@steipete/summarize-core 0.7.0 → 0.7.1

package/dist/cjs/content/transcript/providers/podcast/media.js

@@ -0,0 +1,355 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.transcribeMediaUrl = transcribeMediaUrl;
+exports.probeRemoteMedia = probeRemoteMedia;
+exports.downloadCappedBytes = downloadCappedBytes;
+exports.downloadToFile = downloadToFile;
+exports.normalizeHeaderType = normalizeHeaderType;
+exports.parseContentLength = parseContentLength;
+exports.filenameFromUrl = filenameFromUrl;
+exports.formatBytes = formatBytes;
+const node_crypto_1 = require("node:crypto");
+const node_fs_1 = require("node:fs");
+const node_os_1 = require("node:os");
+const node_path_1 = require("node:path");
+const whisper_js_1 = require("../../../../transcription/whisper.js");
+const constants_js_1 = require("./constants.js");
+async function transcribeMediaUrl({ fetchImpl, url, filenameHint, durationSecondsHint, openaiApiKey, falApiKey, notes, progress, }) {
+    const canChunk = await (0, whisper_js_1.isFfmpegAvailable)();
+    const providerHint = (await (0, whisper_js_1.isWhisperCppReady)())
+        ? 'cpp'
+        : openaiApiKey && falApiKey
+            ? 'openai->fal'
+            : openaiApiKey
+                ? 'openai'
+                : falApiKey
+                    ? 'fal'
+                    : 'unknown';
+    const head = await probeRemoteMedia(fetchImpl, url);
+    if (head.contentLength !== null && head.contentLength > constants_js_1.MAX_REMOTE_MEDIA_BYTES) {
+        throw new Error(`Remote media too large (${formatBytes(head.contentLength)}). Limit is ${formatBytes(constants_js_1.MAX_REMOTE_MEDIA_BYTES)}.`);
+    }
+    const mediaType = head.mediaType ?? 'application/octet-stream';
+    const filename = head.filename ?? filenameHint;
+    const totalBytes = head.contentLength;
+    progress?.onProgress?.({
+        kind: 'transcript-media-download-start',
+        url: progress.url,
+        service: progress.service,
+        mediaUrl: url,
+        totalBytes,
+    });
+    const modelId = providerHint === 'cpp'
+        ? ((await (0, whisper_js_1.resolveWhisperCppModelNameForDisplay)()) ?? 'whisper.cpp')
+        : openaiApiKey && falApiKey
+            ? 'whisper-1->fal-ai/wizper'
+            : openaiApiKey
+                ? 'whisper-1'
+                : falApiKey
+                    ? 'fal-ai/wizper'
+                    : null;
+    if (!canChunk) {
+        const bytes = await downloadCappedBytes(fetchImpl, url, whisper_js_1.MAX_OPENAI_UPLOAD_BYTES, {
+            totalBytes,
+            onProgress: (downloadedBytes) => progress?.onProgress?.({
+                kind: 'transcript-media-download-progress',
+                url: progress.url,
+                service: progress.service,
+                downloadedBytes,
+                totalBytes,
+            }),
+        });
+        progress?.onProgress?.({
+            kind: 'transcript-media-download-done',
+            url: progress.url,
+            service: progress.service,
+            downloadedBytes: bytes.byteLength,
+            totalBytes,
+        });
+        progress?.onProgress?.({
+            kind: 'transcript-whisper-start',
+            url: progress.url,
+            service: progress.service,
+            providerHint,
+            modelId,
+            totalDurationSeconds: durationSecondsHint,
+            parts: null,
+        });
+        notes.push(`Transcribed first ${formatBytes(bytes.byteLength)} only (ffmpeg not available)`);
+        const transcript = await (0, whisper_js_1.transcribeMediaWithWhisper)({
+            bytes,
+            mediaType,
+            filename,
+            openaiApiKey,
+            falApiKey,
+            totalDurationSeconds: durationSecondsHint,
+            onProgress: (event) => {
+                progress?.onProgress?.({
+                    kind: 'transcript-whisper-progress',
+                    url: progress.url,
+                    service: progress.service,
+                    processedDurationSeconds: event.processedDurationSeconds,
+                    totalDurationSeconds: event.totalDurationSeconds,
+                    partIndex: event.partIndex,
+                    parts: event.parts,
+                });
+            },
+        });
+        if (transcript.notes.length > 0)
+            notes.push(...transcript.notes);
+        return { text: transcript.text, provider: transcript.provider, error: transcript.error };
+    }
+    if (head.contentLength !== null && head.contentLength <= whisper_js_1.MAX_OPENAI_UPLOAD_BYTES) {
+        const bytes = await downloadCappedBytes(fetchImpl, url, whisper_js_1.MAX_OPENAI_UPLOAD_BYTES, {
+            totalBytes,
+            onProgress: (downloadedBytes) => progress?.onProgress?.({
+                kind: 'transcript-media-download-progress',
+                url: progress.url,
+                service: progress.service,
+                downloadedBytes,
+                totalBytes,
+            }),
+        });
+        progress?.onProgress?.({
+            kind: 'transcript-media-download-done',
+            url: progress.url,
+            service: progress.service,
+            downloadedBytes: bytes.byteLength,
+            totalBytes,
+        });
+        progress?.onProgress?.({
+            kind: 'transcript-whisper-start',
+            url: progress.url,
+            service: progress.service,
+            providerHint,
+            modelId,
+            totalDurationSeconds: durationSecondsHint,
+            parts: null,
+        });
+        const transcript = await (0, whisper_js_1.transcribeMediaWithWhisper)({
+            bytes,
+            mediaType,
+            filename,
+            openaiApiKey,
+            falApiKey,
+            totalDurationSeconds: durationSecondsHint,
+            onProgress: (event) => {
+                progress?.onProgress?.({
+                    kind: 'transcript-whisper-progress',
+                    url: progress.url,
+                    service: progress.service,
+                    processedDurationSeconds: event.processedDurationSeconds,
+                    totalDurationSeconds: event.totalDurationSeconds,
+                    partIndex: event.partIndex,
+                    parts: event.parts,
+                });
+            },
+        });
+        if (transcript.notes.length > 0)
+            notes.push(...transcript.notes);
+        return { text: transcript.text, provider: transcript.provider, error: transcript.error };
+    }
+    const tmpFile = (0, node_path_1.join)((0, node_os_1.tmpdir)(), `summarize-podcast-${(0, node_crypto_1.randomUUID)()}.bin`);
+    try {
+        const downloadedBytes = await downloadToFile(fetchImpl, url, tmpFile, {
+            totalBytes,
+            onProgress: (nextDownloadedBytes) => progress?.onProgress?.({
+                kind: 'transcript-media-download-progress',
+                url: progress.url,
+                service: progress.service,
+                downloadedBytes: nextDownloadedBytes,
+                totalBytes,
+            }),
+        });
+        progress?.onProgress?.({
+            kind: 'transcript-media-download-done',
+            url: progress.url,
+            service: progress.service,
+            downloadedBytes,
+            totalBytes,
+        });
+        const probedDurationSeconds = durationSecondsHint ?? (await (0, whisper_js_1.probeMediaDurationSecondsWithFfprobe)(tmpFile));
+        progress?.onProgress?.({
+            kind: 'transcript-whisper-start',
+            url: progress.url,
+            service: progress.service,
+            providerHint,
+            modelId,
+            totalDurationSeconds: probedDurationSeconds,
+            parts: null,
+        });
+        const transcript = await (0, whisper_js_1.transcribeMediaFileWithWhisper)({
+            filePath: tmpFile,
+            mediaType,
+            filename,
+            openaiApiKey,
+            falApiKey,
+            totalDurationSeconds: probedDurationSeconds,
+            onProgress: (event) => {
+                progress?.onProgress?.({
+                    kind: 'transcript-whisper-progress',
+                    url: progress.url,
+                    service: progress.service,
+                    processedDurationSeconds: event.processedDurationSeconds,
+                    totalDurationSeconds: event.totalDurationSeconds,
+                    partIndex: event.partIndex,
+                    parts: event.parts,
+                });
+            },
+        });
+        if (transcript.notes.length > 0)
+            notes.push(...transcript.notes);
+        return { text: transcript.text, provider: transcript.provider, error: transcript.error };
+    }
+    finally {
+        await node_fs_1.promises.unlink(tmpFile).catch(() => { });
+    }
+}
+async function probeRemoteMedia(fetchImpl, url) {
+    try {
+        const res = await fetchImpl(url, {
+            method: 'HEAD',
+            redirect: 'follow',
+            signal: AbortSignal.timeout(constants_js_1.TRANSCRIPTION_TIMEOUT_MS),
+        });
+        if (!res.ok)
+            throw new Error('head failed');
+        const contentLength = parseContentLength(res.headers.get('content-length'));
+        const mediaType = normalizeHeaderType(res.headers.get('content-type'));
+        const filename = filenameFromUrl(url);
+        return { contentLength, mediaType, filename };
+    }
+    catch {
+        return { contentLength: null, mediaType: null, filename: filenameFromUrl(url) };
+    }
+}
+async function downloadCappedBytes(fetchImpl, url, maxBytes, options) {
+    const res = await fetchImpl(url, {
+        redirect: 'follow',
+        headers: { Range: `bytes=0-${maxBytes - 1}` },
+        signal: AbortSignal.timeout(constants_js_1.TRANSCRIPTION_TIMEOUT_MS),
+    });
+    if (!res.ok) {
+        throw new Error(`Download failed (${res.status})`);
+    }
+    const body = res.body;
+    if (!body) {
+        const arrayBuffer = await res.arrayBuffer();
+        return new Uint8Array(arrayBuffer.slice(0, maxBytes));
+    }
+    const reader = body.getReader();
+    const chunks = [];
+    let total = 0;
+    let lastReported = 0;
+    try {
+        while (total < maxBytes) {
+            const { value, done } = await reader.read();
+            if (done)
+                break;
+            if (!value)
+                continue;
+            const remaining = maxBytes - total;
+            const next = value.byteLength > remaining ? value.slice(0, remaining) : value;
+            chunks.push(next);
+            total += next.byteLength;
+            if (total - lastReported >= 64 * 1024) {
+                lastReported = total;
+                options?.onProgress?.(total);
+            }
+            if (total >= maxBytes)
+                break;
+        }
+    }
+    finally {
+        await reader.cancel().catch(() => { });
+    }
+    options?.onProgress?.(total);
+    const out = new Uint8Array(total);
+    let offset = 0;
+    for (const chunk of chunks) {
+        out.set(chunk, offset);
+        offset += chunk.byteLength;
+    }
+    return out;
+}
+async function downloadToFile(fetchImpl, url, filePath, options) {
+    const res = await fetchImpl(url, {
+        redirect: 'follow',
+        signal: AbortSignal.timeout(constants_js_1.TRANSCRIPTION_TIMEOUT_MS),
+    });
+    if (!res.ok) {
+        throw new Error(`Download failed (${res.status})`);
+    }
+    const body = res.body;
+    if (!body) {
+        const bytes = new Uint8Array(await res.arrayBuffer());
+        await node_fs_1.promises.writeFile(filePath, bytes);
+        options?.onProgress?.(bytes.byteLength);
+        return bytes.byteLength;
+    }
+    const handle = await node_fs_1.promises.open(filePath, 'w');
+    let downloadedBytes = 0;
+    let lastReported = 0;
+    try {
+        const reader = body.getReader();
+        try {
+            while (true) {
+                const { value, done } = await reader.read();
+                if (done)
+                    break;
+                if (!value)
+                    continue;
+                await handle.write(value);
+                downloadedBytes += value.byteLength;
+                if (downloadedBytes - lastReported >= 128 * 1024) {
+                    lastReported = downloadedBytes;
+                    options?.onProgress?.(downloadedBytes);
+                }
+            }
+            options?.onProgress?.(downloadedBytes);
+        }
+        finally {
+            await reader.cancel().catch(() => { });
+        }
+    }
+    finally {
+        await handle.close().catch(() => { });
+    }
+    return downloadedBytes;
+}
+function normalizeHeaderType(value) {
+    if (!value)
+        return null;
+    const trimmed = value.trim();
+    if (!trimmed)
+        return null;
+    return trimmed.split(';')[0]?.trim().toLowerCase() ?? null;
+}
+function parseContentLength(value) {
+    if (!value)
+        return null;
+    const parsed = Number(value);
+    return Number.isFinite(parsed) && parsed > 0 ? Math.floor(parsed) : null;
+}
+function filenameFromUrl(url) {
+    try {
+        const parsed = new URL(url);
+        const base = parsed.pathname.split('/').pop() ?? '';
+        return base.trim().length > 0 ? base : null;
+    }
+    catch {
+        return null;
+    }
+}
+function formatBytes(bytes) {
+    const units = ['B', 'KB', 'MB', 'GB'];
+    let value = bytes;
+    let idx = 0;
+    while (value >= 1024 && idx < units.length - 1) {
+        value /= 1024;
+        idx += 1;
+    }
+    const decimals = value >= 10 || idx === 0 ? 0 : 1;
+    return `${value.toFixed(decimals)}${units[idx]}`;
+}
+//# sourceMappingURL=media.js.map

package/dist/cjs/content/transcript/providers/podcast/media.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"media.js","sourceRoot":"","sources":["../../../../../../src/content/transcript/providers/podcast/media.ts"],"names":[],"mappings":";;AA4BA,gDA+NC;AAED,4CAkBC;AAED,kDAmDC;AAED,wCA8CC;AAED,kDAKC;AAED,gDAIC;AAED,0CAQC;AAED,kCAUC;AAvZD,6CAAwC;AACxC,qCAAwC;AACxC,qCAAgC;AAChC,yCAAgC;AAChC,qEAQ6C;AAE7C,iDAAiF;AAc1E,KAAK,UAAU,kBAAkB,CAAC,EACvC,SAAS,EACT,GAAG,EACH,YAAY,EACZ,mBAAmB,EACnB,YAAY,EACZ,SAAS,EACT,KAAK,EACL,QAAQ,GAcT;IACC,MAAM,QAAQ,GAAG,MAAM,IAAA,8BAAiB,GAAE,CAAA;IAC1C,MAAM,YAAY,GAChB,CAAC,MAAM,IAAA,8BAAiB,GAAE,CAAC;QACzB,CAAC,CAAC,KAAK;QACP,CAAC,CAAC,YAAY,IAAI,SAAS;YACzB,CAAC,CAAC,aAAa;YACf,CAAC,CAAC,YAAY;gBACZ,CAAC,CAAC,QAAQ;gBACV,CAAC,CAAC,SAAS;oBACT,CAAC,CAAC,KAAK;oBACP,CAAC,CAAC,SAAS,CAAA;IAErB,MAAM,IAAI,GAAG,MAAM,gBAAgB,CAAC,SAAS,EAAE,GAAG,CAAC,CAAA;IACnD,IAAI,IAAI,CAAC,aAAa,KAAK,IAAI,IAAI,IAAI,CAAC,aAAa,GAAG,qCAAsB,EAAE,CAAC;QAC/E,MAAM,IAAI,KAAK,CACb,2BAA2B,WAAW,CAAC,IAAI,CAAC,aAAa,CAAC,eAAe,WAAW,CAAC,qCAAsB,CAAC,GAAG,CAChH,CAAA;IACH,CAAC;IAED,MAAM,SAAS,GAAG,IAAI,CAAC,SAAS,IAAI,0BAA0B,CAAA;IAC9D,MAAM,QAAQ,GAAG,IAAI,CAAC,QAAQ,IAAI,YAAY,CAAA;IAC9C,MAAM,UAAU,GAAG,IAAI,CAAC,aAAa,CAAA;IAErC,QAAQ,EAAE,UAAU,EAAE,CAAC;QACrB,IAAI,EAAE,iCAAiC;QACvC,GAAG,EAAE,QAAQ,CAAC,GAAG;QACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;QACzB,QAAQ,EAAE,GAAG;QACb,UAAU;KACX,CAAC,CAAA;IAEF,MAAM,OAAO,GACX,YAAY,KAAK,KAAK;QACpB,CAAC,CAAC,CAAC,CAAC,MAAM,IAAA,iDAAoC,GAAE,CAAC,IAAI,aAAa,CAAC;QACnE,CAAC,CAAC,YAAY,IAAI,SAAS;YACzB,CAAC,CAAC,0BAA0B;YAC5B,CAAC,CAAC,YAAY;gBACZ,CAAC,CAAC,WAAW;gBACb,CAAC,CAAC,SAAS;oBACT,CAAC,CAAC,eAAe;oBACjB,CAAC,CAAC,IAAI,CAAA;IAChB,IAAI,CAAC,QAAQ,EAAE,CAAC;QACd,MAAM,KAAK,GAAG,MAAM,mBAAmB,CAAC,SAAS,EAAE,GAAG,EAAE,oCAAuB,EAAE;YAC/E,UAAU;YACV,UAAU,EAAE,CAAC,eAAe,EAAE,EAAE,CAC9B,QAAQ,EAAE,UAAU,EAAE,CAAC;gBACrB,IAAI,EAAE,oCAAoC;gBAC1C,GAAG,EAAE,QAAQ,CAAC,GAAG;gBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;gBACzB,eAAe;gBACf,UAAU;aACX,CAAC;SACL,CAAC,CAAA;QACF,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,gCAAgC;YACtC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,eAAe,EAAE,KAAK,CAAC,UAAU;YACjC,UAAU;SACX,CAAC,CAAA;QACF,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,0BAA0B;YAChC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,YAAY;YACZ,OAAO;YACP,oBAAoB,EAAE,mBAAmB;YACzC,KAAK,EAAE,IAAI;SACZ,CAAC,CAAA;QACF,KAAK,CAAC,IAAI,CAAC,qBAAqB,WAAW,CAAC,KAAK,CAAC,UAAU,CAAC,8BAA8B,CAAC,CAAA;QAC5F,MAAM,UAAU,GAAG,MAAM,IAAA,uCAA0B,EAAC;YAClD,KAAK;YACL,SAAS;YACT,QAAQ;YACR,YAAY;YACZ,SAAS;YACT,oBAAoB,EAAE,mBAAmB;YACzC,UAAU,EAAE,CAAC,KAAK,EAAE,EAAE;gBACpB,QAAQ,EAAE,UAAU,EAAE,CAAC;oBACrB,IAAI,EAAE,6BAA6B;oBACnC,GAAG,EAAE,QAAQ,CAAC,GAAG;oBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;oBACzB,wBAAwB,EAAE,KAAK,CAAC,wBAAwB;oBACxD,oBAAoB,EAAE,KAAK,CAAC,oBAAoB;oBAChD,SAAS,EAAE,KAAK,CAAC,SAAS;oBAC1B,KAAK,EAAE,KAAK,CAAC,KAAK;iBACnB,CAAC,CAAA;YACJ,CAAC;SACF,CAAC,CAAA;QACF,IAAI,UAAU,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,GAAG,UAAU,CAAC,KAAK,CAAC,CAAA;QAChE,OAAO,EAAE,IAAI,EAAE,UAAU,CAAC,IAAI,EAAE,QAAQ,EAAE,UAAU,CAAC,QAAQ,EAAE,KAAK,EAAE,UAAU,CAAC,KAAK,EAAE,CAAA;IAC1F,CAAC;IAED,IAAI,IAAI,CAAC,aAAa,KAAK,IAAI,IAAI,IAAI,CAAC,aAAa,IAAI,oCAAuB,EAAE,CAAC;QACjF,MAAM,KAAK,GAAG,MAAM,mBAAmB,CAAC,SAAS,EAAE,GAAG,EAAE,oCAAuB,EAAE;YAC/E,UAAU;YACV,UAAU,EAAE,CAAC,eAAe,EAAE,EAAE,CAC9B,QAAQ,EAAE,UAAU,EAAE,CAAC;gBACrB,IAAI,EAAE,oCAAoC;gBAC1C,GAAG,EAAE,QAAQ,CAAC,GAAG;gBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;gBACzB,eAAe;gBACf,UAAU;aACX,CAAC;SACL,CAAC,CAAA;QACF,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,gCAAgC;YACtC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,eAAe,EAAE,KAAK,CAAC,UAAU;YACjC,UAAU;SACX,CAAC,CAAA;QACF,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,0BAA0B;YAChC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,YAAY;YACZ,OAAO;YACP,oBAAoB,EAAE,mBAAmB;YACzC,KAAK,EAAE,IAAI;SACZ,CAAC,CAAA;QACF,MAAM,UAAU,GAAG,MAAM,IAAA,uCAA0B,EAAC;YAClD,KAAK;YACL,SAAS;YACT,QAAQ;YACR,YAAY;YACZ,SAAS;YACT,oBAAoB,EAAE,mBAAmB;YACzC,UAAU,EAAE,CAAC,KAAK,EAAE,EAAE;gBACpB,QAAQ,EAAE,UAAU,EAAE,CAAC;oBACrB,IAAI,EAAE,6BAA6B;oBACnC,GAAG,EAAE,QAAQ,CAAC,GAAG;oBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;oBACzB,wBAAwB,EAAE,KAAK,CAAC,wBAAwB;oBACxD,oBAAoB,EAAE,KAAK,CAAC,oBAAoB;oBAChD,SAAS,EAAE,KAAK,CAAC,SAAS;oBAC1B,KAAK,EAAE,KAAK,CAAC,KAAK;iBACnB,CAAC,CAAA;YACJ,CAAC;SACF,CAAC,CAAA;QACF,IAAI,UAAU,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,GAAG,UAAU,CAAC,KAAK,CAAC,CAAA;QAChE,OAAO,EAAE,IAAI,EAAE,UAAU,CAAC,IAAI,EAAE,QAAQ,EAAE,UAAU,CAAC,QAAQ,EAAE,KAAK,EAAE,UAAU,CAAC,KAAK,EAAE,CAAA;IAC1F,CAAC;IAED,MAAM,OAAO,GAAG,IAAA,gBAAI,EAAC,IAAA,gBAAM,GAAE,EAAE,qBAAqB,IAAA,wBAAU,GAAE,MAAM,CAAC,CAAA;IACvE,IAAI,CAAC;QACH,MAAM,eAAe,GAAG,MAAM,cAAc,CAAC,SAAS,EAAE,GAAG,EAAE,OAAO,EAAE;YACpE,UAAU;YACV,UAAU,EAAE,CAAC,mBAAmB,EAAE,EAAE,CAClC,QAAQ,EAAE,UAAU,EAAE,CAAC;gBACrB,IAAI,EAAE,oCAAoC;gBAC1C,GAAG,EAAE,QAAQ,CAAC,GAAG;gBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;gBACzB,eAAe,EAAE,mBAAmB;gBACpC,UAAU;aACX,CAAC;SACL,CAAC,CAAA;QACF,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,gCAAgC;YACtC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,eAAe;YACf,UAAU;SACX,CAAC,CAAA;QAEF,MAAM,qBAAqB,GACzB,mBAAmB,IAAI,CAAC,MAAM,IAAA,iDAAoC,EAAC,OAAO,CAAC,CAAC,CAAA;QAC9E,QAAQ,EAAE,UAAU,EAAE,CAAC;YACrB,IAAI,EAAE,0BAA0B;YAChC,GAAG,EAAE,QAAQ,CAAC,GAAG;YACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;YACzB,YAAY;YACZ,OAAO;YACP,oBAAoB,EAAE,qBAAqB;YAC3C,KAAK,EAAE,IAAI;SACZ,CAAC,CAAA;QACF,MAAM,UAAU,GAAG,MAAM,IAAA,2CAA8B,EAAC;YACtD,QAAQ,EAAE,OAAO;YACjB,SAAS;YACT,QAAQ;YACR,YAAY;YACZ,SAAS;YACT,oBAAoB,EAAE,qBAAqB;YAC3C,UAAU,EAAE,CAAC,KAAK,EAAE,EAAE;gBACpB,QAAQ,EAAE,UAAU,EAAE,CAAC;oBACrB,IAAI,EAAE,6BAA6B;oBACnC,GAAG,EAAE,QAAQ,CAAC,GAAG;oBACjB,OAAO,EAAE,QAAQ,CAAC,OAAO;oBACzB,wBAAwB,EAAE,KAAK,CAAC,wBAAwB;oBACxD,oBAAoB,EAAE,KAAK,CAAC,oBAAoB;oBAChD,SAAS,EAAE,KAAK,CAAC,SAAS;oBAC1B,KAAK,EAAE,KAAK,CAAC,KAAK;iBACnB,CAAC,CAAA;YACJ,CAAC;SACF,CAAC,CAAA;QACF,IAAI,UAAU,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,GAAG,UAAU,CAAC,KAAK,CAAC,CAAA;QAChE,OAAO,EAAE,IAAI,EAAE,UAAU,CAAC,IAAI,EAAE,QAAQ,EAAE,UAAU,CAAC,QAAQ,EAAE,KAAK,EAAE,UAAU,CAAC,KAAK,EAAE,CAAA;IAC1F,CAAC;YAAS,CAAC;QACT,MAAM,kBAAE,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,KAAK,CAAC,GAAG,EAAE,GAAE,CAAC,CAAC,CAAA;IAC1C,CAAC;AACH,CAAC;AAEM,KAAK,UAAU,gBAAgB,CACpC,SAAuB,EACvB,GAAW;IAEX,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,SAAS,CAAC,GAAG,EAAE;YAC/B,MAAM,EAAE,MAAM;YACd,QAAQ,EAAE,QAAQ;YAClB,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,uCAAwB,CAAC;SACtD,CAAC,CAAA;QACF,IAAI,CAAC,GAAG,CAAC,EAAE;YAAE,MAAM,IAAI,KAAK,CAAC,aAAa,CAAC,CAAA;QAC3C,MAAM,aAAa,GAAG,kBAAkB,CAAC,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,gBAAgB,CAAC,CAAC,CAAA;QAC3E,MAAM,SAAS,GAAG,mBAAmB,CAAC,GAAG,CAAC,OAAO,CAAC,GAAG,CAAC,cAAc,CAAC,CAAC,CAAA;QACtE,MAAM,QAAQ,GAAG,eAAe,CAAC,GAAG,CAAC,CAAA;QACrC,OAAO,EAAE,aAAa,EAAE,SAAS,EAAE,QAAQ,EAAE,CAAA;IAC/C,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,EAAE,aAAa,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,QAAQ,EAAE,eAAe,CAAC,GAAG,CAAC,EAAE,CAAA;IACjF,CAAC;AACH,CAAC;AAEM,KAAK,UAAU,mBAAmB,CACvC,SAAuB,EACvB,GAAW,EACX,QAAgB,EAChB,OAAgG;IAEhG,MAAM,GAAG,GAAG,MAAM,SAAS,CAAC,GAAG,EAAE;QAC/B,QAAQ,EAAE,QAAQ;QAClB,OAAO,EAAE,EAAE,KAAK,EAAE,WAAW,QAAQ,GAAG,CAAC,EAAE,EAAE;QAC7C,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,uCAAwB,CAAC;KACtD,CAAC,CAAA;IACF,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;QACZ,MAAM,IAAI,KAAK,CAAC,oBAAoB,GAAG,CAAC,MAAM,GAAG,CAAC,CAAA;IACpD,CAAC;IACD,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAA;IACrB,IAAI,CAAC,IAAI,EAAE,CAAC;QACV,MAAM,WAAW,GAAG,MAAM,GAAG,CAAC,WAAW,EAAE,CAAA;QAC3C,OAAO,IAAI,UAAU,CAAC,WAAW,CAAC,KAAK,CAAC,CAAC,EAAE,QAAQ,CAAC,CAAC,CAAA;IACvD,CAAC;IAED,MAAM,MAAM,GAAG,IAAI,CAAC,SAAS,EAAE,CAAA;IAC/B,MAAM,MAAM,GAAiB,EAAE,CAAA;IAC/B,IAAI,KAAK,GAAG,CAAC,CAAA;IACb,IAAI,YAAY,GAAG,CAAC,CAAA;IACpB,IAAI,CAAC;QACH,OAAO,KAAK,GAAG,QAAQ,EAAE,CAAC;YACxB,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,GAAG,MAAM,MAAM,CAAC,IAAI,EAAE,CAAA;YAC3C,IAAI,IAAI;gBAAE,MAAK;YACf,IAAI,CAAC,KAAK;gBAAE,SAAQ;YACpB,MAAM,SAAS,GAAG,QAAQ,GAAG,KAAK,CAAA;YAClC,MAAM,IAAI,GAAG,KAAK,CAAC,UAAU,GAAG,SAAS,CAAC,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC,CAAC,CAAC,KAAK,CAAA;YAC7E,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;YACjB,KAAK,IAAI,IAAI,CAAC,UAAU,CAAA;YACxB,IAAI,KAAK,GAAG,YAAY,IAAI,EAAE,GAAG,IAAI,EAAE,CAAC;gBACtC,YAAY,GAAG,KAAK,CAAA;gBACpB,OAAO,EAAE,UAAU,EAAE,CAAC,KAAK,CAAC,CAAA;YAC9B,CAAC;YACD,IAAI,KAAK,IAAI,QAAQ;gBAAE,MAAK;QAC9B,CAAC;IACH,CAAC;YAAS,CAAC;QACT,MAAM,MAAM,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,GAAE,CAAC,CAAC,CAAA;IACvC,CAAC;IACD,OAAO,EAAE,UAAU,EAAE,CAAC,KAAK,CAAC,CAAA;IAE5B,MAAM,GAAG,GAAG,IAAI,UAAU,CAAC,KAAK,CAAC,CAAA;IACjC,IAAI,MAAM,GAAG,CAAC,CAAA;IACd,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,GAAG,CAAC,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAA;QACtB,MAAM,IAAI,KAAK,CAAC,UAAU,CAAA;IAC5B,CAAC;IACD,OAAO,GAAG,CAAA;AACZ,CAAC;AAEM,KAAK,UAAU,cAAc,CAClC,SAAuB,EACvB,GAAW,EACX,QAAgB,EAChB,OAAgG;IAEhG,MAAM,GAAG,GAAG,MAAM,SAAS,CAAC,GAAG,EAAE;QAC/B,QAAQ,EAAE,QAAQ;QAClB,MAAM,EAAE,WAAW,CAAC,OAAO,CAAC,uCAAwB,CAAC;KACtD,CAAC,CAAA;IACF,IAAI,CAAC,GAAG,CAAC,EAAE,EAAE,CAAC;QACZ,MAAM,IAAI,KAAK,CAAC,oBAAoB,GAAG,CAAC,MAAM,GAAG,CAAC,CAAA;IACpD,CAAC;IACD,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAA;IACrB,IAAI,CAAC,IAAI,EAAE,CAAC;QACV,MAAM,KAAK,GAAG,IAAI,UAAU,CAAC,MAAM,GAAG,CAAC,WAAW,EAAE,CAAC,CAAA;QACrD,MAAM,kBAAE,CAAC,SAAS,CAAC,QAAQ,EAAE,KAAK,CAAC,CAAA;QACnC,OAAO,EAAE,UAAU,EAAE,CAAC,KAAK,CAAC,UAAU,CAAC,CAAA;QACvC,OAAO,KAAK,CAAC,UAAU,CAAA;IACzB,CAAC;IAED,MAAM,MAAM,GAAG,MAAM,kBAAE,CAAC,IAAI,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAA;IAC3C,IAAI,eAAe,GAAG,CAAC,CAAA;IACvB,IAAI,YAAY,GAAG,CAAC,CAAA;IACpB,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAI,CAAC,SAAS,EAAE,CAAA;QAC/B,IAAI,CAAC;YACH,OAAO,IAAI,EAAE,CAAC;gBACZ,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,GAAG,MAAM,MAAM,CAAC,IAAI,EAAE,CAAA;gBAC3C,IAAI,IAAI;oBAAE,MAAK;gBACf,IAAI,CAAC,KAAK;oBAAE,SAAQ;gBACpB,MAAM,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,CAAA;gBACzB,eAAe,IAAI,KAAK,CAAC,UAAU,CAAA;gBACnC,IAAI,eAAe,GAAG,YAAY,IAAI,GAAG,GAAG,IAAI,EAAE,CAAC;oBACjD,YAAY,GAAG,eAAe,CAAA;oBAC9B,OAAO,EAAE,UAAU,EAAE,CAAC,eAAe,CAAC,CAAA;gBACxC,CAAC;YACH,CAAC;YACD,OAAO,EAAE,UAAU,EAAE,CAAC,eAAe,CAAC,CAAA;QACxC,CAAC;gBAAS,CAAC;YACT,MAAM,MAAM,CAAC,MAAM,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,GAAE,CAAC,CAAC,CAAA;QACvC,CAAC;IACH,CAAC;YAAS,CAAC;QACT,MAAM,MAAM,CAAC,KAAK,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,GAAE,CAAC,CAAC,CAAA;IACtC,CAAC;IACD,OAAO,eAAe,CAAA;AACxB,CAAC;AAED,SAAgB,mBAAmB,CAAC,KAAoB;IACtD,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAA;IACvB,MAAM,OAAO,GAAG,KAAK,CAAC,IAAI,EAAE,CAAA;IAC5B,IAAI,CAAC,OAAO;QAAE,OAAO,IAAI,CAAA;IACzB,OAAO,OAAO,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,WAAW,EAAE,IAAI,IAAI,CAAA;AAC5D,CAAC;AAED,SAAgB,kBAAkB,CAAC,KAAoB;IACrD,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAA;IACvB,MAAM,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,CAAA;IAC5B,OAAO,MAAM,CAAC,QAAQ,CAAC,MAAM,CAAC,IAAI,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,IAAI,CAAA;AAC1E,CAAC;AAED,SAAgB,eAAe,CAAC,GAAW;IACzC,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAA;QAC3B,MAAM,IAAI,GAAG,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,EAAE,IAAI,EAAE,CAAA;QACnD,OAAO,IAAI,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAA;IAC7C,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAA;IACb,CAAC;AACH,CAAC;AAED,SAAgB,WAAW,CAAC,KAAa;IACvC,MAAM,KAAK,GAAG,CAAC,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC,CAAA;IACrC,IAAI,KAAK,GAAG,KAAK,CAAA;IACjB,IAAI,GAAG,GAAG,CAAC,CAAA;IACX,OAAO,KAAK,IAAI,IAAI,IAAI,GAAG,GAAG,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC/C,KAAK,IAAI,IAAI,CAAA;QACb,GAAG,IAAI,CAAC,CAAA;IACV,CAAC;IACD,MAAM,QAAQ,GAAG,KAAK,IAAI,EAAE,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAA;IACjD,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,QAAQ,CAAC,GAAG,KAAK,CAAC,GAAG,CAAC,EAAE,CAAA;AAClD,CAAC"}

package/dist/cjs/content/transcript/providers/podcast/results.js

@@ -0,0 +1,32 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.joinNotes = joinNotes;
+exports.buildWhisperResult = buildWhisperResult;
+function joinNotes(notes) {
+    return notes.length > 0 ? notes.join('; ') : null;
+}
+function buildWhisperResult({ attemptedProviders, notes, outcome, metadata, includeProviderOnFailure = false, }) {
+    if (outcome.text) {
+        return {
+            text: outcome.text,
+            source: 'whisper',
+            attemptedProviders,
+            notes: joinNotes(notes),
+            metadata: {
+                ...metadata,
+                transcriptionProvider: outcome.provider,
+            },
+        };
+    }
+    const failureMetadata = includeProviderOnFailure && outcome.provider
+        ? { ...metadata, transcriptionProvider: outcome.provider }
+        : metadata;
+    return {
+        text: null,
+        source: null,
+        attemptedProviders,
+        notes: outcome.error?.message ?? null,
+        metadata: failureMetadata,
+    };
+}
+//# sourceMappingURL=results.js.map

package/dist/cjs/content/transcript/providers/podcast/results.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"results.js","sourceRoot":"","sources":["../../../../../../src/content/transcript/providers/podcast/results.ts"],"names":[],"mappings":";;AAGA,8BAEC;AAED,gDAsCC;AA1CD,SAAgB,SAAS,CAAC,KAAe;IACvC,OAAO,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,CAAA;AACnD,CAAC;AAED,SAAgB,kBAAkB,CAAC,EACjC,kBAAkB,EAClB,KAAK,EACL,OAAO,EACP,QAAQ,EACR,wBAAwB,GAAG,KAAK,GAOjC;IACC,IAAI,OAAO,CAAC,IAAI,EAAE,CAAC;QACjB,OAAO;YACL,IAAI,EAAE,OAAO,CAAC,IAAI;YAClB,MAAM,EAAE,SAAS;YACjB,kBAAkB;YAClB,KAAK,EAAE,SAAS,CAAC,KAAK,CAAC;YACvB,QAAQ,EAAE;gBACR,GAAG,QAAQ;gBACX,qBAAqB,EAAE,OAAO,CAAC,QAAQ;aACxC;SACF,CAAA;IACH,CAAC;IAED,MAAM,eAAe,GACnB,wBAAwB,IAAI,OAAO,CAAC,QAAQ;QAC1C,CAAC,CAAC,EAAE,GAAG,QAAQ,EAAE,qBAAqB,EAAE,OAAO,CAAC,QAAQ,EAAE;QAC1D,CAAC,CAAC,QAAQ,CAAA;IAEd,OAAO;QACL,IAAI,EAAE,IAAI;QACV,MAAM,EAAE,IAAI;QACZ,kBAAkB;QAClB,KAAK,EAAE,OAAO,CAAC,KAAK,EAAE,OAAO,IAAI,IAAI;QACrC,QAAQ,EAAE,eAAe;KAC1B,CAAA;AACH,CAAC"}

package/dist/cjs/content/transcript/providers/podcast/rss.js

@@ -0,0 +1,262 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.looksLikeRssOrAtomFeed = looksLikeRssOrAtomFeed;
+exports.extractEnclosureFromFeed = extractEnclosureFromFeed;
+exports.extractEnclosureForEpisode = extractEnclosureForEpisode;
+exports.extractItemDurationSeconds = extractItemDurationSeconds;
+exports.decodeXmlEntities = decodeXmlEntities;
+exports.normalizeLooseTitle = normalizeLooseTitle;
+exports.tryFetchTranscriptFromFeedXml = tryFetchTranscriptFromFeedXml;
+const constants_js_1 = require("./constants.js");
+function looksLikeRssOrAtomFeed(xml) {
+    const head = xml.slice(0, 4096).trimStart().toLowerCase();
+    if (head.startsWith('<rss') || head.includes('<rss'))
+        return true;
+    if (head.startsWith('<?xml') && (head.includes('<rss') || head.includes('<feed')))
+        return true;
+    if (head.startsWith('<feed') || head.includes('<feed'))
+        return true;
+    return false;
+}
+function extractEnclosureFromFeed(xml) {
+    const items = xml.match(/<item\b[\s\S]*?<\/item>/gi) ?? [];
+    for (const item of items) {
+        const enclosureUrl = extractEnclosureUrlFromItem(item);
+        if (!enclosureUrl)
+            continue;
+        return { enclosureUrl, durationSeconds: extractItemDurationSeconds(item) };
+    }
+    const enclosureMatch = xml.match(/<enclosure\b[^>]*\burl\s*=\s*(['"])([^'"]+)\1/i);
+    if (enclosureMatch?.[2]) {
+        return { enclosureUrl: enclosureMatch[2], durationSeconds: extractItemDurationSeconds(xml) };
+    }
+    const atomMatch = xml.match(/<link\b[^>]*\brel\s*=\s*(['"])enclosure\1[^>]*\bhref\s*=\s*(['"])([^'"]+)\2/i);
+    if (atomMatch?.[3]) {
+        return { enclosureUrl: atomMatch[3], durationSeconds: extractItemDurationSeconds(xml) };
+    }
+    return null;
+}
+function extractEnclosureForEpisode(feedXml, episodeTitle) {
+    const normalizedTarget = normalizeLooseTitle(episodeTitle);
+    const items = feedXml.match(/<item\b[\s\S]*?<\/item>/gi) ?? [];
+    for (const item of items) {
+        const title = extractItemTitle(item);
+        if (!title)
+            continue;
+        if (normalizeLooseTitle(title) !== normalizedTarget)
+            continue;
+        const enclosureUrl = extractEnclosureUrlFromItem(item);
+        if (!enclosureUrl)
+            continue;
+        return { enclosureUrl, durationSeconds: extractItemDurationSeconds(item) };
+    }
+    return null;
+}
+function extractItemDurationSeconds(itemXml) {
+    const match = itemXml.match(/<itunes:duration>([\s\S]*?)<\/itunes:duration>/i);
+    if (!match?.[1])
+        return null;
+    const raw = match[1]
+        .replaceAll(/<!\[CDATA\[/gi, '')
+        .replaceAll(/\]\]>/g, '')
+        .trim();
+    if (!raw)
+        return null;
+    // common forms: "HH:MM:SS", "MM:SS", "SS"
+    if (/^\d+$/.test(raw)) {
+        const seconds = Number(raw);
+        return Number.isFinite(seconds) && seconds > 0 ? seconds : null;
+    }
+    const parts = raw
+        .split(':')
+        .map((value) => value.trim())
+        .filter(Boolean);
+    if (parts.length < 2 || parts.length > 3)
+        return null;
+    const nums = parts.map((value) => Number(value));
+    if (nums.some((n) => !Number.isFinite(n) || n < 0))
+        return null;
+    const seconds = (() => {
+        if (nums.length === 3) {
+            const [hours, minutes, secondsRaw] = nums;
+            if (hours === undefined || minutes === undefined || secondsRaw === undefined)
+                return null;
+            return Math.round(hours * 3600 + minutes * 60 + secondsRaw);
+        }
+        const [minutes, secondsRaw] = nums;
+        if (minutes === undefined || secondsRaw === undefined)
+            return null;
+        return Math.round(minutes * 60 + secondsRaw);
+    })();
+    if (seconds === null)
+        return null;
+    return seconds > 0 ? seconds : null;
+}
+function decodeXmlEntities(value) {
+    return value
+        .replaceAll(/&amp;/gi, '&')
+        .replaceAll(/&#38;/g, '&')
+        .replaceAll(/&lt;/gi, '<')
+        .replaceAll(/&gt;/gi, '>')
+        .replaceAll(/&quot;/gi, '"')
+        .replaceAll(/&apos;/gi, "'");
+}
+function normalizeLooseTitle(value) {
+    return value
+        .toLowerCase()
+        .normalize('NFKD')
+        .replaceAll(/\p{Diacritic}+/gu, '')
+        .replaceAll(/[^a-z0-9]+/g, ' ')
+        .trim();
+}
+async function tryFetchTranscriptFromFeedXml({ fetchImpl, feedXml, episodeTitle, notes, }) {
+    const items = feedXml.match(/<item\b[\s\S]*?<\/item>/gi) ?? [];
+    const normalizedTarget = episodeTitle ? normalizeLooseTitle(episodeTitle) : null;
+    for (const item of items) {
+        if (normalizedTarget) {
+            const title = extractItemTitle(item);
+            if (!title || normalizeLooseTitle(title) !== normalizedTarget)
+                continue;
+        }
+        const candidates = extractPodcastTranscriptCandidatesFromItem(item);
+        const preferred = selectPreferredTranscriptCandidate(candidates);
+        if (!preferred) {
+            if (normalizedTarget)
+                return null;
+            continue;
+        }
+        const transcriptUrl = decodeXmlEntities(preferred.url);
+        try {
+            const res = await fetchImpl(transcriptUrl, {
+                redirect: 'follow',
+                signal: AbortSignal.timeout(constants_js_1.TRANSCRIPTION_TIMEOUT_MS),
+                headers: { accept: 'text/vtt,text/plain,application/json;q=0.9,*/*;q=0.8' },
+            });
+            if (!res.ok)
+                throw new Error(`transcript fetch failed (${res.status})`);
+            const contentType = res.headers.get('content-type')?.toLowerCase().split(';')[0]?.trim() ?? null;
+            const hintedType = preferred.type?.toLowerCase().split(';')[0]?.trim() ?? null;
+            const effectiveType = hintedType ?? contentType;
+            const body = await res.text();
+            const text = (() => {
+                if (effectiveType === 'application/json' || transcriptUrl.toLowerCase().endsWith('.json')) {
+                    try {
+                        return jsonTranscriptToPlainText(JSON.parse(body));
+                    }
+                    catch {
+                        return null;
+                    }
+                }
+                if (effectiveType === 'text/vtt' || transcriptUrl.toLowerCase().endsWith('.vtt')) {
+                    const plain = vttToPlainText(body);
+                    return plain.length > 0 ? plain : null;
+                }
+                const plain = body.trim();
+                return plain.length > 0 ? plain : null;
+            })();
+            if (!text) {
+                if (normalizedTarget)
+                    return null;
+                continue;
+            }
+            notes.push('Used RSS <podcast:transcript> (skipped Whisper)');
+            return { text, transcriptUrl, transcriptType: effectiveType };
+        }
+        catch (error) {
+            if (normalizedTarget) {
+                notes.push(`RSS <podcast:transcript> fetch failed: ${error instanceof Error ? error.message : String(error)}`);
+                return null;
+            }
+        }
+    }
+    return null;
+}
+function extractEnclosureUrlFromItem(xml) {
+    const enclosureMatch = xml.match(/<enclosure\b[^>]*\burl\s*=\s*(['"])([^'"]+)\1/i);
+    if (enclosureMatch?.[2])
+        return enclosureMatch[2];
+    const atomMatch = xml.match(/<link\b[^>]*\brel\s*=\s*(['"])enclosure\1[^>]*\bhref\s*=\s*(['"])([^'"]+)\2/i);
+    if (atomMatch?.[3])
+        return atomMatch[3];
+    return null;
+}
+function extractItemTitle(itemXml) {
+    const match = itemXml.match(/<title>([\s\S]*?)<\/title>/i);
+    if (!match?.[1])
+        return null;
+    const raw = match[1]
+        .replaceAll(/<!\[CDATA\[/gi, '')
+        .replaceAll(/\]\]>/g, '')
+        .trim();
+    return raw.length > 0 ? raw : null;
+}
+function extractPodcastTranscriptCandidatesFromItem(itemXml) {
+    const matches = itemXml.matchAll(/<podcast:transcript\b[^>]*\burl\s*=\s*(['"])([^'"]+)\1[^>]*>/gi);
+    const results = [];
+    for (const match of matches) {
+        const tag = match[0];
+        const url = match[2]?.trim();
+        if (!url)
+            continue;
+        const type = tag.match(/\btype\s*=\s*(['"])([^'"]+)\1/i)?.[2]?.trim() ?? null;
+        results.push({ url, type });
+    }
+    return results;
+}
+function selectPreferredTranscriptCandidate(candidates) {
+    if (candidates.length === 0)
+        return null;
+    const normalized = candidates.map((c) => ({
+        ...c,
+        type: c.type?.toLowerCase().split(';')[0]?.trim() ?? null,
+    }));
+    const json = normalized.find((c) => c.type === 'application/json' || c.url.toLowerCase().endsWith('.json'));
+    if (json)
+        return json;
+    const vtt = normalized.find((c) => c.type === 'text/vtt' || c.url.toLowerCase().endsWith('.vtt'));
+    if (vtt)
+        return vtt;
+    return normalized[0] ?? null;
+}
+function vttToPlainText(raw) {
+    const lines = raw
+        .replace(/\r\n/g, '\n')
+        .split('\n')
+        .map((line) => line.trim())
+        .filter((line) => line.length > 0)
+        .filter((line) => line.toUpperCase() !== 'WEBVTT')
+        .filter((line) => !/^\d{2}:\d{2}:\d{2}\.\d{3}\s+-->\s+\d{2}:\d{2}:\d{2}\.\d{3}/.test(line))
+        .filter((line) => !/^\d+$/.test(line))
+        .filter((line) => !/^(NOTE|STYLE|REGION)\b/i.test(line));
+    return lines.join('\n').trim();
+}
+function jsonTranscriptToPlainText(payload) {
+    if (Array.isArray(payload)) {
+        const parts = payload
+            .map((row) => (row && typeof row === 'object' ? row.text : null))
+            .filter((t) => typeof t === 'string')
+            .map((t) => t.trim())
+            .filter(Boolean);
+        const text = parts.join('\n').trim();
+        return text.length > 0 ? text : null;
+    }
+    if (payload && typeof payload === 'object') {
+        const record = payload;
+        if (typeof record.transcript === 'string' && record.transcript.trim())
+            return record.transcript.trim();
+        if (typeof record.text === 'string' && record.text.trim())
+            return record.text.trim();
+        const segments = record.segments;
+        if (Array.isArray(segments)) {
+            const parts = segments
+                .map((row) => row && typeof row === 'object' ? row.text : null)
+                .filter((t) => typeof t === 'string')
+                .map((t) => t.trim())
+                .filter(Boolean);
+            const text = parts.join('\n').trim();
+            return text.length > 0 ? text : null;
+        }
+    }
+    return null;
+}
+//# sourceMappingURL=rss.js.map