Files
HealthLog/src/lib/measurements/create-from-telegram.ts
T
Marc-André Bombeck 9c8fb8c7c8 feat(telegram): unique chat binding and numeric measurement capture
Two related changes to the Telegram bot's data path.

One chat ⇄ one account. `User.telegramChatId` is now UNIQUE (migration
0189, defensive pre-clean of any pre-existing duplicate binding). The
inbound webhook resolves the user from the chat id, so a chat shared
across two accounts would route a reply or button tap ambiguously. The
index is over a nullable column, so an unlinked account (null chat id)
never collides — only two accounts binding the same non-null chat id are
rejected.

Numeric measurement capture. A numeric reply to a measurement reminder is
now captured as a Measurement (`source = TELEGRAM`, added to the enum in
0189). On "Erledigt" for a reminder that names a single-value metric, the
bot offers an optional value prompt via force_reply; a numeric reply is
written through the shared `logTelegramMeasurement` helper — same range
guard, canonical unit, cache + rollup refresh as the manual route. The
reply binds to the reminder and user through a TelegramPromptContext keyed
on the prompt message id (the same strict chat/message binding the
mood-note path uses); the userId is never read from the payload. Comma
decimals are tolerated; non-numeric and out-of-range replies get a calm
hint and write nothing; the reply self-cleans on the ~30-min sweep. Blood
pressure (two values) is out of scope for the single-numeric path and
keeps its satisfy-only flow.
2026-06-21 21:06:42 +02:00

165 lines
5.7 KiB
TypeScript

/**
* v1.19.2 — server-authoritative measurement capture from the Telegram bot.
*
* The HTTP measurement-create path (`POST /api/measurements`) is bound to a
* cookie/Bearer session via `requireAuth`; the Telegram webhook resolves the
* user from the linked chat instead. This helper holds the shared write so
* the two entry points agree on the canonical shape: the same range guard,
* the same canonical unit, the same cache invalidation + DAY-rollup
* recompute the manual route fires. `source` is pinned to the `TELEGRAM`
* `MeasurementSource` value (added in migration 0189).
*
* The `userId` is NEVER taken from the Telegram payload — the caller passes
* the id it resolved from the stored `telegramChatId` binding. The capture
* is correlated to a specific `MeasurementReminder` (carrying the expected
* `measurementType`) through a `TelegramPromptContext` row, so a numeric
* reply can only ever land the metric that reminder asked for, for that
* user. BP (two values) is out of scope for the single-numeric-reply path.
*/
import type { Prisma, PrismaClient } from "@/generated/prisma/client";
import type { MeasurementType } from "@/generated/prisma/client";
import { prisma as defaultPrisma } from "@/lib/db";
import {
getUnitForType,
validateMeasurementRange,
} from "@/lib/validations/measurement";
import { invalidateUserMeasurements } from "@/lib/cache/invalidate";
import { recomputeBucketsForMeasurement } from "@/lib/rollups/measurement-rollups";
import { getEvent } from "@/lib/logging/context";
/**
* The single-value measurement types a numeric Telegram reply can capture.
*
* Excludes BLOOD_PRESSURE_SYS — a BP reading needs systolic AND diastolic,
* which a single numeric reply can't express unambiguously. The BP reminder
* keeps its existing satisfy-only "done" flow; a value capture for it is a
* documented follow-up (a guided two-number prompt).
*/
const TELEGRAM_CAPTURABLE_TYPES: ReadonlySet<MeasurementType> =
new Set<MeasurementType>([
"WEIGHT",
"PULSE",
"BLOOD_GLUCOSE",
"OXYGEN_SATURATION",
"BODY_TEMPERATURE",
"BODY_FAT",
"FAT_MASS",
"FAT_FREE_MASS",
"MUSCLE_MASS",
"LEAN_BODY_MASS",
"BONE_MASS",
"TOTAL_BODY_WATER",
"VISCERAL_FAT",
"BODY_MASS_INDEX",
]);
/**
* True when a numeric Telegram reply can capture a Measurement for this
* reminder's target type. The webhook gates the value-prompt on this so a
* BP reminder (or a free-text Vorsorge with no `measurementType`) never
* offers an ambiguous numeric capture.
*/
export function isTelegramCapturableType(
type: MeasurementType | null | undefined,
): type is MeasurementType {
return type != null && TELEGRAM_CAPTURABLE_TYPES.has(type);
}
export interface TelegramMeasurementResult {
/** `ok` — a row was written (or already existed under the dedup key).
* `invalid_number` — the reply did not parse to a finite number.
* `out_of_range` — the value fell outside the type's plausible range.
* `unsupported_type` — the type is not single-value capturable. */
status: "ok" | "invalid_number" | "out_of_range" | "unsupported_type";
}
/**
* Parse a free-text Telegram reply into a finite number. Tolerates a
* comma decimal separator (de-DE keyboards) and surrounding whitespace;
* rejects anything with trailing non-numeric noise so "72 bpm" doesn't
* silently capture 72.
*/
export function parseTelegramNumber(text: string): number | null {
const normalised = text.trim().replace(",", ".");
if (!/^-?\d+(\.\d+)?$/.test(normalised)) return null;
const n = Number(normalised);
return Number.isFinite(n) ? n : null;
}
/**
* Capture a measurement on behalf of a Telegram-linked user from a numeric
* reply to a measurement-reminder prompt.
*
* Idempotent on `externalId` (the per-prompt key the caller derives from
* `telegram:measure:<chatId>:<promptMsgId>`): a redelivered reply converges
* onto the same row via the `(userId, type, source, externalId)` unique
* instead of minting a duplicate.
*/
export async function logTelegramMeasurement(input: {
userId: string;
type: MeasurementType;
rawText: string;
tz: string | null;
/** Stable per-prompt id for Telegram redelivery dedup. */
externalId: string;
client?: PrismaClient | Prisma.TransactionClient;
}): Promise<TelegramMeasurementResult> {
const prisma = (input.client ?? defaultPrisma) as PrismaClient;
if (!isTelegramCapturableType(input.type)) {
return { status: "unsupported_type" };
}
const value = parseTelegramNumber(input.rawText);
if (value === null) {
return { status: "invalid_number" };
}
if (validateMeasurementRange(input.type, value) !== null) {
return { status: "out_of_range" };
}
const measuredAt = new Date();
const existing = await prisma.measurement.findUnique({
where: {
userId_type_source_externalId: {
userId: input.userId,
type: input.type,
source: "TELEGRAM",
externalId: input.externalId,
},
},
select: { id: true },
});
if (existing) {
return { status: "ok" };
}
await prisma.measurement.create({
data: {
userId: input.userId,
type: input.type,
value,
unit: getUnitForType(input.type),
source: "TELEGRAM",
measuredAt,
externalId: input.externalId,
},
});
invalidateUserMeasurements(input.userId, { evict: true });
// Best-effort rollup refresh — a cache tier, never a write-path invariant.
try {
await recomputeBucketsForMeasurement(input.userId, input.type, measuredAt);
} catch (rollupErr) {
getEvent()?.addMeta(
"telegram_measurement_rollup_failed",
rollupErr instanceof Error ? rollupErr.message : String(rollupErr),
);
}
return { status: "ok" };
}