screen-locker/screen_locker/_manual_sync.py

"""Ingest phone-/PC-synced manual workouts into ``log.json``.

A manual workout logged on another device (phone form, or the PC form pushed
from elsewhere) arrives via crdt-sync as a ``kind="manual_workout"`` payload
(see :func:`screen_locker._manual_workout.build_sync_payload`). This module
turns those synced payloads into ordinary signed ``manual_workout`` entries in
``log.json`` so they count toward the weekly minimum and — via the
optional ``on_ingested`` callback — earn the same shutdown/debt reward a
live-logged workout would.

Trust model: the PC never trusts the phone's derived fields. Every synced
manual is re-validated with the same :func:`validate_manual_workout` the local
form uses, re-built with :func:`build_entry`, and only then HMAC-signed by the
PC. A phone can therefore at most add a budget-limited, honour-system
``manual_workout`` — never a ``phone_verified`` one (the session path skips
manual records, see :mod:`screen_locker._workout_sync`).
"""

from __future__ import annotations

import logging
from typing import TYPE_CHECKING

from screen_locker._log_io import load_workout_log
from screen_locker._log_mixin import write_signed_entry
from screen_locker._manual_workout import (
    MANUAL_WORKOUT_SYNC_KIND,
    SPORT_OTHER,
    ManualWorkoutDraft,
    build_entry,
    is_budget_exhausted,
    validate_manual_workout,
)

if TYPE_CHECKING:
    from collections.abc import Callable, Iterable, Mapping
    from pathlib import Path

    OnIngestedCallback = Callable[[dict, "list[dict]"], None]

_logger = logging.getLogger(__name__)

# Key under which the source sync record id is stored in the ingested
# ``workout_data`` — makes re-ingestion of the same record idempotent.
_SYNC_ID_FIELD = "sync_record_id"

# The raw user inputs every real manual workout carries. A payload holding none
# of them is a metadata stub, not a damaged workout: there is nothing to
# recover and no amount of retrying will change that.
_SUBSTANTIVE_FIELDS = frozenset(
    {
        "sport",
        "start_time",
        "end_time",
        "location_name",
        "transport_method",
        "cost",
        "rpe",
        "went_well",
        "to_improve",
        "overall_feeling",
    }
)


def is_empty_stub(payload: Mapping[str, object]) -> bool:
    """True when ``payload`` carries no workout content at all.

    Observed in the wild as ``{"type", "kind", "date"}`` and nothing else --
    a record the phone created without ever attaching the workout. Every sync
    cycle re-reported it as malformed, so a permanent, unfixable condition
    produced an unbounded stream of identical warnings (four every 15 minutes),
    which is how the genuinely actionable lines got lost in the noise.
    """
    return not (_SUBSTANTIVE_FIELDS & set(payload))


def _coerce_int(value: object) -> int:
    """Coerce a JSON scalar to int; raise for a non-numeric (skips the record)."""
    if isinstance(value, (int, str)):
        return int(value)
    raise TypeError(value)


def reconstruct_draft(payload: Mapping[str, object]) -> ManualWorkoutDraft | None:
    """Rebuild a :class:`ManualWorkoutDraft` from a synced manual payload.

    Returns None if a required raw field is missing or mistyped, so a malformed
    record is skipped rather than crashing ingestion. Only the raw user inputs
    are read back — the derived fields (``source``, ``duration_minutes``,
    ``type``) are recomputed by :func:`build_entry` on the PC.
    """
    try:
        sport = str(payload["sport"])
        activity_type_other = (
            str(payload.get("activity_type", "")) if sport == SPORT_OTHER else ""
        )
        return ManualWorkoutDraft(
            sport=sport,
            start_time=str(payload["start_time"]),
            end_time=str(payload["end_time"]),
            location_name=str(payload["location_name"]),
            transport_method=str(payload["transport_method"]),
            cost=str(payload["cost"]),
            rpe=_coerce_int(payload["rpe"]),
            went_well=str(payload["went_well"]),
            to_improve=str(payload["to_improve"]),
            overall_feeling=str(payload["overall_feeling"]),
            reservation_phone=str(payload.get("reservation_phone", "")),
            techniques_practiced=str(payload.get("techniques_practiced", "")),
            warm_up_minutes=str(payload.get("warm_up_minutes", "")),
            pain_or_injury=str(payload.get("pain_or_injury", "none")),
            matches_won=_coerce_int(payload.get("matches_won", 0)),
            matches_lost=_coerce_int(payload.get("matches_lost", 0)),
            sets_won=_coerce_int(payload.get("sets_won", 0)),
            sets_lost=_coerce_int(payload.get("sets_lost", 0)),
            racket=str(payload.get("racket", "")),
            balls=str(payload.get("balls", "")),
            activity_type_other=activity_type_other,
            activity_details=str(payload.get("activity_details", "")),
            equipment=str(payload.get("equipment", "")),
        )
    except (KeyError, TypeError, ValueError) as exc:
        _logger.warning(
            "Synced manual workout payload is malformed (%s: %s) — SKIPPING "
            "this record, so it will not be logged or counted on the PC",
            type(exc).__name__,
            exc,
        )
        return None


def _draft_or_report(
    record_id: str, payload: Mapping[str, object]
) -> ManualWorkoutDraft | None:
    """Rebuild the draft, or report why it cannot be rebuilt and return None.

    The two failure modes are deliberately reported at different levels, because
    only one of them is actionable:

    * an **empty stub** carries no workout at all (seen as ``{type, kind,
      date}``) -- nothing to recover, nothing to fix, so re-reporting it at
      ``warning`` on every 15-minute sync cycle only teaches the reader to
      ignore this logger;
    * a **malformed** payload has real workout fields that would not parse,
      which means genuine credit may be going missing -- that stays loud.
    """
    if is_empty_stub(payload):
        _logger.info(
            "Manual record %s is an empty stub (keys: %s) — no workout data to "
            "ingest; skipping permanently, no credit is being lost",
            record_id,
            sorted(payload),
        )
        return None
    draft = reconstruct_draft(payload)
    if draft is None:
        _logger.warning(
            "Manual record %s is malformed — it HAS workout fields but they "
            "could not be parsed, so real credit may be lost",
            record_id,
        )
    return draft


def _already_ingested(logs: dict[str, list[dict]], record_id: str) -> bool:
    """True if any logged entry already carries this sync record id.

    Iterates the per-day lists (a day may hold several workouts); a cheap
    early-out before reconstructing the draft. The write chokepoint also dedups
    by ``workout_id``, so this and that are two guards on the same idempotency.
    """
    for entries in logs.values():
        for entry in entries:
            workout_data = entry.get("workout_data", {})
            if (
                isinstance(workout_data, dict)
                and workout_data.get(_SYNC_ID_FIELD) == record_id
            ):
                return True
    return False


def ingest_manual_records(
    log_file: Path,
    records: Iterable[tuple[str, Mapping[str, object]]],
    *,
    today: str | None = None,
    on_ingested: OnIngestedCallback | None = None,
) -> list[str]:
    """Ingest synced manual workouts into ``log.json``.

    For each ``(record_id, payload)`` tagged ``kind="manual_workout"``:
    reconstruct + re-validate the draft on the PC, enforce the rate budget, then
    HMAC-sign and APPEND it under its own ``date`` (a day may hold several
    workouts). Idempotent: dedup by ``record_id`` here and by ``workout_id`` in
    the write chokepoint, so re-syncing the same record is a no-op.

    When ``on_ingested`` is given, it's called ``on_ingested(entry, prior_entries)``
    right after each newly-appended record — ``prior_entries`` being that
    entry's own date's entries written before it — so the caller (see
    ``screen_lock.ScreenLocker._ingest_synced_manual_workouts``) can apply the
    identical live-workout shutdown/debt reward. This applies regardless of
    whether ``date`` is today or back-dated: there's only one current shutdown
    config, so a back-dated sync still pushes it exactly as a live workout
    would have that day.

    Returns the record ids actually ingested.
    """
    ingested: list[str] = []
    for record_id, payload in records:
        if payload.get("kind") != MANUAL_WORKOUT_SYNC_KIND:
            continue
        date = payload.get("date")
        if not isinstance(date, str):
            _logger.warning("Manual record %s has no date — skipped", record_id)
            continue
        if _already_ingested(load_workout_log(log_file), record_id):
            continue
        draft = _draft_or_report(record_id, payload)
        if draft is None:
            continue
        error = validate_manual_workout(draft)
        if error is not None:
            _logger.warning("Manual record %s invalid: %s", record_id, error)
            continue
        if is_budget_exhausted(log_file, today=today):
            _logger.info("Manual-workout budget exhausted — %s not ingested", record_id)
            continue
        entry = build_entry(draft)
        entry[_SYNC_ID_FIELD] = record_id
        result = write_signed_entry(log_file, date, entry)
        if not result.appended:
            # This is dedup working, not credit being lost: the workout is
            # already in the log under this same workout_id. It was logged at
            # warning and re-reported on every 15-minute cycle, which read like
            # a recurring fault and buried the lines that did need attention.
            _logger.info(
                "Manual record %s is already logged under the same workout_id "
                "— skipped as a duplicate (its credit is already counted)",
                record_id,
            )
            continue
        ingested.append(record_id)
        if on_ingested is not None:
            on_ingested(entry, result.prior_entries)
    return ingested