@tienne/gestalt 0.101.1 → 0.103.0

package/dist/src/digest/format.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"format.d.ts","sourceRoot":"","sources":["../../../src/digest/format.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,aAAa,CAAC;AAE9C,8CAA8C;AAC9C,wBAAgB,UAAU,CAAC,GAAG,EAAE,MAAM,EAAE,SAAS,UAAQ,GAAG,MAAM,CAQjE;AAED,sCAAsC;AACtC,wBAAgB,UAAU,CAAC,IAAI,EAAE,UAAU,CAAC,MAAM,CAAC,EAAE,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAG/E;AAED,kDAAkD;AAClD,wBAAgB,cAAc,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,CAEpD;AAED,eAAO,MAAM,aAAa,EAAE,MAAM,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC,eAAe,CAAC,EAAE,MAAM,CAI7E,CAAC"}

package/dist/src/digest/format.js

@@ -0,0 +1,26 @@
+/** 1시간이 넘는 영상만 시를 붙인다. 스크립트 패널 폭을 아끼려는 것이다 */
+export function formatTime(sec, withHours = false) {
+    const s = Math.max(0, Math.floor(sec));
+    const h = Math.floor(s / 3600);
+    const m = Math.floor((s % 3600) / 60);
+    const r = String(s % 60).padStart(2, '0');
+    return h > 0 || withHours
+        ? `${h}:${String(m).padStart(2, '0')}:${r}`
+        : `${String(m).padStart(2, '0')}:${r}`;
+}
+/** 유튜브 영상이면 그 시각으로 여는 주소, 아니면 null */
+export function watchUrlAt(meta, sec) {
+    if (meta.source !== 'youtube' || !meta.videoId)
+        return null;
+    return `https://www.youtube.com/watch?v=${meta.videoId}&t=${Math.floor(Math.max(0, sec))}s`;
+}
+/** 작성자가 "1.2." 같은 번호를 붙여 와도 렌더러가 다시 매기므로 떼어 낸다 */
+export function stripNumbering(title) {
+    return title.replace(/^\s*\d+(?:\.\d+)*\.?\s+/, '');
+}
+export const CAPTION_LABEL = {
+    manual: '수동 자막',
+    auto: '자동 자막',
+    whisper: '음성 인식(whisper)',
+};
+//# sourceMappingURL=format.js.map

package/dist/src/digest/format.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"format.js","sourceRoot":"","sources":["../../../src/digest/format.ts"],"names":[],"mappings":"AAEA,8CAA8C;AAC9C,MAAM,UAAU,UAAU,CAAC,GAAW,EAAE,SAAS,GAAG,KAAK;IACvD,MAAM,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC;IACvC,MAAM,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,IAAI,CAAC,CAAC;IAC/B,MAAM,CAAC,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC,CAAC;IACtC,MAAM,CAAC,GAAG,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC;IAC1C,OAAO,CAAC,GAAG,CAAC,IAAI,SAAS;QACvB,CAAC,CAAC,GAAG,CAAC,IAAI,MAAM,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,IAAI,CAAC,EAAE;QAC3C,CAAC,CAAC,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC;AAC3C,CAAC;AAED,sCAAsC;AACtC,MAAM,UAAU,UAAU,CAAC,IAAwB,EAAE,GAAW;IAC9D,IAAI,IAAI,CAAC,MAAM,KAAK,SAAS,IAAI,CAAC,IAAI,CAAC,OAAO;QAAE,OAAO,IAAI,CAAC;IAC5D,OAAO,mCAAmC,IAAI,CAAC,OAAO,MAAM,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,GAAG,CAAC,CAAC,GAAG,CAAC;AAC9F,CAAC;AAED,kDAAkD;AAClD,MAAM,UAAU,cAAc,CAAC,KAAa;IAC1C,OAAO,KAAK,CAAC,OAAO,CAAC,yBAAyB,EAAE,EAAE,CAAC,CAAC;AACtD,CAAC;AAED,MAAM,CAAC,MAAM,aAAa,GAAwD;IAChF,MAAM,EAAE,OAAO;IACf,IAAI,EAAE,OAAO;IACb,OAAO,EAAE,gBAAgB;CAC1B,CAAC"}

package/dist/src/digest/frames.d.ts

@@ -0,0 +1,26 @@
+import { type Digest, type Transcript } from './schema.js';
+import { type RunCommand } from './ytdlp.js';
+export interface FramesOptions {
+    repoRoot: string;
+    /** 이미 뽑아 둔 프레임도 다시 뽑는다 */
+    force?: boolean;
+    run?: RunCommand;
+}
+export interface FramesResult {
+    framesDir: string;
+    extracted: number;
+    skipped: number;
+    total: number;
+    /** ffmpeg가 오류 없이 끝났는데 파일이 안 생긴 frameSec. 영상 끝을 넘긴 시각이 주로 여기 온다 */
+    missing: number[];
+}
+/** 프레임 파일 이름. 렌더러가 frameSec으로 같은 이름을 찾는다 */
+export declare function frameFileName(sec: number): string;
+export declare function collectFrameSecs(digest: Digest): number[];
+export declare function ensureVideo(run: RunCommand, transcript: Transcript, cacheDir: string, cwd: string): Promise<string>;
+/**
+ * digest.json의 소섹션 frameSec마다 한 장씩 뽑아 frames/에 둔다.
+ * 요약이 끝나야 시각이 정해져서 ingest와 따로 돈다.
+ */
+export declare function extractFrames(slug: string, opts: FramesOptions): Promise<FramesResult>;
+//# sourceMappingURL=frames.d.ts.map

package/dist/src/digest/frames.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"frames.d.ts","sourceRoot":"","sources":["../../../src/digest/frames.ts"],"names":[],"mappings":"AAIA,OAAO,EAAmB,KAAK,MAAM,EAAE,KAAK,UAAU,EAAE,MAAM,aAAa,CAAC;AAE5E,OAAO,EAAiC,KAAK,UAAU,EAAE,MAAM,YAAY,CAAC;AAE5E,MAAM,WAAW,aAAa;IAC5B,QAAQ,EAAE,MAAM,CAAC;IACjB,0BAA0B;IAC1B,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,GAAG,CAAC,EAAE,UAAU,CAAC;CAClB;AAED,MAAM,WAAW,YAAY;IAC3B,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,kEAAkE;IAClE,OAAO,EAAE,MAAM,EAAE,CAAC;CACnB;AAED,4CAA4C;AAC5C,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAEjD;AAED,wBAAgB,gBAAgB,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,CAKzD;AAkBD,wBAAsB,WAAW,CAC/B,GAAG,EAAE,UAAU,EACf,UAAU,EAAE,UAAU,EACtB,QAAQ,EAAE,MAAM,EAChB,GAAG,EAAE,MAAM,GACV,OAAO,CAAC,MAAM,CAAC,CAoBjB;AAED;;;GAGG;AACH,wBAAsB,aAAa,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,aAAa,GAAG,OAAO,CAAC,YAAY,CAAC,CAuD5F"}

package/dist/src/digest/frames.js

@@ -0,0 +1,109 @@
+import { existsSync, mkdirSync, mkdtempSync, readdirSync, rmSync, statSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+import { fileURLToPath } from 'node:url';
+import { youtubeWatchUrl } from './schema.js';
+import { DigestStore } from './store.js';
+import { runCommand, timeoutFor, ytdlp } from './ytdlp.js';
+/** 프레임 파일 이름. 렌더러가 frameSec으로 같은 이름을 찾는다 */
+export function frameFileName(sec) {
+    return `t${String(Math.round(sec * 1000)).padStart(8, '0')}.jpg`;
+}
+export function collectFrameSecs(digest) {
+    const secs = digest.reports.flatMap((r) => r.sections.flatMap((s) => s.subsections.map((sub) => sub.frameSec)));
+    return [...new Set(secs)].sort((a, b) => a - b);
+}
+/**
+ * 프레임만 뽑으면 되니 소리 없는 낮은 화질 영상을 받는다.
+ * 403이 나면 web_embedded로 다시 받는데, 그 클라이언트는 480p 이하 포맷이 빠지는 경우가 있어 136(720p)과 18(360p)까지 넓힌다.
+ */
+const FORMAT = 'bv*[height<=480][ext=mp4]/b[height<=480][ext=mp4]/18';
+const FORMAT_RETRY = '135/134/136/18';
+// yt-dlp가 중단되면 video.mp4.part나 video.mp4.ytdl 같은 보조 파일이 남는다. 완성본 확장자만 캐시로 본다
+const CACHED_VIDEO = /^video\.(mp4|webm|mkv)$/;
+function findCachedVideo(cacheDir) {
+    if (!existsSync(cacheDir))
+        return null;
+    const hit = readdirSync(cacheDir).find((f) => CACHED_VIDEO.test(f));
+    return hit ? join(cacheDir, hit) : null;
+}
+export async function ensureVideo(run, transcript, cacheDir, cwd) {
+    const { meta } = transcript;
+    if (meta.source === 'file') {
+        if (!meta.url)
+            throw new Error('transcript에 원본 영상 경로가 없어요.');
+        return fileURLToPath(meta.url);
+    }
+    const cached = findCachedVideo(cacheDir);
+    if (cached)
+        return cached;
+    if (!meta.videoId)
+        throw new Error('transcript에 videoId가 없어요.');
+    mkdirSync(cacheDir, { recursive: true });
+    // url 필드 대신 검증된 videoId로 주소를 다시 만들고 '--' 뒤에 둬서 옵션으로 읽히지 않게 한다
+    const out = ['-o', join(cacheDir, 'video.%(ext)s'), '--', youtubeWatchUrl(meta.videoId)];
+    await ytdlp(run, ['-f', FORMAT, ...out], ['-f', FORMAT_RETRY, ...out], {
+        cwd,
+        timeoutMs: timeoutFor(meta.durationSec, 1),
+    });
+    const got = findCachedVideo(cacheDir);
+    if (!got)
+        throw new Error('영상을 받지 못했어요.');
+    return got;
+}
+/**
+ * digest.json의 소섹션 frameSec마다 한 장씩 뽑아 frames/에 둔다.
+ * 요약이 끝나야 시각이 정해져서 ingest와 따로 돈다.
+ */
+export async function extractFrames(slug, opts) {
+    const run = opts.run ?? runCommand;
+    const store = new DigestStore(opts.repoRoot);
+    const t = store.readTranscript(slug);
+    if (!t.ok)
+        throw new Error(t.message);
+    const transcript = t.transcript;
+    const digest = store.loadDigest(slug);
+    if (!digest) {
+        throw new Error(`digest.json이 없거나 형식이 틀렸어요. 먼저 요약을 쓰고 validate를 통과시켜 주세요: ${slug}`);
+    }
+    const secs = collectFrameSecs(digest);
+    const framesDir = store.framesDir(slug);
+    mkdirSync(framesDir, { recursive: true });
+    const todo = secs.filter((sec) => {
+        const p = join(framesDir, frameFileName(sec));
+        return opts.force || !existsSync(p) || statSync(p).size === 0;
+    });
+    if (todo.length === 0) {
+        return { framesDir, extracted: 0, skipped: secs.length, total: secs.length, missing: [] };
+    }
+    // cacheDir는 레포 안이라 cwd로 쓰면 그 레포가 커밋한 yt-dlp.conf를 읽는다
+    const cwd = mkdtempSync(join(tmpdir(), 'gestalt-digest-run-'));
+    try {
+        const video = await ensureVideo(run, transcript, store.videoCacheDir(slug), cwd);
+        // 긴 영상일수록 뒤쪽 시각을 찾는 데 오래 걸려서 상한을 길이에 비례시킨다
+        const frameTimeout = timeoutFor(transcript.meta.durationSec, 0.05, 60);
+        const missing = [];
+        for (const sec of todo) {
+            const file = join(framesDir, frameFileName(sec));
+            await run('ffmpeg', [
+                ...['-v', 'error', '-y', '-ss', String(sec), '-i', video],
+                ...['-frames:v', '1', '-vf', "scale='min(960,iw)':-2", '-q:v', '5'],
+                file,
+            ], { cwd, timeoutMs: frameTimeout });
+            // 영상 끝을 넘긴 시각이면 ffmpeg가 파일 없이 0으로 끝나기도 한다
+            if (!existsSync(file) || statSync(file).size === 0)
+                missing.push(sec);
+        }
+        return {
+            framesDir,
+            extracted: todo.length - missing.length,
+            skipped: secs.length - todo.length,
+            total: secs.length,
+            missing,
+        };
+    }
+    finally {
+        rmSync(cwd, { recursive: true, force: true });
+    }
+}
+//# sourceMappingURL=frames.js.map

package/dist/src/digest/frames.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"frames.js","sourceRoot":"","sources":["../../../src/digest/frames.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,SAAS,EAAE,WAAW,EAAE,WAAW,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,SAAS,CAAC;AAC5F,OAAO,EAAE,MAAM,EAAE,MAAM,SAAS,CAAC;AACjC,OAAO,EAAE,IAAI,EAAE,MAAM,WAAW,CAAC;AACjC,OAAO,EAAE,aAAa,EAAE,MAAM,UAAU,CAAC;AACzC,OAAO,EAAE,eAAe,EAAgC,MAAM,aAAa,CAAC;AAC5E,OAAO,EAAE,WAAW,EAAE,MAAM,YAAY,CAAC;AACzC,OAAO,EAAE,UAAU,EAAE,UAAU,EAAE,KAAK,EAAmB,MAAM,YAAY,CAAC;AAkB5E,4CAA4C;AAC5C,MAAM,UAAU,aAAa,CAAC,GAAW;IACvC,OAAO,IAAI,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,IAAI,CAAC,CAAC,CAAC,QAAQ,CAAC,CAAC,EAAE,GAAG,CAAC,MAAM,CAAC;AACnE,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,MAAc;IAC7C,MAAM,IAAI,GAAG,MAAM,CAAC,OAAO,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CACxC,CAAC,CAAC,QAAQ,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,EAAE,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC,CACpE,CAAC;IACF,OAAO,CAAC,GAAG,IAAI,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;AAClD,CAAC;AAED;;;GAGG;AACH,MAAM,MAAM,GAAG,sDAAsD,CAAC;AACtE,MAAM,YAAY,GAAG,gBAAgB,CAAC;AAEtC,6EAA6E;AAC7E,MAAM,YAAY,GAAG,yBAAyB,CAAC;AAE/C,SAAS,eAAe,CAAC,QAAgB;IACvC,IAAI,CAAC,UAAU,CAAC,QAAQ,CAAC;QAAE,OAAO,IAAI,CAAC;IACvC,MAAM,GAAG,GAAG,WAAW,CAAC,QAAQ,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;IACpE,OAAO,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC;AAC1C,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,WAAW,CAC/B,GAAe,EACf,UAAsB,EACtB,QAAgB,EAChB,GAAW;IAEX,MAAM,EAAE,IAAI,EAAE,GAAG,UAAU,CAAC;IAC5B,IAAI,IAAI,CAAC,MAAM,KAAK,MAAM,EAAE,CAAC;QAC3B,IAAI,CAAC,IAAI,CAAC,GAAG;YAAE,MAAM,IAAI,KAAK,CAAC,4BAA4B,CAAC,CAAC;QAC7D,OAAO,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACjC,CAAC;IACD,MAAM,MAAM,GAAG,eAAe,CAAC,QAAQ,CAAC,CAAC;IACzC,IAAI,MAAM;QAAE,OAAO,MAAM,CAAC;IAE1B,IAAI,CAAC,IAAI,CAAC,OAAO;QAAE,MAAM,IAAI,KAAK,CAAC,2BAA2B,CAAC,CAAC;IAChE,SAAS,CAAC,QAAQ,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IACzC,8DAA8D;IAC9D,MAAM,GAAG,GAAG,CAAC,IAAI,EAAE,IAAI,CAAC,QAAQ,EAAE,eAAe,CAAC,EAAE,IAAI,EAAE,eAAe,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC;IACzF,MAAM,KAAK,CAAC,GAAG,EAAE,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,GAAG,CAAC,EAAE,CAAC,IAAI,EAAE,YAAY,EAAE,GAAG,GAAG,CAAC,EAAE;QACrE,GAAG;QACH,SAAS,EAAE,UAAU,CAAC,IAAI,CAAC,WAAW,EAAE,CAAC,CAAC;KAC3C,CAAC,CAAC;IACH,MAAM,GAAG,GAAG,eAAe,CAAC,QAAQ,CAAC,CAAC;IACtC,IAAI,CAAC,GAAG;QAAE,MAAM,IAAI,KAAK,CAAC,cAAc,CAAC,CAAC;IAC1C,OAAO,GAAG,CAAC;AACb,CAAC;AAED;;;GAGG;AACH,MAAM,CAAC,KAAK,UAAU,aAAa,CAAC,IAAY,EAAE,IAAmB;IACnE,MAAM,GAAG,GAAG,IAAI,CAAC,GAAG,IAAI,UAAU,CAAC;IACnC,MAAM,KAAK,GAAG,IAAI,WAAW,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC;IAC7C,MAAM,CAAC,GAAG,KAAK,CAAC,cAAc,CAAC,IAAI,CAAC,CAAC;IACrC,IAAI,CAAC,CAAC,CAAC,EAAE;QAAE,MAAM,IAAI,KAAK,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;IACtC,MAAM,UAAU,GAAG,CAAC,CAAC,UAAU,CAAC;IAChC,MAAM,MAAM,GAAG,KAAK,CAAC,UAAU,CAAC,IAAI,CAAC,CAAC;IACtC,IAAI,CAAC,MAAM,EAAE,CAAC;QACZ,MAAM,IAAI,KAAK,CACb,4DAA4D,IAAI,EAAE,CACnE,CAAC;IACJ,CAAC;IAED,MAAM,IAAI,GAAG,gBAAgB,CAAC,MAAM,CAAC,CAAC;IACtC,MAAM,SAAS,GAAG,KAAK,CAAC,SAAS,CAAC,IAAI,CAAC,CAAC;IACxC,SAAS,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAC1C,MAAM,IAAI,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,EAAE;QAC/B,MAAM,CAAC,GAAG,IAAI,CAAC,SAAS,EAAE,aAAa,CAAC,GAAG,CAAC,CAAC,CAAC;QAC9C,OAAO,IAAI,CAAC,KAAK,IAAI,CAAC,UAAU,CAAC,CAAC,CAAC,IAAI,QAAQ,CAAC,CAAC,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC;IAChE,CAAC,CAAC,CAAC;IACH,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACtB,OAAO,EAAE,SAAS,EAAE,SAAS,EAAE,CAAC,EAAE,OAAO,EAAE,IAAI,CAAC,MAAM,EAAE,KAAK,EAAE,IAAI,CAAC,MAAM,EAAE,OAAO,EAAE,EAAE,EAAE,CAAC;IAC5F,CAAC;IAED,sDAAsD;IACtD,MAAM,GAAG,GAAG,WAAW,CAAC,IAAI,CAAC,MAAM,EAAE,EAAE,qBAAqB,CAAC,CAAC,CAAC;IAC/D,IAAI,CAAC;QACH,MAAM,KAAK,GAAG,MAAM,WAAW,CAAC,GAAG,EAAE,UAAU,EAAE,KAAK,CAAC,aAAa,CAAC,IAAI,CAAC,EAAE,GAAG,CAAC,CAAC;QACjF,2CAA2C;QAC3C,MAAM,YAAY,GAAG,UAAU,CAAC,UAAU,CAAC,IAAI,CAAC,WAAW,EAAE,IAAI,EAAE,EAAE,CAAC,CAAC;QACvE,MAAM,OAAO,GAAa,EAAE,CAAC;QAC7B,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;YACvB,MAAM,IAAI,GAAG,IAAI,CAAC,SAAS,EAAE,aAAa,CAAC,GAAG,CAAC,CAAC,CAAC;YACjD,MAAM,GAAG,CACP,QAAQ,EACR;gBACE,GAAG,CAAC,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,KAAK,EAAE,MAAM,CAAC,GAAG,CAAC,EAAE,IAAI,EAAE,KAAK,CAAC;gBACzD,GAAG,CAAC,WAAW,EAAE,GAAG,EAAE,KAAK,EAAE,wBAAwB,EAAE,MAAM,EAAE,GAAG,CAAC;gBACnE,IAAI;aACL,EACD,EAAE,GAAG,EAAE,SAAS,EAAE,YAAY,EAAE,CACjC,CAAC;YACF,0CAA0C;YAC1C,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,IAAI,KAAK,CAAC;gBAAE,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACxE,CAAC;QACD,OAAO;YACL,SAAS;YACT,SAAS,EAAE,IAAI,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM;YACvC,OAAO,EAAE,IAAI,CAAC,MAAM,GAAG,IAAI,CAAC,MAAM;YAClC,KAAK,EAAE,IAAI,CAAC,MAAM;YAClB,OAAO;SACR,CAAC;IACJ,CAAC;YAAS,CAAC;QACT,MAAM,CAAC,GAAG,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;IAChD,CAAC;AACH,CAAC"}

package/dist/src/digest/index.d.ts

@@ -0,0 +1,25 @@
+export * from './schema.js';
+export * from './lint.js';
+export * from './store.js';
+export * from './captions.js';
+export * from './ingest.js';
+export * from './frames.js';
+export * from './format.js';
+export * from './render-md.js';
+export * from './render-html.js';
+export * from './render.js';
+export * from './server.js';
+export { findSegmentIndex } from './viewer-logic.js';
+export * from './prose.js';
+export * from './reports.js';
+export * from './merge.js';
+export * from './slice.js';
+export * from './translate.js';
+export * from './ocr.js';
+export * from './blocks.js';
+export * from './templates.js';
+export * from './lint-extension.js';
+export * from './render-blocks.js';
+export * from './comments.js';
+export * from './svg.js';
+//# sourceMappingURL=index.d.ts.map

package/dist/src/digest/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../../src/digest/index.ts"],"names":[],"mappings":"AAAA,cAAc,aAAa,CAAC;AAC5B,cAAc,WAAW,CAAC;AAC1B,cAAc,YAAY,CAAC;AAC3B,cAAc,eAAe,CAAC;AAC9B,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,kBAAkB,CAAC;AACjC,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,OAAO,EAAE,gBAAgB,EAAE,MAAM,mBAAmB,CAAC;AACrD,cAAc,YAAY,CAAC;AAC3B,cAAc,cAAc,CAAC;AAC7B,cAAc,YAAY,CAAC;AAC3B,cAAc,YAAY,CAAC;AAC3B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,UAAU,CAAC;AACzB,cAAc,aAAa,CAAC;AAC5B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,qBAAqB,CAAC;AACpC,cAAc,oBAAoB,CAAC;AACnC,cAAc,eAAe,CAAC;AAC9B,cAAc,UAAU,CAAC"}

package/dist/src/digest/index.js

@@ -0,0 +1,25 @@
+export * from './schema.js';
+export * from './lint.js';
+export * from './store.js';
+export * from './captions.js';
+export * from './ingest.js';
+export * from './frames.js';
+export * from './format.js';
+export * from './render-md.js';
+export * from './render-html.js';
+export * from './render.js';
+export * from './server.js';
+export { findSegmentIndex } from './viewer-logic.js';
+export * from './prose.js';
+export * from './reports.js';
+export * from './merge.js';
+export * from './slice.js';
+export * from './translate.js';
+export * from './ocr.js';
+export * from './blocks.js';
+export * from './templates.js';
+export * from './lint-extension.js';
+export * from './render-blocks.js';
+export * from './comments.js';
+export * from './svg.js';
+//# sourceMappingURL=index.js.map

package/dist/src/digest/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../../src/digest/index.ts"],"names":[],"mappings":"AAAA,cAAc,aAAa,CAAC;AAC5B,cAAc,WAAW,CAAC;AAC1B,cAAc,YAAY,CAAC;AAC3B,cAAc,eAAe,CAAC;AAC9B,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,kBAAkB,CAAC;AACjC,cAAc,aAAa,CAAC;AAC5B,cAAc,aAAa,CAAC;AAC5B,OAAO,EAAE,gBAAgB,EAAE,MAAM,mBAAmB,CAAC;AACrD,cAAc,YAAY,CAAC;AAC3B,cAAc,cAAc,CAAC;AAC7B,cAAc,YAAY,CAAC;AAC3B,cAAc,YAAY,CAAC;AAC3B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,UAAU,CAAC;AACzB,cAAc,aAAa,CAAC;AAC5B,cAAc,gBAAgB,CAAC;AAC/B,cAAc,qBAAqB,CAAC;AACpC,cAAc,oBAAoB,CAAC;AACnC,cAAc,eAAe,CAAC;AAC9B,cAAc,UAAU,CAAC"}

package/dist/src/digest/ingest.d.ts

@@ -0,0 +1,65 @@
+import { type CaptionSource } from './schema.js';
+import { type RunCommand } from './ytdlp.js';
+export interface IngestOptions {
+    repoRoot: string;
+    /** 자막 언어 우선순위. 없으면 영상 원어를 쓴다 */
+    lang?: string;
+    slug?: string;
+    /** 자막이 있어도 whisper로 받아쓴다 */
+    whisper?: boolean;
+    whisperModel?: string;
+    run?: RunCommand;
+}
+export interface IngestResult {
+    slug: string;
+    transcriptPath: string;
+    title: string;
+    durationSec: number;
+    captionSource: CaptionSource;
+    /** 실제로 쓴 자막 트랙 이름(en-orig 같은). whisper면 모델 이름 */
+    captionTrack: string;
+    segmentCount: number;
+    /**
+     * 같은 slug를 다시 받아 문장이 바뀌면 기존 digest.json의 cite가 엉뚱한 문장을 가리킨다.
+     * 그래서 옮겨 두고 그 경로와 이전 문장 수를 싣는다
+     */
+    staleDigest?: {
+        movedTo: string;
+        previousSegmentCount: number | null;
+    };
+}
+/** watch?v=, youtu.be/, shorts/, embed/, live/ 꼴에서 영상 id를 꺼낸다. 유튜브 주소가 아니면 null */
+export declare function parseYoutubeId(input: string): string | null;
+interface YtTrack {
+    ext?: string;
+}
+interface YtMeta {
+    id: string;
+    title?: string;
+    channel?: string;
+    uploader?: string;
+    duration?: number;
+    upload_date?: string;
+    language?: string | null;
+    webpage_url?: string;
+    subtitles?: Record<string, YtTrack[]>;
+    automatic_captions?: Record<string, YtTrack[]>;
+}
+export interface CaptionPick {
+    kind: 'manual' | 'auto';
+    key: string;
+    lang: string;
+}
+/**
+ * 수동 자막 → 자동 자막 순으로 고른다. 자동 자막은 157개 언어가 기계 번역으로 딸려 오는데
+ * 영상 원어 트랙만 `-orig`가 붙어서 그걸 먼저 잡는다. 번역 자막은 문장 시각이 어긋나 쓰지 않는다.
+ * 자동 더빙이 붙은 영상은 더빙 언어마다 `-orig`가 생겨서 원어에 맞는 것을 먼저 고른다.
+ */
+export declare function pickCaption(meta: YtMeta, prefLang?: string): CaptionPick | null;
+/**
+ * 유튜브 주소나 로컬 영상 파일에서 문장 단위 스크립트를 만들어 transcript.json으로 저장한다.
+ * 자막 순서는 수동 자막 → 자동 자막(문장 병합) → faster-whisper다.
+ */
+export declare function ingest(input: string, opts: IngestOptions): Promise<IngestResult>;
+export {};
+//# sourceMappingURL=ingest.d.ts.map

package/dist/src/digest/ingest.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"ingest.d.ts","sourceRoot":"","sources":["../../../src/digest/ingest.ts"],"names":[],"mappings":"AAeA,OAAO,EAGL,KAAK,aAAa,EAEnB,MAAM,aAAa,CAAC;AAErB,OAAO,EAML,KAAK,UAAU,EAChB,MAAM,YAAY,CAAC;AAEpB,MAAM,WAAW,aAAa;IAC5B,QAAQ,EAAE,MAAM,CAAC;IACjB,gCAAgC;IAChC,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,4BAA4B;IAC5B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,GAAG,CAAC,EAAE,UAAU,CAAC;CAClB;AAED,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE,MAAM,CAAC;IACb,cAAc,EAAE,MAAM,CAAC;IACvB,KAAK,EAAE,MAAM,CAAC;IACd,WAAW,EAAE,MAAM,CAAC;IACpB,aAAa,EAAE,aAAa,CAAC;IAC7B,iDAAiD;IACjD,YAAY,EAAE,MAAM,CAAC;IACrB,YAAY,EAAE,MAAM,CAAC;IACrB;;;OAGG;IACH,WAAW,CAAC,EAAE;QAAE,OAAO,EAAE,MAAM,CAAC;QAAC,oBAAoB,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,CAAC;CACxE;AAMD,mFAAmF;AACnF,wBAAgB,cAAc,CAAC,KAAK,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAkB3D;AAED,UAAU,OAAO;IACf,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AACD,UAAU,MAAM;IACd,EAAE,EAAE,MAAM,CAAC;IACX,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,QAAQ,CAAC,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,SAAS,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,EAAE,CAAC,CAAC;IACtC,kBAAkB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,EAAE,CAAC,CAAC;CAChD;AAED,MAAM,WAAW,WAAW;IAC1B,IAAI,EAAE,QAAQ,GAAG,MAAM,CAAC;IACxB,GAAG,EAAE,MAAM,CAAC;IACZ,IAAI,EAAE,MAAM,CAAC;CACd;AAUD;;;;GAIG;AACH,wBAAgB,WAAW,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,CAAC,EAAE,MAAM,GAAG,WAAW,GAAG,IAAI,CAmB/E;AA2PD;;;GAGG;AACH,wBAAsB,MAAM,CAAC,KAAK,EAAE,MAAM,EAAE,IAAI,EAAE,aAAa,GAAG,OAAO,CAAC,YAAY,CAAC,CA8CtF"}

package/dist/src/digest/ingest.js

@@ -0,0 +1,303 @@
+import { createHash } from 'node:crypto';
+import { existsSync, mkdtempSync, readFileSync, readdirSync, renameSync, rmSync, statSync, } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { basename, dirname, extname, join, resolve } from 'node:path';
+import { pathToFileURL } from 'node:url';
+import { log } from '../core/log.js';
+import { captionsToSentences, cuesToWords, mergeSentences } from './captions.js';
+import { TRANSCRIPT_SCHEMA_VERSION, youtubeWatchUrl, } from './schema.js';
+import { DigestStore, isValidSlug, slugFromMeta } from './store.js';
+import { CommandError, QUICK_TIMEOUT_MS, runCommand, timeoutFor, ytdlp, } from './ytdlp.js';
+const DEFAULT_WHISPER_MODEL = 'large-v3';
+const YOUTUBE_ID = /^[A-Za-z0-9_-]{11}$/;
+/** watch?v=, youtu.be/, shorts/, embed/, live/ 꼴에서 영상 id를 꺼낸다. 유튜브 주소가 아니면 null */
+export function parseYoutubeId(input) {
+    let url;
+    try {
+        url = new URL(input);
+    }
+    catch {
+        return null;
+    }
+    const host = url.hostname.replace(/^(www|m|music)\./, '');
+    let id = null;
+    if (host === 'youtu.be')
+        id = url.pathname.slice(1).split('/')[0] ?? null;
+    else if (host === 'youtube.com' || host === 'youtube-nocookie.com') {
+        if (url.pathname === '/watch')
+            id = url.searchParams.get('v');
+        else {
+            const m = url.pathname.match(/^\/(?:shorts|embed|live|v)\/([^/?#]+)/);
+            id = m ? m[1] : null;
+        }
+    }
+    return id && YOUTUBE_ID.test(id) ? id : null;
+}
+const langOf = (key) => key.replace(/-orig$/, '');
+const matches = (key, pref) => key === pref || key.startsWith(`${pref}-`) || langOf(key) === pref;
+const primary = (tag) => tag.split('-')[0].toLowerCase();
+// 메타 원어는 `en-US`인데 트랙은 `en-orig`처럼 지역 태그만 다를 때가 있어 주 언어로 한 번 더 맞춘다
+const findPref = (keys, pref) => keys.find((k) => matches(k, pref)) ?? keys.find((k) => primary(langOf(k)) === primary(pref));
+/**
+ * 수동 자막 → 자동 자막 순으로 고른다. 자동 자막은 157개 언어가 기계 번역으로 딸려 오는데
+ * 영상 원어 트랙만 `-orig`가 붙어서 그걸 먼저 잡는다. 번역 자막은 문장 시각이 어긋나 쓰지 않는다.
+ * 자동 더빙이 붙은 영상은 더빙 언어마다 `-orig`가 생겨서 원어에 맞는 것을 먼저 고른다.
+ */
+export function pickCaption(meta, prefLang) {
+    const prefs = [prefLang, meta.language ?? undefined].filter((p) => !!p);
+    const manual = Object.keys(meta.subtitles ?? {}).filter((k) => k !== 'live_chat');
+    const auto = Object.keys(meta.automatic_captions ?? {});
+    for (const p of prefs) {
+        const k = findPref(manual, p);
+        if (k)
+            return { kind: 'manual', key: k, lang: langOf(k) };
+    }
+    if (prefs.length === 0 && manual.length === 1) {
+        return { kind: 'manual', key: manual[0], lang: langOf(manual[0]) };
+    }
+    const origs = auto.filter((k) => k.endsWith('-orig'));
+    const orig = prefs.map((p) => findPref(origs, p)).find((k) => !!k) ?? origs[0];
+    if (orig)
+        return { kind: 'auto', key: orig, lang: langOf(orig) };
+    for (const p of prefs) {
+        if (auto.includes(p))
+            return { kind: 'auto', key: p, lang: p };
+    }
+    return null;
+}
+const toIsoDate = (yyyymmdd) => yyyymmdd && /^\d{8}$/.test(yyyymmdd)
+    ? `${yyyymmdd.slice(0, 4)}-${yyyymmdd.slice(4, 6)}-${yyyymmdd.slice(6, 8)}`
+    : undefined;
+const round2 = (n) => Math.round(n * 100) / 100;
+function toSegments(cues) {
+    let prev = 0;
+    return cues.map((c, i) => {
+        // 수동 자막은 큐가 겹치기도 해서 시각이 뒤로 가는 경우만 앞 문장에 맞춘다
+        prev = Math.max(prev, round2(c.startSec));
+        return { i, startSec: prev, text: c.text };
+    });
+}
+function findFile(dir, prefix, exts) {
+    const hit = readdirSync(dir)
+        .filter((f) => f.startsWith(prefix) && exts.includes(extname(f).toLowerCase()))
+        .sort();
+    return hit.length > 0 ? join(dir, hit[0]) : null;
+}
+// 모델을 받을 때 진행 로그가 stdout에 섞일 수 있어서 결과는 마지막 줄 한 줄로만 낸다
+const WHISPER_SCRIPT = `
+import json, sys
+from faster_whisper import WhisperModel
+audio, model_name, lang = sys.argv[1], sys.argv[2], sys.argv[3] or None
+model = WhisperModel(model_name, device="auto", compute_type="auto")
+segments, info = model.transcribe(audio, language=lang, vad_filter=True)
+out = [{"start": s.start, "text": s.text.strip()} for s in segments if s.text.strip()]
+sys.stdout.write("\\n" + json.dumps({"language": info.language, "segments": out}) + "\\n")
+`;
+const mmss = (sec) => `${Math.floor(sec / 60)}:${String(Math.floor(sec % 60)).padStart(2, '0')}`;
+async function transcribeWithWhisper(run, cwd, audioPath, model, durationSec, lang) {
+    log(`faster-whisper ${model} 모델로 ${mmss(durationSec)} 분량을 받아써요. ` +
+        'GPU가 없으면 영상 길이보다 오래 걸릴 수 있어요.');
+    let stdout;
+    try {
+        ({ stdout } = await run('python3', ['-c', WHISPER_SCRIPT, audioPath, model, lang ?? ''], {
+            cwd,
+            // CPU로 큰 모델을 돌리면 영상 길이의 몇 배가 걸린다. 멈춘 프로세스만 끊을 만큼 넉넉히 둔다
+            timeoutMs: timeoutFor(durationSec, 20, 600),
+        }));
+    }
+    catch (e) {
+        // faster-whisper 의존 모듈(ctranslate2, av)이 빠진 경우는 설치 안내가 틀려서 원래 오류를 그대로 올린다
+        if (e instanceof CommandError && /No module named ['"]faster_whisper/.test(e.stderr)) {
+            throw new Error('자막이 없어서 faster-whisper로 받아쓰려 했는데 설치돼 있지 않아요. ' +
+                '`python3 -m pip install faster-whisper`로 설치한 뒤 다시 돌려 주세요.', { cause: e });
+        }
+        throw e;
+    }
+    const last = stdout.trim().split('\n').pop() ?? '';
+    let parsed;
+    try {
+        parsed = JSON.parse(last);
+    }
+    catch {
+        throw new Error(`받아쓰기 결과를 JSON으로 못 읽었어요: ${stdout.trim().slice(0, 200)}`);
+    }
+    return {
+        lang: parsed.language,
+        cues: parsed.segments.map((s) => ({ startSec: s.start, text: s.text })),
+    };
+}
+async function probeDuration(run, cwd, file) {
+    const { stdout } = await run('ffprobe', ['-v', 'error', '-show_entries', 'format=duration', '-of', 'csv=p=0', file], { cwd, timeoutMs: QUICK_TIMEOUT_MS });
+    const sec = Number(stdout.trim());
+    if (!Number.isFinite(sec) || sec <= 0)
+        throw new Error(`영상 길이를 못 읽었어요: ${file}`);
+    return sec;
+}
+async function collectYoutube(videoId, opts, run, work) {
+    const watchUrl = youtubeWatchUrl(videoId);
+    const quick = { cwd: work, timeoutMs: QUICK_TIMEOUT_MS };
+    const { stdout } = await ytdlp(run, ['-J', '--skip-download', watchUrl], undefined, quick);
+    const meta = JSON.parse(stdout);
+    const title = meta.title?.trim() || videoId;
+    const base = {
+        source: 'youtube',
+        videoId,
+        url: watchUrl,
+        title,
+        ...(meta.channel || meta.uploader ? { channel: meta.channel ?? meta.uploader } : {}),
+        durationSec: meta.duration ?? 0,
+        ...(toIsoDate(meta.upload_date) ? { publishedAt: toIsoDate(meta.upload_date) } : {}),
+    };
+    const slug = opts.slug ?? slugFromMeta(videoId, title);
+    const pick = opts.whisper ? null : pickCaption(meta, opts.lang);
+    if (pick) {
+        const flag = pick.kind === 'manual' ? '--write-subs' : '--write-auto-subs';
+        const subArgs = [
+            ...['--skip-download', flag, '--sub-langs', pick.key, '--sub-format', 'vtt/srt/best'],
+            ...['-o', join(work, 'sub.%(ext)s'), watchUrl],
+        ];
+        await ytdlp(run, subArgs, undefined, quick);
+        const file = findFile(work, 'sub.', ['.vtt', '.srt']);
+        const cues = file ? captionsToSentences(readFileSync(file, 'utf-8')) : [];
+        if (cues.length > 0) {
+            return {
+                slug,
+                meta: { ...base, lang: pick.lang, captionSource: pick.kind },
+                cues,
+                track: pick.key,
+            };
+        }
+    }
+    const model = opts.whisperModel ?? DEFAULT_WHISPER_MODEL;
+    await ytdlp(run, ['-f', 'bestaudio/best', '-o', join(work, 'audio.%(ext)s'), watchUrl], undefined, { cwd: work, timeoutMs: timeoutFor(base.durationSec, 1) });
+    const audio = findFile(work, 'audio.', ['.m4a', '.webm', '.mp4', '.opus', '.mp3', '.ogg']);
+    if (!audio)
+        throw new Error('오디오를 받지 못했어요.');
+    const w = await transcribeWithWhisper(run, work, audio, model, base.durationSec, opts.lang ?? meta.language ?? undefined);
+    return {
+        slug,
+        meta: { ...base, lang: w.lang, captionSource: 'whisper' },
+        cues: mergeSentences(cuesToWords(w.cues)),
+        track: model,
+    };
+}
+async function collectFile(file, opts, run, work) {
+    const stem = basename(file, extname(file));
+    const durationSec = await probeDuration(run, work, file);
+    // 다른 폴더의 같은 이름 영상과 섞이지 않게 경로 해시를 slug 끝에 붙인다
+    const hash = createHash('sha1').update(file).digest('hex').slice(0, 8);
+    const slug = opts.slug ?? slugFromMeta(hash, stem);
+    const base = {
+        source: 'file',
+        url: pathToFileURL(file).href,
+        title: stem,
+        durationSec: round2(durationSec),
+    };
+    const sidecar = opts.whisper ? null : findFile(dirname(file), `${stem}.`, ['.vtt', '.srt']);
+    if (sidecar) {
+        const cues = captionsToSentences(readFileSync(sidecar, 'utf-8'));
+        if (cues.length > 0) {
+            return {
+                slug,
+                meta: { ...base, lang: opts.lang ?? 'und', captionSource: 'manual' },
+                cues,
+                track: basename(sidecar),
+            };
+        }
+    }
+    const model = opts.whisperModel ?? DEFAULT_WHISPER_MODEL;
+    const w = await transcribeWithWhisper(run, work, file, model, base.durationSec, opts.lang);
+    return {
+        slug,
+        meta: { ...base, lang: w.lang, captionSource: 'whisper' },
+        cues: mergeSentences(cuesToWords(w.cues)),
+        track: model,
+    };
+}
+const sameSegments = (a, b) => a.length === b.length &&
+    a.every((s, i) => s.startSec === b[i].startSec && s.text === b[i].text);
+/**
+ * 문장이 바뀌었는데 digest.json을 두면 cite가 옛 번호로 엉뚱한 장면을 가리킨다.
+ * 지우지 않고 옮겨 두면 다음 validate가 DIGEST_NOT_FOUND로 요약을 새로 쓰게 이끈다
+ */
+function moveStaleDigest(store, next) {
+    const digestPath = store.digestPath(next.slug);
+    if (!existsSync(digestPath))
+        return undefined;
+    const prev = store.readTranscript(next.slug);
+    if (prev.ok && sameSegments(prev.transcript.segments, next.segments))
+        return undefined;
+    const stamp = new Date().toISOString().replace(/[:.]/g, '-');
+    const movedTo = join(store.dir(next.slug), `digest.stale-${stamp}.json`);
+    renameSync(digestPath, movedTo);
+    return { movedTo, previousSegmentCount: prev.ok ? prev.transcript.segments.length : null };
+}
+function restoreStaleDigest(store, slug, movedTo, cause) {
+    let why;
+    try {
+        renameSync(movedTo, store.digestPath(slug));
+        return;
+    }
+    catch (restoreErr) {
+        why = restoreErr instanceof Error ? restoreErr.message : String(restoreErr);
+    }
+    // catch 밖에서 던져야 cause에 되돌리기 오류가 아닌 저장 오류를 둘 수 있다.
+    // CLI는 message만 찍어서 저장이 왜 실패했는지와 digest가 어디 남았는지, 왜 못 되돌렸는지를 message에 다 넣는다
+    const reason = cause instanceof Error ? cause.message : String(cause);
+    throw new Error(`${reason} (옮겨 둔 digest.json을 되돌리지 못해 ${movedTo}에 남아 있어요: ${why})`, { cause });
+}
+/**
+ * 유튜브 주소나 로컬 영상 파일에서 문장 단위 스크립트를 만들어 transcript.json으로 저장한다.
+ * 자막 순서는 수동 자막 → 자동 자막(문장 병합) → faster-whisper다.
+ */
+export async function ingest(input, opts) {
+    const run = opts.run ?? runCommand;
+    if (opts.slug !== undefined && !isValidSlug(opts.slug)) {
+        throw new Error(`slug는 소문자, 숫자, 하이픈만 쓸 수 있어요: ${opts.slug}`);
+    }
+    const videoId = parseYoutubeId(input);
+    const filePath = videoId ? null : resolve(input);
+    if (!videoId && !(filePath && existsSync(filePath) && statSync(filePath).isFile())) {
+        throw new Error(`유튜브 주소도 아니고 파일도 찾지 못했어요: ${input}`);
+    }
+    const work = mkdtempSync(join(tmpdir(), 'gestalt-digest-'));
+    try {
+        const c = videoId
+            ? await collectYoutube(videoId, opts, run, work)
+            : await collectFile(filePath, opts, run, work);
+        if (c.cues.length === 0)
+            throw new Error('스크립트 문장이 하나도 안 나왔어요.');
+        const transcript = {
+            schemaVersion: TRANSCRIPT_SCHEMA_VERSION,
+            slug: c.slug,
+            meta: c.meta,
+            segments: toSegments(c.cues),
+        };
+        const store = new DigestStore(opts.repoRoot);
+        const staleDigest = moveStaleDigest(store, transcript);
+        let transcriptPath;
+        try {
+            transcriptPath = store.saveTranscript(transcript);
+        }
+        catch (e) {
+            // 저장 뒤에 옮기면 rename 실패 때 새 transcript 옆에 옛 digest가 남는다. 옮긴 걸 되돌려 옛 짝을 지킨다
+            if (staleDigest)
+                restoreStaleDigest(store, c.slug, staleDigest.movedTo, e);
+            throw e;
+        }
+        return {
+            slug: c.slug,
+            transcriptPath,
+            title: c.meta.title,
+            durationSec: c.meta.durationSec,
+            captionSource: c.meta.captionSource,
+            captionTrack: c.track,
+            segmentCount: transcript.segments.length,
+            ...(staleDigest ? { staleDigest } : {}),
+        };
+    }
+    finally {
+        rmSync(work, { recursive: true, force: true });
+    }
+}
+//# sourceMappingURL=ingest.js.map