stats: add hourly views — continuous timeline + hour-of-day pattern

Closes #45.

Two new server-side aggregates in computeStats(), same JS-bucketing
approach as the existing by_day: by_hour (localHourBucket — literal clock
time, no business-day rollover, feeds a left-to-right timeline) and
by_hour_of_day (localHourOfDay — every day's hour 0-23 summed together,
zero-filled to a full 24-entry axis so a quiet hour reads as zero, not a
missing data point).

Client: two more revenue charts on /stats — 'Umsatz nach Stunde' (timeline)
and 'Umsatz nach Tageszeit' (pattern). Revenue only, not also a tx-count
variant, to keep the page from growing a chart per metric per granularity.

Build clean both sides. Manually verified by_hour_of_day returns all 24
zero-filled buckets against a fresh DB.

server/src/time.ts

@@ -79,6 +79,27 @@ export function formatLocal(d: Date, tz: string = TZ): string {
   return `${p.year}-${pad(p.month)}-${pad(p.day)} ${pad(p.hour)}:${pad(p.minute)}:${pad(p.second)}`;
 }
 
+/**
+ * Local wall-clock hour bucket as `YYYY-MM-DD HH:00`, for the continuous
+ * hourly timeline in /stats (#45) — literal clock time, no business-day
+ * rollover (unlike businessDay() below): the point of this bucket is "what
+ * hour did this actually happen", not which sales night it counts toward.
+ */
+export function localHourBucket(d: Date, tz: string = TZ): string {
+  const p = localParts(d, tz);
+  const pad = (n: number) => String(n).padStart(2, '0');
+  return `${p.year}-${pad(p.month)}-${pad(p.day)} ${pad(p.hour)}:00`;
+}
+
+/**
+ * Local hour-of-day (0-23), for the "which hour is busiest" pattern
+ * summed across every day in /stats (#45) — deliberately not business-day
+ * aware either, same reasoning as localHourBucket().
+ */
+export function localHourOfDay(d: Date, tz: string = TZ): number {
+  return localParts(d, tz).hour;
+}
+
 /**
  * The business day a timestamp belongs to, as `YYYY-MM-DD`.
  * Hours before BUSINESS_DAY_CUTOFF_HOUR are attributed to the previous day.