Files
HealthLog/src/lib/validations/backup.ts
T
Marc-André Bombeck 893db887e9 Record why a document's index attempt failed and say so
A content-index attempt that failed (raster failure, a scan with no
text layer, a provider error) recorded its reason only in the server's
observability stream. In the vault, hasContentIndex false looked
identical for a document nobody ever tried and one that failed three
times, and the detail sheet had nothing to say about either.

Apply the summaryState pattern to the index. Two additive columns on
InboundDocument (migration 0333) record when the last attempt finished
and why it produced no index, as a closed reason set: local-empty,
local-unsupported, decrypt-error, raster-failed, provider-error,
pdf-needs-anthropic, empty-transcription. The writer sits at the choke
point the auto job and the corpus backfill share, and in the manual
index route; a success clears the reason. The provider path's specific
failure now survives the local fallback, since "no text" is less
actionable than "the PDF would not render". rasterizePdf names its
failure class so a build that cannot rasterize at all is distinguished
from a PDF that would not render. New guard in both index paths: an
empty provider transcription never becomes a successful empty index.

The detail sheet renders the reason in plain words (muted, six
locales). The DTO and OpenAPI carry the two fields; the disaster
recovery backup and restore carry them verbatim like their summary
siblings.

Refs #776
2026-08-13 12:40:50 +02:00

1248 lines
49 KiB
TypeScript

/**
* Zod schemas + helpers for the per-user `DataBackup` JSON payload.
*
* Single source of truth shared by:
* - the pg-boss `data-backup` worker that writes a backup,
* - `GET /api/admin/backups/[id]/download` that streams it back as JSON,
* - `POST /api/admin/backups/upload` that ingests an admin-supplied file,
* - `POST /api/admin/backups/[id]/restore` that re-creates DB rows from it.
*
* The schema is intentionally permissive about *extra* fields (`.passthrough`)
* so older snapshots written before a column was added still parse and so
* future minor additions don't break old admins. Required fields are
* deliberately tight — any drift surfaces as a validation error rather than
* a silent data-loss restore.
*
* `schemaVersion` is the migration handle. Bump when the on-disk shape
* changes incompatibly. The current writer (worker) historically did NOT
* include this field; `parseBackupPayload()` defaults it to "1" so legacy
* blobs continue to round-trip and the upload validator can still produce
* a useful summary for them.
*/
import { z } from "zod/v4";
import {
AllergyCategory,
AllergySeverity,
AllergyStatus,
AllergyType,
CervicalMucus,
CervixFirmness,
CervixOpening,
CervixPosition,
ContraceptiveKind,
CycleTrackingGoal,
DocumentSummaryState,
EncounterKind,
EncounterStatus,
FamilyRelationship,
FlowLevel,
GlucoseContext,
HomeTestResult,
IllnessLifecycle,
IllnessType,
InboundDocumentKind,
InboundDocumentStatus,
InjectionSite,
IntakeAttributionSource,
IntakeSource,
MeasurementAggregationProvenance,
MeasurementSource,
MeasurementType,
MedicationCategory,
MedicationDeliveryForm,
MedicationScheduleType,
MedicationSideEffectCategory,
MedicationSideEffectEntry,
OvulationTest,
RhythmClassification,
SecondarySymptom,
SleepStage,
VaccinationSite,
} from "@/generated/prisma/enums";
import {
DEFAULT_HEALTH_PROFILE_AI_SECTIONS,
healthProfileAiSectionSchema,
healthProfileFactKindSchema,
isHealthProfileFactValue,
} from "@/lib/validations/health-profile-facts";
import {
advanceDirectiveStatusSchema,
emergencyBloodTypeSchema,
organDonorStatusSchema,
} from "@/lib/validations/emergency-profile";
export const BACKUP_SCHEMA_VERSION = "2" as const;
const LEGACY_BACKUP_SCHEMA_VERSION = "1" as const;
const isoDateTime = z
.string()
.min(1)
.refine((s) => !Number.isNaN(Date.parse(s)), {
message: "Expected ISO-8601 date-time string",
});
const base64BytesSchema = z
.string()
.min(1)
.refine(
(value) => value.length % 4 === 0 && /^[A-Za-z0-9+/]+={0,2}$/.test(value),
{ message: "Expected base64-encoded encrypted bytes" },
);
const measurementSchema = z
.object({
id: z.string().min(1).optional(),
type: z.enum(MeasurementType),
value: z.number(),
valueMin: z.number().nullable().optional(),
valueMax: z.number().nullable().optional(),
unit: z.string().min(1),
measuredAt: isoDateTime,
source: z.enum(MeasurementSource).optional(),
notes: z.string().nullable().optional(),
notesEncrypted: base64BytesSchema.nullable().optional(),
externalId: z.string().nullable().optional(),
externalSourceVersion: z.string().nullable().optional(),
// Authority of a `stats:` aggregate. `.optional()` keeps every backup
// written before this field rode the payload parseable.
aggregationProvenance: z
.enum(MeasurementAggregationProvenance)
.nullable()
.optional(),
glucoseContext: z.enum(GlucoseContext).nullable().optional(),
sleepStage: z.enum(SleepStage).nullable().optional(),
rhythmClassification: z.enum(RhythmClassification).nullable().optional(),
deviceType: z.string().nullable().optional(),
syncVersion: z.number().int().optional(),
deletedAt: isoDateTime.nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const medicationScheduleSchema = z
.object({
id: z.string().min(1).optional(),
windowStart: z.string().min(1),
windowEnd: z.string().min(1),
label: z.string().nullable().optional(),
dose: z.string().nullable().optional(),
// #219 — per-schedule units per dose. Serialised as a Decimal string in a
// DR file (or a number in a hand-authored one); NULL means inherit.
unitsPerDose: z.union([z.string(), z.number()]).nullable().optional(),
daysOfWeek: z.string().nullable().optional(),
timesOfDay: z.array(z.string()).optional(),
reminderGraceMinutes: z.number().int().nullable().optional(),
rrule: z.string().nullable().optional(),
rollingIntervalDays: z.number().int().nullable().optional(),
scheduleType: z.enum(MedicationScheduleType).optional(),
cyclicOnWeeks: z.number().int().nullable().optional(),
cyclicOffWeeks: z.number().int().nullable().optional(),
doseWindows: z.unknown().nullable().optional(),
})
.passthrough();
/**
* One recorded side effect, carried inside its medication.
*
* `notes` is the decrypted note in a portable export and the row's legacy
* plaintext column in a canonical DR file; `notesEncrypted` is the base64
* ciphertext and rides only in the DR case. The restore prefers the ciphertext
* and encrypts the plaintext when that is all the file has, so neither shape
* loses the note and neither writes plaintext back into the column.
*/
const medicationSideEffectSchema = z
.object({
id: z.string().min(1).optional(),
occurredAt: isoDateTime,
category: z.enum(MedicationSideEffectCategory),
entry: z.enum(MedicationSideEffectEntry),
severity: z.number().int().min(1).max(5),
notes: z.string().nullable().optional(),
notesEncrypted: base64BytesSchema.nullable().optional(),
createdAt: isoDateTime.optional(),
})
.passthrough();
const medicationSchema = z
.object({
id: z.string().min(1).optional(),
name: z.string().min(1),
dose: z.string(),
treatmentClass: z.enum(MedicationCategory).optional(),
dosesPerUnit: z.number().int().nullable().optional(),
unitsPerDose: z.string().min(1).optional(),
active: z.boolean().optional(),
notificationsEnabled: z.boolean().optional(),
pausedAt: isoDateTime.nullable().optional(),
snoozedUntil: isoDateTime.nullable().optional(),
startsOn: isoDateTime.nullable().optional(),
endsOn: isoDateTime.nullable().optional(),
oneShot: z.boolean().optional(),
asNeeded: z.boolean().optional(),
deliveryForm: z.enum(MedicationDeliveryForm).optional(),
trackInjectionSites: z.boolean().optional(),
allowedInjectionSites: z.array(z.enum(InjectionSite)).optional(),
liveActivityEnabled: z.boolean().optional(),
criticalAlarmEnabled: z.boolean().optional(),
atcCode: z.string().nullable().optional(),
rxNormCode: z.string().nullable().optional(),
lowStockNotifiedAt: isoDateTime.nullable().optional(),
lowStockNotifiedThresholdDays: z.number().int().nullable().optional(),
reorderLeadDays: z.number().int().nullable().optional(),
externalSource: z.enum(IntakeSource).nullable().optional(),
externalId: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
schedules: z.array(medicationScheduleSchema).default([]),
// Defaulted so a file written before side effects rode the wire still
// parses, and a drug with none writes [].
sideEffects: z.array(medicationSideEffectSchema).default([]),
})
.passthrough();
const intakeEventSchema = z
.object({
id: z.string().min(1).optional(),
medicationId: z.string().min(1).optional(),
medication: z.string().min(1),
scheduledFor: isoDateTime,
takenAt: isoDateTime.nullable().optional(),
skipped: z.boolean().optional(),
autoMissed: z.boolean().optional(),
attributionSource: z.enum(IntakeAttributionSource).optional(),
source: z.enum(IntakeSource).optional(),
idempotencyKey: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
injectionSite: z.enum(InjectionSite).nullable().optional(),
doseTaken: z.string().nullable().optional(),
inventoryConsumption: z.unknown().nullable().optional(),
externalId: z.string().nullable().optional(),
updatedAt: isoDateTime.optional(),
syncVersion: z.number().int().optional(),
deletedAt: isoDateTime.nullable().optional(),
})
.passthrough();
/**
* `MoodEntry.tags` is stored as a JSON-array-as-string in the
* `mood_entries.tags` column ("[\"work\",\"sleep\"]"). The previous
* schema accepted any `string` here, so a malformed blob in a backup
* (or one tampered with mid-restore) would land in the DB and crash
* downstream readers that `JSON.parse` it. We now refine to one of
* `null` / empty-string (legacy null wire format) / a JSON string
* that parses to a `string[]`. v1.4.15 H2.
*/
const moodEntryTagsSchema = z
.union([z.null(), z.string()])
.nullable()
.optional()
.refine(
(v) => {
if (v == null || v === "") return true;
try {
const parsed = JSON.parse(v) as unknown;
return (
Array.isArray(parsed) && parsed.every((x) => typeof x === "string")
);
} catch {
return false;
}
},
{ message: "tags must be null, empty, or a JSON array of strings" },
);
const moodFactorSchema = z
.object({
key: z.string().min(1),
rating: z.number().int(),
})
.passthrough();
const moodEntrySchema = z
.object({
id: z.string().min(1).optional(),
date: z.string().min(1),
mood: z.string().min(1),
score: z.number().int().min(0).max(10),
// The five level-A values. Optional and nullable: a file written before
// they existed carries none, and an entry whose sliders were never touched
// carries nulls. Absent restores as NULL, never as a defaulted midpoint —
// a restore that invents an answer is worse than one that admits none.
a1: z.number().int().min(0).max(10).nullable().optional(),
a2: z.number().int().min(0).max(10).nullable().optional(),
a3: z.number().int().min(0).max(10).nullable().optional(),
a4: z.number().int().min(0).max(10).nullable().optional(),
a5: z.number().int().min(0).max(10).nullable().optional(),
tags: moodEntryTagsSchema,
// The free text, in both storage shapes. A portable export carries the
// decrypted `note`; a disaster-recovery file carries `noteEncrypted` plus
// whatever legacy plaintext the row still holds. Same pair, and same
// handling on restore, as `Measurement.notes` / `notesEncrypted`.
note: z.string().nullable().optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
source: z.string().min(1).optional(),
externalId: z.string().nullable().optional(),
loggedAt: isoDateTime,
// The IANA zone the `date` string is anchored to. Absent means the legacy
// Europe/Berlin reading, which is what a row written before v1.4.25 means
// by a NULL here — so an old file keeps its old meaning exactly.
tz: z.string().nullable().optional(),
syncedAt: isoDateTime.optional(),
syncVersion: z.number().int().optional(),
deletedAt: isoDateTime.nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
factors: z.array(moodFactorSchema).default([]),
// BINARY structured-tag keys. Separate from `factors`, which keeps its
// RATED-only meaning, so every file written before this one parses
// unchanged and restores identically: absent reads as none.
structuredTags: z.array(z.string().min(1)).default([]),
// The day context, when the entry had one. Optional and nullable at the
// top: a file written before this existed carries no `context` key at all
// and must parse and restore exactly as it did. Every field inside is
// optional for the same reason a context row's columns are nullable —
// only the sections somebody opened say anything.
//
// The two multi-selects ride as the stored JSON string rather than as
// arrays, so the restore writes the column back verbatim and a file cannot
// arrive with a list the parser and the column disagree about. Nothing is
// re-validated against the vocabulary here: a backup restores what the
// account had, and a key retired between the export and the restore is
// still that person's answer.
context: z
.object({
workStatus: z.string().nullable().optional(),
workMinutes: z.number().int().nullable().optional(),
overtimeMinutes: z.number().int().nullable().optional(),
workLoad: z.number().int().nullable().optional(),
workSatisfaction: z.number().int().nullable().optional(),
contactCircles: z.string().nullable().optional(),
contactForm: z.string().nullable().optional(),
contactExtent: z.string().nullable().optional(),
contactQuality: z.number().int().nullable().optional(),
contactSupport: z.number().int().nullable().optional(),
leisureCategories: z.string().nullable().optional(),
leisureMinutes: z.number().int().nullable().optional(),
leisureJoy: z.number().int().nullable().optional(),
leisureRecovery: z.number().int().nullable().optional(),
eventType: z.string().nullable().optional(),
eventValence: z.number().int().nullable().optional(),
eventAt: isoDateTime.nullable().optional(),
// Same pair as the entry's own note: a portable file carries the
// plain text, a disaster-recovery file the ciphertext.
note: z.string().nullable().optional(),
notesEncrypted: base64BytesSchema.nullable().optional(),
})
.passthrough()
.nullable()
.optional(),
})
.passthrough();
/* ── v1.15.0 cycle-tracking backup shapes ──────────────────────────── */
/**
* One menstrual-cycle span. `startDate` is the natural per-user key
* (matching the `(userId, startDate)` unique), so a restore upserts on it.
* Predicted (forecast) rows are excluded from the backup — only observed
* history round-trips.
*/
const cycleSpanSchema = z
.object({
id: z.string().min(1).optional(),
startDate: z.string().min(1),
endDate: z.string().nullable().optional(),
periodEndDate: z.string().nullable().optional(),
lengthDays: z.number().int().nullable().optional(),
ovulationDate: z.string().nullable().optional(),
ovulationConfirmed: z.boolean().optional(),
isPredicted: z.boolean().optional(),
tz: z.string().nullable().optional(),
syncVersion: z.number().int().optional(),
deletedAt: isoDateTime.nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
/**
* One cycle day-log. `notesEncrypted` is carried as the AES-256-GCM
* ciphertext envelope verbatim — the backup never decrypts it, so the
* owner's free-text note round-trips encrypted (and a wrong-surface leak is
* impossible). `symptomKeys` carries the seeded catalogue keys so the
* restore can re-link without exporting internal join ids, and
* `symptomSeverities` carries how hard each of them hit.
*/
/**
* The intensity recorded against one of a day's symptoms.
*
* A sparse annotation over `symptomKeys`, not a replacement for it: the keys
* still say which symptoms the day had, this says how hard the ones that were
* rated hit. Only rated links appear, so a missing entry means the person never
* put a number on it. `severity` is left as a plain integer here on purpose —
* the 1-4 range is checked when the row is written, so a file carrying a value
* outside it loses that one intensity instead of failing to restore at all.
*/
const cycleSymptomSeveritySchema = z
.object({
key: z.string().min(1),
severity: z.number().int().nullable().optional(),
})
.passthrough();
const cycleDayLogSchema = z
.object({
id: z.string().min(1).optional(),
date: z.string().min(1),
cycleId: z.string().nullable().optional(),
flow: z.enum(FlowLevel).nullable().optional(),
intermenstrualBleeding: z.boolean().optional(),
basalBodyTempC: z.number().nullable().optional(),
temperatureExcluded: z.boolean().optional(),
ovulationTest: z.enum(OvulationTest).nullable().optional(),
cervicalMucus: z.enum(CervicalMucus).nullable().optional(),
cervixPosition: z.enum(CervixPosition).nullable().optional(),
cervixFirmness: z.enum(CervixFirmness).nullable().optional(),
cervixOpening: z.enum(CervixOpening).nullable().optional(),
sexualActivity: z.boolean().optional(),
protectedSex: z.boolean().nullable().optional(),
pregnancyTest: z.enum(HomeTestResult).nullable().optional(),
progesteroneTest: z.enum(HomeTestResult).nullable().optional(),
contraceptive: z.enum(ContraceptiveKind).nullable().optional(),
sensitiveEncrypted: z.string().nullable().optional(),
notesEncrypted: z.string().nullable().optional(),
source: z.enum(MeasurementSource).optional(),
externalId: z.string().nullable().optional(),
tz: z.string().nullable().optional(),
syncVersion: z.number().int().optional(),
deletedAt: isoDateTime.nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
symptomKeys: z.array(z.string()).default([]),
symptomSeverities: z.array(cycleSymptomSeveritySchema).default([]),
})
.passthrough();
/** Cycle-tracking preferences (one row per user). */
const cycleProfileSchema = z
.object({
id: z.string().min(1).optional(),
goal: z.enum(CycleTrackingGoal).optional(),
cycleTrackingEnabled: z.boolean().nullable().optional(),
typicalCycleLength: z.number().int().nullable().optional(),
typicalPeriodLength: z.number().int().nullable().optional(),
lutealPhaseLength: z.number().int().nullable().optional(),
secondarySymptom: z.enum(SecondarySymptom).nullable().optional(),
predictionEnabled: z.boolean().optional(),
rawChartMode: z.boolean().optional(),
discreetNotifications: z.boolean().optional(),
sensitiveCategoryEncryption: z.boolean().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
/**
* A mood tag the account created, as opposed to the seeded catalogue.
*
* Carried for the same reason as a custom cycle symptom, but the failure it
* prevents is louder: the restore resolves an entry's rated factors by key and
* throws on one it cannot find, so an account with a single custom rated tag
* could not be restored at all.
*/
const customMoodTagSchema = z
.object({
id: z.string().min(1).optional(),
key: z.string().min(1),
labelKey: z.string().min(1),
categoryId: z.string().min(1),
// "BINARY" | "RATED", enforced in app code rather than a DB enum.
kind: z.string().min(1),
isActive: z.boolean().default(true),
icon: z.string().nullable().optional(),
sortOrder: z.number().int().default(0),
// The user's own words for the tag, encrypted at rest and carried
// verbatim. Without it the tag comes back as a bare key and the person
// who named it "Migräne" gets `custom:cm3x9…` instead.
labelEncrypted: z.string().nullable().optional(),
// The scale a RATED factor was recorded on. `inverse` marks a factor
// where a HIGH value is bad (stress, conflict). Dropping it does not lose
// a label, it silently reverses the meaning of every rating already
// stored against that tag.
scaleMin: z.number().int().default(1),
scaleMax: z.number().int().default(5),
inverse: z.boolean().default(false),
})
.passthrough();
/**
* A day's total for one nutrient, from one source.
*
* The export has written these since v1.29 and the canonical schema never
* declared them, so `.passthrough()` carried the key through parsing and no
* reader ever looked at it. The restore therefore dropped every water and
* vitamin total on the floor while reporting success.
*/
const nutrientDaySchema = z
.object({
day: z.string().min(1),
nutrient: z.string().min(1),
amount: z.number(),
unit: z.string().min(1),
// Closed set in app code rather than a DB enum; kept as a string here so a
// future source does not make an old file unparseable.
source: z.string().min(1),
})
.passthrough();
/**
* A symptom the account created, as opposed to the seeded catalogue.
*
* Carried because a day-log's symptom link resolves by key: without the
* definition the key resolves to nothing on the restoring instance, and the
* link is lost. The seeded rows are deliberately NOT carried — every instance
* already has them, and shipping them would mean a restore could rewrite
* another instance's reference data.
*/
const customCycleSymptomSchema = z
.object({
id: z.string().min(1).optional(),
key: z.string().min(1),
labelKey: z.string().min(1),
categoryId: z.string().min(1),
icon: z.string().nullable().optional(),
sortOrder: z.number().int().default(0),
isActive: z.boolean().default(true),
// Same reasoning as the mood tag's: the user's own words live here.
labelEncrypted: z.string().nullable().optional(),
})
.passthrough();
/**
* The account's durable self-context — one row per user.
*
* Portable exports carry the decrypted free text; disaster-recovery payloads
* carry the AES-256-GCM envelopes as base64 and leave the plaintext fields
* null. The restore prefers ciphertext when the file has it and re-encrypts
* the plaintext otherwise, which is the same contract the lab note uses.
*/
const healthProfileBackupSchema = z
.object({
id: z.string().min(1).optional(),
aboutMe: z.string().nullable().default(null),
conditions: z.string().nullable().default(null),
allergies: z.string().nullable().default(null),
coachFocus: z.string().nullable().default(null),
aiIncludedSections: z
.array(healthProfileAiSectionSchema)
.default([...DEFAULT_HEALTH_PROFILE_AI_SECTIONS]),
// Emergency profile: three plaintext enums (carried by value in both
// purposes), three free-text columns following the ciphertext-or-plaintext
// split of the self-context fields above.
emergencyBloodType: emergencyBloodTypeSchema.nullable().default(null),
organDonorStatus: organDonorStatusSchema.nullable().default(null),
advanceDirectiveStatus: advanceDirectiveStatusSchema
.nullable()
.default(null),
emergencyContacts: z.string().nullable().default(null),
emergencyImplants: z.string().nullable().default(null),
emergencyNote: z.string().nullable().default(null),
emergencyContactsEncrypted: base64BytesSchema.nullable().optional(),
emergencyImplantsEncrypted: base64BytesSchema.nullable().optional(),
emergencyNoteEncrypted: base64BytesSchema.nullable().optional(),
aboutMeEncrypted: base64BytesSchema.nullable().optional(),
conditionsEncrypted: base64BytesSchema.nullable().optional(),
allergiesEncrypted: base64BytesSchema.nullable().optional(),
coachFocusEncrypted: base64BytesSchema.nullable().optional(),
pendingQuestionsEncrypted: base64BytesSchema.nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const healthProfileFactBackupSchema = z
.object({
id: z.string().min(1),
kind: healthProfileFactKindSchema,
value: z.string().nullable().default(null),
valueEncrypted: base64BytesSchema.optional(),
validFrom: isoDateTime,
validUntil: isoDateTime.nullable(),
provenance: z.enum(["USER_REPORTED", "USER_CORRECTION"]),
supersededByRevisionId: z.string().min(1).nullable(),
createdAt: isoDateTime,
})
.passthrough()
.superRefine((fact, ctx) => {
if (fact.valueEncrypted === undefined) {
if (
fact.value === null ||
!isHealthProfileFactValue(fact.kind, fact.value)
) {
ctx.addIssue({
code: "custom",
path: ["value"],
message: "A portable profile fact requires a valid readable value",
});
}
}
if (
fact.validUntil !== null &&
Date.parse(fact.validUntil) <= Date.parse(fact.validFrom)
) {
ctx.addIssue({
code: "custom",
path: ["validUntil"],
message: "validUntil must be later than validFrom",
});
}
});
/** One reading of a user-defined metric, nested under the metric that owns it. */
const customMetricEntryBackupSchema = z
.object({
id: z.string().min(1).optional(),
value: z.number(),
unit: z.string().min(1),
measuredAt: isoDateTime,
note: z.string().nullable().default(null),
createdAt: isoDateTime.optional(),
})
.passthrough();
/**
* A metric the account defined itself.
*
* The one class of series no integration can ever re-sync: nobody else has the
* definition and nobody else has the readings. It was classified as carried
* from the day the backup plan was written and carried by nothing, so a
* restore rebuilt the account without it and reported success.
*
* Readings are nested rather than flat-with-a-name-reference so the restore
* has no parent to look up and therefore no lookup to miss.
*/
const customMetricBackupSchema = z
.object({
id: z.string().min(1).optional(),
name: z.string().min(1),
unit: z.string().min(1),
targetLow: z.number().nullable().default(null),
targetHigh: z.number().nullable().default(null),
decimals: z.number().int().nullable().default(null),
description: z.string().nullable().default(null),
correlationEnabled: z.boolean().default(false),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
deletedAt: isoDateTime.nullable().optional(),
entries: z.array(customMetricEntryBackupSchema).default([]),
})
.passthrough();
const correlationPatternBackupSchema = z
.object({
id: z.string().min(1).optional(),
canonicalKey: z.string().regex(/^p1:[a-f0-9]{64}$/),
family: z.string().min(1),
factorKey: z.string().min(1),
outcomeKey: z.string().min(1),
lagDays: z.number().int().min(0),
sampleSize: z.number().int().positive(),
effectSize: z.number().finite(),
pValue: z.number().min(0).max(1),
qValue: z.number().min(0).max(1).nullable().default(null),
evidenceHash: z.string().regex(/^[a-f0-9]{64}$/),
isCurrent: z.boolean(),
lastComputedAt: isoDateTime,
dismissedAt: isoDateTime.nullable().default(null),
dismissedEvidenceHash: z
.string()
.regex(/^[a-f0-9]{64}$/)
.nullable()
.default(null),
dismissedEffectSize: z.number().finite().nullable().default(null),
dismissedSampleSize: z.number().int().positive().nullable().default(null),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
/**
* One local day's cumulative curve for one metric.
*
* The drain writes this row and deletes the per-sample rows it was folded from
* in the same transaction, so past the grace window the file is the only copy.
* `hourlyCumulative` is left unconstrained in length here on purpose: the
* restore checks it against the slot count it shares with the reader and
* throws naming the day, which says what is wrong with one row instead of
* making a whole file unparseable.
*/
const intradayProfileBackupSchema = z
.object({
id: z.string().min(1).optional(),
type: z.enum(MeasurementType),
dateKey: z.string().regex(/^\d{4}-\d{2}-\d{2}$/),
hourlyCumulative: z.array(z.number()),
dayTotal: z.number(),
sampleCount: z.number().int(),
// The zone the day was cut on. The reader compares it against the
// account's current zone and drops the mismatched days, so a restore that
// dropped this column would make old curves look comparable when their
// hours mean something else.
timezone: z.string().min(1),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const healthScoreRecordBackupSchema = z
.object({
id: z.string().min(1).optional(),
dayKey: z.string().regex(/^\d{4}-\d{2}-\d{2}$/),
// The zone the day was cut on. An account that moved zones has days of
// different lengths behind it, and a restore that guessed would make them
// look comparable.
timezone: z.string().min(1),
composite: z.number().int().min(0).max(100),
// The same closed set the column's CHECK constraint carries. A band the
// schema let through and the database refused would fail the whole
// restore transaction on one bad row.
band: z.enum(["green", "yellow", "red"]),
scoreVersion: z.number().int(),
composition: z.array(z.string().min(1)).min(1),
pillarScores: z.record(z.string(), z.number()),
inputFingerprint: z.string().regex(/^[0-9a-f]{64}$/),
configVersion: z.number().int().nullable().default(null),
configChangedAt: isoDateTime.nullable().default(null),
computedAt: isoDateTime,
createdAt: isoDateTime.optional(),
})
.passthrough();
const appSettingsBackupSchema = z
.object({
id: z.string().min(1),
registrationEnabled: z.boolean(),
mfaRequired: z.boolean(),
defaultLocale: z.string(),
telegramGlobal: z.boolean(),
ntfyGlobal: z.boolean(),
webPushGlobal: z.boolean(),
webPushVapidPublicKey: z.string().nullable(),
webPushVapidPrivateKeyEncrypted: z.string().nullable(),
webPushVapidSubject: z.string().nullable(),
apiGlobal: z.boolean(),
umamiEnabled: z.boolean(),
umamiScriptUrl: z.string().nullable(),
umamiWebsiteId: z.string().nullable(),
glitchtipEnabled: z.boolean(),
glitchtipDsn: z.string().nullable(),
glitchtipEnvironment: z.string().nullable(),
reminderLateMinutes: z.number().int(),
reminderMissedMinutes: z.number().int(),
adminAiKeyEncrypted: z.string().nullable(),
adminAiModel: z.string(),
adminAiBaseUrl: z.string(),
adminCodexAccessTokenEncrypted: z.string().nullable(),
adminCodexRefreshTokenEncrypted: z.string().nullable(),
adminCodexAccountIdEncrypted: z.string().nullable(),
adminCodexTokenExpiresAt: isoDateTime.nullable(),
adminCodexConnectedAt: isoDateTime.nullable(),
adminCodexConnectionStatus: z.string(),
adminAiInsightsFeedbackSummary: z.unknown().nullable(),
defaultUserTimezone: z.string().nullable(),
assistantEnabled: z.boolean(),
assistantCoachEnabled: z.boolean(),
assistantBriefingEnabled: z.boolean(),
assistantInsightStatusEnabled: z.boolean(),
assistantCorrelationsEnabled: z.boolean(),
moduleAvailabilityJson: z.unknown().nullable(),
documentMaxFileBytes: z.number().int(),
documentQuotaBytes: z.string().regex(/^\d+$/),
})
.passthrough();
/* ── Structured-record disaster-recovery shapes ─────────────────────
*
* These shapes serve both the historical portable export and the canonical
* weekly/off-host disaster-recovery payload. Portable document entries remain
* metadata-only. Canonical entries additionally carry encrypted BYTEA values
* as base64 plus the codec/hash fields required to recreate InboundDocument
* without decrypting or fabricating content.
*/
const labResultBackupSchema = z
.object({
id: z.string().min(1).optional(),
panel: z.string().nullable().optional(),
analyte: z.string().min(1),
value: z.number().nullable().optional(),
valueText: z.string().nullable().optional(),
unit: z.string().min(1),
referenceLow: z.number().nullable().optional(),
referenceHigh: z.number().nullable().optional(),
// Optional so a backup written before these columns existed still
// restores; absent reads as "this reading had no source window on file".
sourceReferenceLow: z.number().nullable().optional(),
sourceReferenceHigh: z.number().nullable().optional(),
sourceReferenceText: z.string().nullable().optional(),
takenAt: isoDateTime,
source: z.string().min(1),
biomarkerName: z.string().nullable().optional(),
biomarkerId: z.string().nullable().optional(),
note: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
deletedAt: isoDateTime.nullable().optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const biomarkerBackupSchema = z
.object({
id: z.string().min(1).optional(),
name: z.string().min(1),
unit: z.string().min(1),
lowerBound: z.number().nullable().optional(),
upperBound: z.number().nullable().optional(),
panel: z.string().nullable().optional(),
hidden: z.boolean().optional(),
context: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const illnessSymptomBackupSchema = z
.object({
key: z.string().min(1),
severity: z.number().int().nullable().optional(),
})
.passthrough();
const illnessDayLogBackupSchema = z
.object({
id: z.string().min(1).optional(),
episodeId: z.string().min(1).optional(),
date: z.string().min(1),
functionalImpact: z.number().int().nullable().optional(),
feverC: z.number().nullable().optional(),
symptoms: z.array(illnessSymptomBackupSchema).default([]),
note: z.string().nullable().optional(),
updatedAt: isoDateTime.optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
tz: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
deletedAt: isoDateTime.nullable().optional(),
})
.passthrough();
const illnessEpisodeBackupSchema = z
.object({
id: z.string().min(1),
label: z.string().min(1),
type: z.enum(IllnessType),
lifecycle: z.enum(IllnessLifecycle),
onsetAt: isoDateTime,
resolvedAt: isoDateTime.nullable().optional(),
// Self-referencing flare/exacerbation link, carried as the exported
// episode's own id — never resolved against another user's rows.
parentConditionId: z.string().nullable().optional(),
note: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
deletedAt: isoDateTime.nullable().optional(),
updatedAt: isoDateTime.optional(),
dayLogs: z.array(illnessDayLogBackupSchema).default([]),
})
.passthrough();
/**
* A practitioner, an encounter, and the edges between an encounter and the
* things it produced.
*
* `id` is required on both records, unlike the lab result above: a link
* addresses an encounter and its far side by id and nothing else on either row
* is unique enough to rebuild the reference from. The ciphertext columns ride
* verbatim as base64 — a visit note is never decrypted into the file.
*/
const practitionerBackupSchema = z
.object({
id: z.string().min(1),
name: z.string().min(1),
specialty: z.string().nullable().optional(),
practice: z.string().nullable().optional(),
location: z.string().nullable().optional(),
phone: z.string().nullable().optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
createdAt: isoDateTime,
updatedAt: isoDateTime,
deletedAt: isoDateTime.nullable().optional(),
})
.passthrough();
const encounterBackupSchema = z
.object({
id: z.string().min(1),
occurredAt: isoDateTime,
status: z.enum(EncounterStatus),
kind: z.enum(EncounterKind),
practitionerId: z.string().nullable().optional(),
reasonEncrypted: base64BytesSchema.nullable().optional(),
outcomeEncrypted: base64BytesSchema.nullable().optional(),
// Carried so the loss is visible in the file. The restore resolves it to
// NULL and names the drop: the reminder it points at does not travel.
reminderId: z.string().nullable().optional(),
createdAt: isoDateTime,
updatedAt: isoDateTime,
deletedAt: isoDateTime.nullable().optional(),
})
.passthrough();
const encounterLinkBackupSchema = z
.object({
encounterId: z.string().min(1),
targetId: z.string().min(1),
createdAt: isoDateTime,
})
.passthrough();
/**
* One administered dose.
*
* `antigenSlug` is a free string here rather than an enum of the catalogue
* this release ships, deliberately: a slug the catalogue has since dropped
* must still restore, and the renderer degrades to `vaccineName`. Validating
* it at the restore boundary would refuse a file the app itself wrote.
*/
const vaccinationBackupSchema = z
.object({
id: z.string().min(1),
occurredAt: isoDateTime,
antigenSlug: z.string().nullable().optional(),
vaccineName: z.string().nullable().optional(),
doseNumber: z.number().int().nullable().optional(),
seriesDoses: z.number().int().nullable().optional(),
lotNumber: z.string().nullable().optional(),
site: z.enum(VaccinationSite).nullable().optional(),
practitionerId: z.string().nullable().optional(),
encounterId: z.string().nullable().optional(),
// Carried so the loss is visible in the file. The restore resolves it to
// NULL and names the drop: the reminder it points at does not travel.
reminderId: z.string().nullable().optional(),
noteEncrypted: base64BytesSchema.nullable().optional(),
createdAt: isoDateTime,
updatedAt: isoDateTime,
deletedAt: isoDateTime.nullable().optional(),
})
.passthrough();
const vaccinationLinkBackupSchema = z
.object({
vaccinationId: z.string().min(1),
targetId: z.string().min(1),
createdAt: isoDateTime,
})
.passthrough();
const allergyBackupSchema = z
.object({
id: z.string().min(1),
substance: z.string().min(1),
category: z.enum(AllergyCategory),
type: z.enum(AllergyType),
severity: z.enum(AllergySeverity).nullable().optional(),
status: z.enum(AllergyStatus),
onsetAt: isoDateTime.nullable().optional(),
reaction: z.string().nullable().optional(),
note: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
reactionEncrypted: base64BytesSchema.nullable().optional(),
notesEncrypted: base64BytesSchema.nullable().optional(),
deletedAt: isoDateTime.nullable().optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const familyHistoryBackupSchema = z
.object({
id: z.string().min(1),
relationship: z.enum(FamilyRelationship),
condition: z.string().min(1),
ageAtOnset: z.number().int().nullable().optional(),
note: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const workoutBackupSchema = z
.object({
id: z.string().min(1).optional(),
sportType: z.string().min(1),
startedAt: isoDateTime,
endedAt: isoDateTime,
durationSec: z.number().int(),
totalEnergyKcal: z.number().nullable().optional(),
totalDistanceM: z.number().nullable().optional(),
avgHeartRate: z.number().int().nullable().optional(),
maxHeartRate: z.number().int().nullable().optional(),
minHeartRate: z.number().int().nullable().optional(),
stepCount: z.number().int().nullable().optional(),
elevationM: z.number().nullable().optional(),
pauseDurationSec: z.number().int().nullable().optional(),
source: z.enum(MeasurementSource),
externalId: z.string().nullable().optional(),
metadata: z.json().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const documentBackupSchema = z
.object({
id: z.string().min(1),
kind: z.enum(InboundDocumentKind),
title: z.string().nullable().optional(),
filename: z.string().nullable().optional(),
mimeType: z.string().min(1),
byteSize: z.number().int(),
status: z.enum(InboundDocumentStatus),
reportDate: z.string().nullable().optional(),
documentDate: z.string().nullable().optional(),
contentEncrypted: base64BytesSchema.optional(),
contentSha256: z
.string()
.regex(/^[a-f0-9]{64}$/)
.nullable()
.optional(),
contentCodec: z.string().min(1).optional(),
providerType: z.string().nullable().optional(),
errorReason: z.string().nullable().optional(),
summaryEncrypted: base64BytesSchema.nullable().optional(),
summaryGeneratedAt: isoDateTime.nullable().optional(),
summaryState: z.enum(DocumentSummaryState).optional(),
// Refs #776 — the index-attempt record (canonical DR payloads only).
// Free string on purpose: a future outcome value must not fail a restore.
lastIndexAttemptAt: isoDateTime.nullable().optional(),
lastIndexOutcome: z.string().nullable().optional(),
summary: z.string().nullable().optional(),
createdAt: isoDateTime.optional(),
updatedAt: isoDateTime.optional(),
})
.passthrough();
const backupManifestSchema = z
.object({
documents: z
.object({ included: z.string().min(1), note: z.string().min(1) })
.passthrough(),
workouts: z
.object({ included: z.string().min(1), note: z.string().min(1) })
.passthrough(),
})
.passthrough();
/**
* Wire shape — exactly what the pg-boss worker writes today, plus a
* `schemaVersion` field that newer writers stamp explicitly. Older blobs
* without the field default to v1 in `parseBackupPayload`.
*/
export const backupPayloadSchema = z
.object({
schemaVersion: z.string().min(1).default(LEGACY_BACKUP_SCHEMA_VERSION),
exportedAt: isoDateTime,
userId: z.string().min(1),
appSettings: appSettingsBackupSchema.nullable().default(null),
measurements: z.array(measurementSchema).default([]),
medications: z.array(medicationSchema).default([]),
intakeEvents: z.array(intakeEventSchema).default([]),
moodEntries: z.array(moodEntrySchema).default([]),
// v1.15.0 — cycle-tracking tables. Default to empty arrays / null so a
// pre-v1.15 backup (no cycle keys) still round-trips unchanged.
cycleProfile: cycleProfileSchema.nullable().default(null),
cycles: z.array(cycleSpanSchema).default([]),
cycleDayLogs: z.array(cycleDayLogSchema).default([]),
// Defaulted rather than required: files written before this field existed
// are still valid, and an account with no custom symptoms writes [].
customSymptoms: z.array(customCycleSymptomSchema).default([]),
customMoodTags: z.array(customMoodTagSchema).default([]),
// Structured records default to empty arrays so older backups remain
// parseable. Canonical DR writers add stable ids and encrypted document
// fields; portable exports retain the metadata-only subset.
labResults: z.array(labResultBackupSchema).default([]),
biomarkers: z.array(biomarkerBackupSchema).default([]),
illnessEpisodes: z.array(illnessEpisodeBackupSchema).default([]),
allergies: z.array(allergyBackupSchema).default([]),
familyHistory: z.array(familyHistoryBackupSchema).default([]),
workouts: z.array(workoutBackupSchema).default([]),
documents: z.array(documentBackupSchema).default([]),
nutrientDays: z.array(nutrientDaySchema).default([]),
// Durable self-context and user-defined series. Defaulted so files written
// before either rode the wire still parse; an account with neither writes
// `null` / `[]`.
healthProfile: healthProfileBackupSchema.nullable().default(null),
healthProfileFacts: z.array(healthProfileFactBackupSchema).default([]),
customMetrics: z.array(customMetricBackupSchema).default([]),
correlationPatterns: z.array(correlationPatternBackupSchema).default([]),
// The hourly shape of a cumulative day. Defaulted for the same reason as
// the pair above: a file written before the table existed carries no key.
intradayProfiles: z.array(intradayProfileBackupSchema).default([]),
// The score as it was shown, day by day. Defaulted for the same reason as
// the sections above: a file written before the table existed carries no
// key, and an account whose score never resolved writes [].
healthScoreRecords: z.array(healthScoreRecordBackupSchema).default([]),
// Visits, the address book behind them, and the three link tables.
// Defaulted for the same reason as the sections above: a file written
// before the tables existed carries no key, and an account that has never
// filed a visit writes [].
practitioners: z.array(practitionerBackupSchema).default([]),
encounters: z.array(encounterBackupSchema).default([]),
encounterDocumentLinks: z.array(encounterLinkBackupSchema).default([]),
encounterLabLinks: z.array(encounterLinkBackupSchema).default([]),
encounterConditionLinks: z.array(encounterLinkBackupSchema).default([]),
// The immunization log and the pages it was transcribed from. Defaulted
// for the same reason as the sections above: a file written before the
// tables existed carries no key, and an account with an empty Impfpass
// writes [].
vaccinations: z.array(vaccinationBackupSchema).default([]),
vaccinationDocumentLinks: z.array(vaccinationLinkBackupSchema).default([]),
manifest: backupManifestSchema.nullable().default(null),
})
.passthrough()
.superRefine((payload, ctx) => {
if (payload.schemaVersion !== BACKUP_SCHEMA_VERSION) return;
payload.measurements.forEach((measurement, index) => {
if (!measurement.id) {
ctx.addIssue({
code: "custom",
path: ["measurements", index, "id"],
message: "Canonical v2 measurements require a stable id",
});
}
});
});
export type BackupPayload = z.infer<typeof backupPayloadSchema>;
/**
* Numeric counts of each backed-up record kind. Returned in the
* upload + restore API responses so the admin sees what they
* uploaded/restored without having to download the file again.
*/
export interface BackupSummary {
schemaVersion: string;
userId: string;
exportedAt: string;
measurements: number;
medications: number;
intakeEvents: number;
/** Side effects recorded against a drug, across every medication. */
medicationSideEffects: number;
moodEntries: number;
/** v1.15.0 — observed cycle spans in the backup. */
cycles: number;
/** v1.15.0 — cycle day-logs in the backup. */
cycleDayLogs: number;
/** Lab results in the backup. */
labResults: number;
nutrientDays: number;
/** User-scoped biomarker catalog entries in the backup. */
biomarkers: number;
/** Illness episodes, including flares/exacerbations. */
illnessEpisodes: number;
/** Illness day-logs across every episode. */
illnessDayLogs: number;
/** Allergy/intolerance records in the backup. */
allergies: number;
/** Family-history entries in the backup. */
familyHistory: number;
/** Workout summary records in the backup. */
workouts: number;
/** Document records (ciphertext included in canonical DR payloads). */
documents: number;
/** 1 when the account's durable self-context rides the file, 0 otherwise. */
healthProfile: number;
/** Effective-dated structured health-profile revisions. */
healthProfileFactRevisions: number;
/** Metrics the account defined itself. */
customMetrics: number;
/** Readings across every user-defined metric. */
customMetricEntries: number;
/** Persisted accepted correlation identities and dismissal decisions. */
correlationPatterns: number;
/** Stored day curves for the cumulative metrics, across every metric. */
intradayProfiles: number;
/** Local days whose health score was written down as it was shown. */
healthScoreRecords: number;
}
export function summarizeBackup(payload: BackupPayload): BackupSummary {
return {
schemaVersion: payload.schemaVersion,
userId: payload.userId,
exportedAt: payload.exportedAt,
measurements: payload.measurements.length,
medications: payload.medications.length,
intakeEvents: payload.intakeEvents.length,
medicationSideEffects: payload.medications.reduce(
(sum, medication) => sum + medication.sideEffects.length,
0,
),
moodEntries: payload.moodEntries.length,
cycles: payload.cycles.length,
cycleDayLogs: payload.cycleDayLogs.length,
nutrientDays: payload.nutrientDays.length,
labResults: payload.labResults.length,
biomarkers: payload.biomarkers.length,
illnessEpisodes: payload.illnessEpisodes.length,
illnessDayLogs: payload.illnessEpisodes.reduce(
(sum, e) => sum + e.dayLogs.length,
0,
),
allergies: payload.allergies.length,
familyHistory: payload.familyHistory.length,
workouts: payload.workouts.length,
documents: payload.documents.length,
healthProfile: payload.healthProfile ? 1 : 0,
healthProfileFactRevisions: payload.healthProfileFacts.length,
customMetrics: payload.customMetrics.length,
customMetricEntries: payload.customMetrics.reduce(
(sum, metric) => sum + metric.entries.length,
0,
),
correlationPatterns: payload.correlationPatterns.length,
intradayProfiles: payload.intradayProfiles.length,
healthScoreRecords: payload.healthScoreRecords.length,
};
}
/**
* Parse a JSON blob (string or already-parsed object) against
* `backupPayloadSchema`, returning a typed payload. Throws ZodError on
* mismatch — the admin route catches it and turns it into a 422 with a
* field-level error list.
*
* Accepts both forms because:
* - the upload route hands us the parsed object after `await req.text()`
* - the restore route reads `DataBackup.data` (decrypted) which is
* always a JSON string straight from the worker.
*/
export function parseBackupPayload(input: string | unknown): BackupPayload {
const raw = typeof input === "string" ? JSON.parse(input) : input;
return backupPayloadSchema.parse(raw);
}
/**
* The schemaVersion the system understands today. Used by the upload
* route to reject inbound files written by a *future* HealthLog instance
* — restoring them under the current code might silently drop data the
* new shape carried.
*/
export function isCompatibleSchemaVersion(version: string): boolean {
return (
version === LEGACY_BACKUP_SCHEMA_VERSION ||
version === BACKUP_SCHEMA_VERSION
);
}