commit 3cb53fc773c769dcda00e05b3b2b4492cfd3a657
parent a9c364f637e3df63fa66dfab46f48edb49c214cb
Author: I Mean I'm Just Saying <imeanimjustsaying@kiwifarms.st>
Date: Tue, 6 Oct 2026 11:21:04 -0400
tracks: labels for tracks YouTube named, and en-<region>-orig
The corpus holds caption tracks YouTube names itself (en-<id>, an uploader's
extra track) and en-US-orig beside the plain ones. A region subtag is two
letters or three digits; anything else after `en-` is "other uploaded
captions", en-<region>-orig is "original audio captions (US)", and a switcher
whose two tracks would read the same adds the id (trackLabels).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Diffstat:
4 files changed, 49 insertions(+), 16 deletions(-)
diff --git a/common/components/TranscriptModal.tsx b/common/components/TranscriptModal.tsx
@@ -36,7 +36,7 @@ import { AgeRestrictedBadge, LivestreamBadge } from "./badges";
import { VirtualRow } from "./VirtualRow";
import { formatTimestamp } from "../lib/vtt";
import { formatDate } from "../lib/format";
-import { trackLabel } from "../lib/captionTracks";
+import { trackLabels } from "../lib/captionTracks";
type DisplayCue = {
start: number;
@@ -831,9 +831,9 @@ function TrackSwitcher({
onChange={(e) => onChange(e.target.value)}
className="bg-transparent text-zinc-200 rounded border border-zinc-700 px-1 py-0.5 text-xs focus:outline-none focus:ring-1 focus:ring-zinc-500"
>
- {tracks.map((t, i) => (
- <option key={t} value={t} className="bg-zinc-900">
- {trackLabel(t)}
+ {trackLabels(tracks).map((label, i) => (
+ <option key={tracks[i]} value={tracks[i]} className="bg-zinc-900">
+ {label}
{i === 0 ? " (default)" : ""}
</option>
))}
diff --git a/common/lib/captionTracks.test.ts b/common/lib/captionTracks.test.ts
@@ -17,6 +17,7 @@ import {
recordTracks,
trackKind,
trackLabel,
+ trackLabels,
uncoveredAltHits,
} from "./captionTracks";
import { readTrackFields, readVideoTracks } from "./captionTracks-server";
@@ -48,7 +49,16 @@ test("labels are plain words derived from the track id", () => {
assert.equal(trackLabel("en"), "uploaded captions");
assert.equal(trackLabel("en-en-US"), "auto-translated captions");
assert.equal(trackLabel("en-GB"), "UK English captions");
- assert.equal(trackLabel("en-x-foo"), "regional captions (en-x-foo)");
+ assert.equal(trackLabel("en-NG"), "regional captions (en-NG)");
+ assert.equal(trackLabel("en-US-orig"), "original audio captions (US)");
+ // A track YouTube named itself: another uploaded one.
+ assert.equal(trackLabel("en-uYU-mmqFLq8"), "other uploaded captions");
+ assert.equal(trackKind("en-JkeT_87f4cc"), "uploaded");
+ assert.deepEqual(trackLabels(["en-orig", "en-aa-1", "en-bb-2"]), [
+ "original audio captions",
+ "other uploaded captions (en-aa-1)",
+ "other uploaded captions (en-bb-2)",
+ ]);
assert.equal(trackLabel("transcription"), "transcription");
assert.equal(trackLabel("pinned"), "chosen captions");
assert.equal(inTrackLabel("en"), "in uploaded captions");
diff --git a/common/lib/captionTracks.ts b/common/lib/captionTracks.ts
@@ -62,24 +62,39 @@ export type TrackKind =
| "pinned"
| "other";
+// A region subtag as YouTube writes one: two letters (en-US) or three digits
+// (en-419). Anything else after `en-` is a track YouTube named itself — an
+// uploader's extra caption track (en-uYU-mmqFLq8).
+const REGION_RE = /^(?:[A-Za-z]{2}|\d{3})$/;
+
export function trackKind(track: string): TrackKind {
if (track === TRANSCRIPTION_TRACK) return "transcription";
if (track === PINNED_TRACK) return "pinned";
- if (track === "en-orig") return "original";
+ if (track === "en-orig" || /^en-[^-]+-orig$/.test(track)) return "original";
if (track === "en") return "uploaded";
if (/^en-en(?:-|$)/.test(track)) return "translated";
- if (/^en-/.test(track)) return "regional";
+ const m = track.match(/^en-(.+)$/);
+ if (m) return REGION_RE.test(m[1]) ? "regional" : "uploaded";
return "other";
}
+// "UK" for en-GB, the code itself for a region this table does not name.
+function regionName(code: string): string {
+ return REGIONS[code.toUpperCase()] ?? code;
+}
+
// A plain label for a track — what a person reads in the switcher and on a
// search hit ("in uploaded captions").
export function trackLabel(track: string): string {
switch (trackKind(track)) {
- case "original":
- return "original audio captions";
+ case "original": {
+ // en-US-orig: the original audio's captions, under a region.
+ const m = track.match(/^en-([^-]+)-orig$/);
+ return m ? `original audio captions (${regionName(m[1])})` : "original audio captions";
+ }
case "uploaded":
- return "uploaded captions";
+ // A track YouTube named (en-uYU-mmqFLq8) is another uploaded one.
+ return track === "en" ? "uploaded captions" : "other uploaded captions";
case "translated":
return "auto-translated captions";
case "transcription":
@@ -87,8 +102,7 @@ export function trackLabel(track: string): string {
case "pinned":
return "chosen captions";
case "regional": {
- const region = track.slice(3);
- const name = REGIONS[region.toUpperCase()];
+ const name = REGIONS[track.slice(3).toUpperCase()];
return name ? `${name} English captions` : `regional captions (${track})`;
}
default:
@@ -96,6 +110,15 @@ export function trackLabel(track: string): string {
}
}
+// The labels of a switcher's tracks, in order: trackLabel, with the id added
+// where two tracks would otherwise read the same (two tracks YouTube named).
+export function trackLabels(tracks: readonly string[]): string[] {
+ const labels = tracks.map(trackLabel);
+ return labels.map((l, i) =>
+ labels.indexOf(l) !== labels.lastIndexOf(l) ? `${l} (${tracks[i]})` : l,
+ );
+}
+
// The track id of a caption file name (transcript.<id>.vtt), or null.
export function trackOfVttFile(filename: string): string | null {
const m = filename.match(/^transcript\.([^.]+)\.vtt$/);
diff --git a/editor/app/channels/[slug]/videos/[id]/components/cards/TranscriptTracksReader.tsx b/editor/app/channels/[slug]/videos/[id]/components/cards/TranscriptTracksReader.tsx
@@ -7,7 +7,7 @@
import { useState, useTransition } from "react";
import type { AltTrack } from "yt-dlp-transcript-common/lib/captionTracks";
-import { trackLabel } from "yt-dlp-transcript-common/lib/captionTracks";
+import { trackLabels } from "yt-dlp-transcript-common/lib/captionTracks";
import { formatTimestamp } from "yt-dlp-transcript-common/lib/vtt";
import { readTranscriptTracksAction } from "../../videoActions";
@@ -69,9 +69,9 @@ export function TranscriptTracksReader({
onChange={(e) => setShown(e.target.value)}
className="rounded border border-border bg-background px-1 py-0.5 text-xs text-foreground"
>
- {tracks.map((t, i) => (
- <option key={t.track} value={t.track}>
- {trackLabel(t.track)}
+ {trackLabels(tracks.map((t) => t.track)).map((label, i) => (
+ <option key={tracks[i].track} value={tracks[i].track}>
+ {label}
{i === 0 ? " (default)" : ""}
</option>
))}
