fix(playlists): one rendering for a playlist total, honest below a minute (E9 R1)

Epic-end deep review, round 1. Six findings survived verification; five of them
were one defect wearing different hats.

formatRuntime floored 1-59s to "0m". Harmless while it only ever formatted a
Plex movie's length, but 12D repurposed it for an aggregate, and a total is the
only input that can land under a minute: a playlist of Shorts reported "0m"
while the rail beside it read "0:45". Fixed in the function, not behind a
`>= 60` guard at each of the seven call sites — the invariant belongs where it
cannot be forgotten. Its contract comment claimed "Plex movie/episode lengths"
and was two epics stale.

The rail rendered the same playlist total as formatDuration (1:23:45) while the
new header rendered formatRuntime (1h 23m) — two shapes of one number, side by
side on one screen. The rail now uses formatRuntime too; formatTotalHours had
already recorded that H:MM:SS is wrong for an aggregate.

The rename input kept min-w-0 without flex-1 in the compacted row, so it was
the only shrinkable child beside a shrink-0 action cluster. The Plex half of
this same epic already had the fix; this restores the symmetry.

plex.browse.items had no plural forms, so a one-item Plex playlist read
"1 items". Pre-existing, but the new header is the call site where count=1 is
ordinary. Both languages get both forms, no key drift.

formatRuntime had no unit tests at all across seven call sites. Added a tier
table plus a sweep asserting no non-zero second ever renders as "0m".

Dropped after 3/3 adversarial refutations: the meta-line segment ordering
(the filtered count is labelled, and count-then-runtime is the convention) and
the null-duration lower bound (the server's own SUM has skipped NULLs since
long before this branch; changing only the client would split them).

frontend/src/lib/format.test.ts

@@ -1,5 +1,5 @@
 import { afterEach, beforeEach, describe, expect, it, vi } from "vitest";
-import { RELATIVE_UNITS, relativeTime, relativeFromMs } from "./format";
+import { RELATIVE_UNITS, formatRuntime, relativeTime, relativeFromMs } from "./format";
 import en from "../i18n/locales/en/time.json";
 import hu from "../i18n/locales/hu/time.json";
 
@@ -108,3 +108,35 @@ describe("relativeFromMs", () => {
     expect(hu).toHaveProperty("justNow");
   });
 });
+
+// formatRuntime went from "one Plex movie's length" to "a playlist's total playtime" (E9 12D), and a
+// TOTAL is the only input that can legitimately be 1-59 seconds. That tier had no coverage and no
+// implementation: 45s floored to "0m", so a playlist of Shorts reported no runtime while the rail
+// beside it read "0:45". The table pins every tier boundary, so the sub-minute branch cannot be
+// dropped again as "dead code" — it is the one branch a whole-playlist figure reaches.
+describe("formatRuntime", () => {
+  const CASES: [number | null | undefined, string][] = [
+    [null, ""],
+    [undefined, ""],
+    [0, ""], // nothing to show — the call sites suppress their separator on the empty string
+    [1, "1s"],
+    [45, "45s"], // one Short: the regression case
+    [59, "59s"], // last second before the minute tier
+    [60, "1m"],
+    [90, "1m"], // floors, never rounds up — a total must not overstate
+    [3599, "59m"],
+    [3600, "1h 0m"], // the zero minutes stay: "1h" alone reads as an approximation
+    [5025, "1h 23m"],
+    [360000, "100h 0m"], // a large aggregate keeps counting hours rather than rolling to days
+  ];
+
+  it.each(CASES)("formats %o as %o", (sec, expected) => {
+    expect(formatRuntime(sec)).toBe(expected);
+  });
+
+  it("never reports a non-zero duration as zero", () => {
+    for (let sec = 1; sec < 3600; sec++) {
+      expect(formatRuntime(sec), `${sec}s`).not.toBe("0m");
+    }
+  });
+});