Open-source AI podcast clipper. Generate vertical clips with face tracking and burned-in captions. CLI, MCP server, web app.
Python ai-toolsauto-captionscaptionscliffmpegmcpmcp-serveropusclip-alternativepodcast-clipspodcastsshortform-videovertical-videowhisper
// Port of remotion/src/chunks.ts, the chunker the renderer burns into the MP4.
// The studio's phone preview is only worth showing if it groups and holds words
// exactly like the render, so the two must be changed together.
export interface PreviewWord {
text: string;
start: number;
end: number;
speaker?: string | null;
}
export interface PreviewChunk {
words: PreviewWord[];
start: number;
end: number;
displayEnd: number;
}
export interface PreviewChunkOptions {
wordsPerChunk: number;
maxCharsPerChunk?: number;
absorbTail?: number;
splitTail?: boolean;
}
export interface TranscriptPreviewWord {
word?: string;
text?: string;
start: number;
end: number;
speaker?: string | null;
}
export interface PreviewClipBounds {
start_second: number;
end_second: number;
}
const BREAK_GAP_SECONDS = 0.8;
const GAP_FILL_MAX_SECONDS = 0.4;
const LAST_CHUNK_HOLD_SECONDS = 1.2;
const TERMINAL_PUNCTUATION = /[.!?…]["')\]]?$/;
export function selectPreviewWords(
transcriptWords: TranscriptPreviewWord[] | null | undefined,
activeClip?: PreviewClipBounds | null,
): PreviewWord[] {
if (!transcriptWords?.length) return [];
const words = activeClip
? transcriptWords.filter(
(word) => word.end > activeClip.start_second && word.start < activeClip.end_second,
)
: transcriptWords;
return words
.map((word) => ({
text: (word.word ?? word.text ?? "").trim(),
start: word.start,
end: word.end,
speaker: word.speaker,
}))
.filter((word) => word.text.length > 0);
}
function breaksAfter(prev: PreviewWord, next: PreviewWord): boolean {
return (
TERMINAL_PUNCTUATION.test(prev.text.trim()) ||
(prev.speaker != null && next.speaker != null && prev.speaker !== next.speaker) ||
next.start - prev.end > BREAK_GAP_SECONDS
);
}
function hasBreakBetween(words: PreviewWord[], from: number, to: number): boolean {
for (let i = from; i < to; i++) {
if (breaksAfter(words[i], words[i + 1])) return true;
}
return false;
}
export function buildPreviewChunks(
words: PreviewWord[],
cfg: PreviewChunkOptions,
clipEnd?: number,
): PreviewChunk[] {
const chunks: PreviewChunk[] = [];
let i = 0;
while (i < words.length) {
let end = i + 1;
let charCount = words[i].text.trim().length;
while (end < words.length && end - i < cfg.wordsPerChunk) {
if (breaksAfter(words[end - 1], words[end])) break;
if (cfg.maxCharsPerChunk != null) {
const nextChars = charCount + 1 + words[end].text.trim().length;
if (nextChars > cfg.maxCharsPerChunk) break;
charCount = nextChars;
}
end += 1;
}
if (
cfg.absorbTail != null &&
end < words.length &&
words.length - end <= cfg.absorbTail &&
!hasBreakBetween(words, end - 1, words.length - 1)
) {
end = words.length;
}
if (cfg.splitTail && words.length - end === 1 && end - i > 2) {
end -= 1;
}
const slice = words.slice(i, end);
const speechEnd = slice[slice.length - 1].end;
chunks.push({ words: slice, start: slice[0].start, end: speechEnd, displayEnd: speechEnd });
i = end;
}
for (let k = 0; k < chunks.length; k++) {
const next = chunks[k + 1];
const hold = next
? Math.min(next.start, chunks[k].end + GAP_FILL_MAX_SECONDS)
: chunks[k].end + LAST_CHUNK_HOLD_SECONDS;
const capped = clipEnd != null ? Math.min(hold, clipEnd) : hold;
chunks[k].displayEnd = Math.max(chunks[k].end, capped);
}
return chunks;
}
export function activePreviewChunk(
chunks: PreviewChunk[],
time: number,
): PreviewChunk | undefined {
let low = 0;
let high = chunks.length - 1;
while (low <= high) {
const middle = Math.floor((low + high) / 2);
const chunk = chunks[middle];
if (time < chunk.start) {
high = middle - 1;
} else if (time >= chunk.displayEnd) {
low = middle + 1;
} else {
return chunk;
}
}
return undefined;
}