Three related fixes, all rooted in the same flaw: authorization and scoping
decided by the caller's role instead of by the environment the data belongs to.
1. Single-appointment access (clinic operations were entirely broken)
AppointmentController::canView/canManage only knew the patient, the owning
doctor and admin -- appointment.clinic was never consulted. A clinic user could
create an appointment through /my/appointment but got 403 on detail, edit,
move, reserve transfer/replace and status change, so nearly every appointment
operation failed in clinic mode.
AppointmentAccessChecker now decides from appointment.clinic: clinic owner,
member doctor (via ClinicDoctorPermissionChecker) and assigned secretary (via
active context + DoctorSecretary) are recognised. Actions reuse the existing
permission vocabulary, so active=false remains the single source of truth for
"collaboration ended". Cancellation is gated separately and an inline status on
PATCH /appointment/{uuid} cannot bypass that gate. The patient is narrowed to
view + cancel.
Also fixed alongside: listByDoctor now serves a clinic manager but scoped to
that clinic; todayStats gained an admin branch and no longer passes an array of
doctor ids as the clinic parameter; PatientController::appointments filters on
appointment.clinic instead of current membership, so deactivating a doctor no
longer erases clinic appointment history from the case file.
The doctor-only active_slot_key was reviewed and deliberately left alone -- a
doctor is one physical person, so adding clinic to the key would permit
double-booking, not fix a bug. Reasoning recorded on the entity.
2. Appointment registration and confirmation
Panel-created appointments are born pending ("ثبت شده") instead of confirmed.
Confirming is now an explicit act: POST /appointment/{uuid}/confirm transitions
the status, files the case file for the appointment's environment (reusing an
existing record or creating one) and registers full or partial payments on the
resulting visit -- all in one transaction.
AppointmentExpiryService would have expired those pending appointments the
moment their slot time passed; findExpiredPending is now limited to online
gateway holds, which are the only pendings carrying a TTL. A pending
appointment still occupies its slot, so the time stays reserved.
The admin panel gets a "قطعی کردن نوبت" modal showing the visit fee, each
selected service, the total, and paid/remaining/status. It is wired inside
AppointmentStatusDropdown, so picking "confirmed" anywhere (timeline, detail,
reserve list, info modal) goes through it and confirmation can never silently
skip the case file and payment.
3. Clinic case-file access
PatientRecordScopeResolver replaces the single-destination role mapping: the
active context decides, so a doctor invited into a clinic finally sees their
patients' records there. A clinic record is per-patient and shared by design,
so "their own patients" is derived from appointments with that doctor in that
clinic rather than from a new column. Clinic secretaries are limited to their
assigned doctors. Read and write share one rule, and out-of-scope records
report 404 so other environments are never disclosed.
Tests: 29 new cases across the three areas (clinic appointment access, confirm
flow, clinic record access). Full suite 466 tests, 2 pre-existing failures
unchanged. API docs updated for all three.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
findByClinicWithFilters left-joined specialties only for filtering, so
toListArray() lazy-loaded them per doctor (N+1). addSelect them and switch the
result fetch to Paginator(fetchJoinCollection: true) so LIMIT still paginates by
doctor.
Test infra: ApiTestCase::countQueries() (via doctrine.debug_data_holder).
Regression: tests/Doctor/ClinicDoctorListNPlusOneTest asserts the query count
does not grow with doctor count (4→10 without the fix).
Also relaxed AppointmentExpiryServiceTest's exact-count assertion (it counts all
stale pendings in the shared db_test, which accumulates) — logged test-isolation
debt as E6.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A non-unique index on (doctor_id, slot_start) plus a count-then-insert check
left a TOCTOU race: two concurrent requests could both pass isSlotTaken and
both insert. wrapInTransaction alone doesn't stop the phantom under InnoDB
REPEATABLE-READ.
Add a nullable, unique active_slot_key on Appointment = "doctorId:slotStart"
while the booking occupies the slot (pending/confirmed — in lockstep with
isSlotTaken); NULL once expired/completed/no_show/cancelled (NULLs don't collide
in a MySQL unique index, so released slots rebook freely). bookAtomically now:
catches the unique violation -> SlotTakenException, and expires lapsed pendings
in-transaction so the ~1-min window before the expiry cron doesn't wrongly block
rebooking. All three booking paths (online / my / admin) routed through it.
Migration backfills one row per (doctor, slot) — the latest id — so the index
builds even on dirty historical data without destructively cancelling bookings.
(Backfill surfaced a real pre-existing double-booked slot in dev data.)
Regression: tests/Appointment/SlotUniquenessTest. Adjusted the expiry-service
test fixture to use distinct slots (one live booking per slot is now enforced).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
AppointmentExpiryService ran one findPendingByAppointment query per expiring
booking. Add PaymentRepository::findPendingByAppointments (one IN query keyed
by appointment id) and use it. Test covers expiry + payment cancellation for
several appointments at once.
