int,
) -> dict:
    """Get tolerance limits for a specific subtask."""
    result = await db.execute(
        select(RecipeSubtask).where(RecipeSubtask.id == subtask_id)
    )
    subtask = result.scalar_one_or_none()
    if subtask is None:
        return {"utl": None, "uwl": None, "lwl": None, "ltl": None, "nominal": None}
    return {
        "utl": float(subtask.utl) if subtask.utl is not None else None,
        "uwl": float(subtask.uwl) if subtask.uwl is not None else None,
        "lwl": float(subtask.lwl) if subtask.lwl is not None else None,
        "ltl": float(subtask.ltl) if subtask.ltl is not None else None,
        "nominal": float(subtask.nominal) if subtask.nominal is not None else None,
    }


@router.get("/summary", response_model=SummaryData)
async def get_summary(
    recipe_id: int = Query(...),
    version_id: int | None = Query(None),
    subtask_id: int | None = Query(None),
    date_from: UtcNaiveInput = Query(None),
    date_to: UtcNaiveInput = Query(None),
    operator_id: int | None = Query(None),
    lot_number: str | None = Query(None),
    serial_number: str | None = Query(None),
    production_run_id: int | None = Query(None),  # #58
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> SummaryData:
    """Get pass/fail/warning summary for filtered measurements."""
    measurements = await _query_measurements(
        db, recipe_id, version_id, subtask_id,
        date_from, date_to, operator_id, lot_number, serial_number,
        production_run_id=production_run_id,  # #58
    )
    pass_fail_values = [m.pass_fail for m in measurements]
    return compute_summary(pass_fail_values)


@router.get("/capability", response_model=CapabilityData)
async def get_capability(
    recipe_id: int = Query(...),
    subtask_id: int = Query(...),
    version_id: int | None = Query(None),
    date_from: UtcNaiveInput = Query(None),
    date_to: UtcNaiveInput = Query(None),
    operator_id: int | None = Query(None),
    lot_number: str | None = Query(None),
    serial_number: str | None = Query(None),
    production_run_id: int | None = Query(None),  # #58
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> CapabilityData:
    """Get capability indices (Cp/Cpk/Pp/Ppk) for a specific subtask."""
    measurements = await _query_measurements(
        db, recipe_id, version_id, subtask_id,
        date_from, date_to, operator_id, lot_number, serial_number,
        production_run_id=production_run_id,  # #58
    )
    values = [float(m.value) for m in measurements]
    tol = await _get_subtask_tolerances(db, subtask_id)
    return compute_capability(values, tol["utl"], tol["ltl"], tol["nominal"])


@router.get("/control-chart", response_model=ControlChartData)
async def get_control_chart(
    recipe_id: int = Query(...),
    subtask_id: int = Query(...),
    version_id: int | None = Query(None),
    date_from: UtcNaiveInput = Query(None),
    date_to: UtcNaiveInput = Query(None),
    operator_id: int | None = Query(None),
    lot_number: str | None = Query(None),
    serial_number: str | None = Query(None),
    production_run_id: int | None = Query(None),  # #58
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> ControlChartData:
    """Get control chart data with UCL/LCL and OOC detection."""
    measurements = await _query_measurements(
        db, recipe_id, version_id, subtask_id,
        date_from, date_to, operator_id, lot_number, serial_number,
        production_run_id=production_run_id,  # #58
    )
    values = [float(m.value) for m in measurements]
    timestamps = [m.measured_at for m in measurements]
    tol = await _get_subtask_tolerances(db, subtask_id)
    return compute_control_chart(
        values, timestamps,
        tol["utl"], tol["uwl"], tol["lwl"], tol["ltl"], tol["nominal"],
    )


@router.get("/histogram", response_model=HistogramData)
async def get_histogram(
    recipe_id: int = Query(...),
    subtask_id: int = Query(...),
    version_id: int | None = Query(None),
    date_from: UtcNaiveInput = Query(None),
    date_to: UtcNaiveInput = Query(None),
    operator_id: int | None = Query(None),
    lot_number: str | None = Query(None),
    serial_number: str | None = Query(None),
    production_run_id: int | None = Query(None),  # #58
    n_bins: int = Query(20, ge=5, le=100),
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> HistogramData:
    """Get histogram data with normal curve overlay."""
    measurements = await _query_measurements(
        db, recipe_id, version_id, subtask_id,
        date_from, date_to, operator_id, lot_number, serial_number,
        production_run_id=production_run_id,  # #58
    )
    values = [float(m.value) for m in measurements]
    return compute_histogram(values, n_bins)


@router.get("/subtasks")
async def get_recipe_subtasks(
    recipe_id: int = Query(...),
    version_id: int | None = Query(None),
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> list[dict]:
    """Get subtasks for a recipe (for filter dropdown).

    Returns subtask id, marker_number, description, and parent task title.
    """
    # Get version_id: use provided or find current version
    if version_id is None:
        ver_result = await db.execute(
            select(RecipeVersion.id).where(
                RecipeVersion.recipe_id == recipe_id,
                RecipeVersion.is_current == True,
            )
        )
        version_id = ver_result.scalar_one_or_none()
        if version_id is None:
            return []

    # Get tasks and subtasks for this version
    tasks_result = await db.execute(
        select(RecipeTask).where(RecipeTask.version_id == version_id)
        .order_by(RecipeTask.order_index)
    )
    tasks = tasks_result.scalars().all()

    subtasks_list = []
    for task in tasks:
        for st in sorted(task.subtasks, key=lambda s: s.marker_number):
            subtasks_list.append({
                "id": st.id,
                "marker_number": st.marker_number,
                "description": st.description,
                "task_title": task.title,
                "nominal": float(st.nominal) if st.nominal is not None else None,
                "utl": float(st.utl) if st.utl is not None else None,
                "ltl": float(st.ltl) if st.ltl is not None else None,
            })

    return subtasks_list


@router.get("/runs")
async def get_recipe_runs(
    recipe_id: int = Query(...),
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> list[dict]:
    """#58: production runs of a recipe (for the "Sessione" filter), newest first."""
    runs = (await db.execute(
        select(ProductionRun)
        .where(ProductionRun.recipe_id == recipe_id)
        .order_by(ProductionRun.started_at.desc(), ProductionRun.id.desc())
    )).scalars().all()
    # #58: conta solo le misure dei run della ricetta, non tutta la tabella.
    run_ids = [r.id for r in runs]
    counts = {}
    if run_ids:
        counts = dict((await db.execute(
            select(Measurement.production_run_id, func.count(Measurement.id))
            .where(Measurement.production_run_id.in_(run_ids))
            .group_by(Measurement.production_run_id)
        )).all())
    operator_ids = {r.operator_id for r in runs}
    names = {}
    if operator_ids:
        names = dict((await db.execute(
            select(User.id, User.display_name).where(User.id.in_(operator_ids))
        )).all())
    return [
        {
            "id": r.id,
            "station_id": r.station_id,
            "station_name": r.station.name if r.station else None,
            "operator_id": r.operator_id,
            "operator_name": names.get(r.operator_id),
            "started_at": r.started_at.isoformat() if r.started_at else None,
            "closed_at": r.closed_at.isoformat() if r.closed_at else None,
            "status": r.status,
            "lot_number": r.lot_number,
            "serial_number": r.serial_number,
            "cycle_count": r.cycle_count,
            "measurement_count": counts.get(r.id, 0),
        }
        for r in runs
    ]


@router.get("/runs/{run_id}/task-results")
async def get_run_task_results(
    run_id: int,
    user: User = Depends(require_metrologist),
    db: AsyncSession = Depends(get_db),
) -> dict:
    """#58: esiti, note, campi e allegati dei task di una sessione di produzione.

    Endpoint di statistica (require_metrologist): GET /api/task-results vuole
    il ruolo MeasurementTec e non si puo' riusare qui.
    """
    run = (await db.execute(
        select(ProductionRun.id).where(ProductionRun.id == run_id)
    )).scalar_one_or_none()
    if run is None:
        raise HTTPException(status_code=404, detail="Sessione non trovata")

    # #58: una query per i risultati della run, una per i task, una per gli utenti.
    results = (await db.execute(
        select(TaskResult)
        .where(TaskResult.production_run_id == run_id)
        .order_by(TaskResult.created_at.asc(), TaskResult.id.asc())
    )).scalars().all()

    task_ids = {r.task_id for r in results}
    tasks = {}
    if task_ids:
        tasks = {t.id: t for t in (await db.execute(
            select(RecipeTask).where(RecipeTask.id.in_(task_ids))
        )).scalars().all()}
    user_ids = {r.user_id for r in results}
    names = {}
    if user_ids:
        names = dict((await db.execute(
            select(User.id, User.display_name).where(User.id.in_(user_ids))
        )).all())

    counts = {"ok": 0, "nok": 0, "done": 0}
    out = []
    for r in results:
        counts[r.outcome] = counts.get(r.outcome, 0) + 1
        task = tasks.get(r.task_id)
        # #58: etichette dai campi del task; se il campo e' stato tolto vale la chiave.
        labels = {
            f.get("key"): f.get("label") or f.get("key")
            for f in (task.fields or []) if isinstance(f, dict)
        } if task else {}
        out.append({
            "id": r.id,
            "task_id": r.task_id,
            "task_title": task.title if task else None,
            "task_type": task.task_type if task else None,
            "outcome": r.outcome,
            "note": r.note,
            "created_at": r.created_at.isoformat() if r.created_at else None,
            "user_name": names.get(r.user_id),
            "values": [
                {"key": k, "label": labels.get(k, k), "value": v}
                for k, v in (r.values or {}).items()
            ],
            "attachments": [
                {
                    "key": a.get("key"),
                    "label": labels.get(a.get("key"), a.get("key")),
                    "name": a.get("name"),
                    "path": a.get("path"),
                }
                for a in (r.attachments or []) if isinstance(a, dict)
            ],
        })
    return {"run_id": run_id, "counts": counts, "results": out}