@storyteller-platform/align 0.0.1

package/dist/process/ranges.js

@@ -0,0 +1,137 @@
+import {
+  __callDispose,
+  __using
+} from "../chunk-BIEQXUOY.js";
+import { randomUUID } from "node:crypto";
+import { mkdir, rm } from "node:fs/promises";
+import { tmpdir } from "node:os";
+import { basename, dirname, extname, join } from "node:path";
+import { detectVoiceActivity } from "@storyteller-platform/ghost-story/vad";
+import { splitFile } from "../common/ffmpeg.js";
+async function getSafeChapterRanges(input, duration, chapters, maxSeconds, signal, logger) {
+  if (!chapters.length) {
+    logger == null ? void 0 : logger.info(
+      `Track is longer than ${maxSeconds / 60} minutes (${duration / 60}m); using VAD to determine safe split points.`
+    );
+    const ranges2 = await getSafeRanges(input, duration, maxSeconds, 0, signal);
+    return ranges2.map((r) => ({ filepath: input, ...r }));
+  }
+  const initialRanges = chapters.map((chapter, index) => {
+    const next = chapters[index + 1];
+    if (!next)
+      return {
+        filepath: chapter.filename,
+        start: chapter.start ?? 0,
+        end: duration
+      };
+    const end = next.filename === chapter.filename ? next.start ?? duration : duration;
+    return {
+      filepath: chapter.filename,
+      start: chapter.start ?? 0,
+      end
+    };
+  });
+  const ranges = [];
+  for (const range of initialRanges) {
+    const chapterDuration = range.end - range.start;
+    if (chapterDuration <= maxSeconds) {
+      ranges.push(range);
+      continue;
+    }
+    logger == null ? void 0 : logger.info(
+      `Chapter is longer than ${maxSeconds / 60} minutes (${duration / 60}m); using VAD to determine safe split points.`
+    );
+    const chapterRanges = await getSafeRanges(
+      range.filepath,
+      chapterDuration,
+      maxSeconds,
+      range.start,
+      signal,
+      logger
+    );
+    ranges.push(
+      ...chapterRanges.map((r) => ({ filepath: range.filepath, ...r }))
+    );
+  }
+  return ranges;
+}
+async function getSafeRanges(input, duration, maxSeconds, start = 0, signal, logger) {
+  var _stack = [];
+  try {
+    const ext = extname(input);
+    const rawFilename = basename(input, ext);
+    const tmpDir = join(tmpdir(), `storyteller-align-silence-${randomUUID()}`);
+    const stack = __using(_stack, new AsyncDisposableStack(), true);
+    stack.defer(async () => {
+      await rm(tmpDir, { recursive: true, force: true });
+    });
+    const ranges = [{ start, end: duration + start }];
+    for (let i = 0; i + 1 < duration / maxSeconds; i++) {
+      if (signal == null ? void 0 : signal.aborted) throw new Error("Aborted");
+      const tmpFilepath = join(tmpDir, i.toString(), `${rawFilename}.wav`);
+      await mkdir(dirname(tmpFilepath), { recursive: true });
+      const approxCutPoint = start + maxSeconds * (i + 1);
+      const searchStart = approxCutPoint - 120;
+      const searchEnd = approxCutPoint;
+      await splitFile(
+        input,
+        tmpFilepath,
+        searchStart,
+        searchEnd,
+        {},
+        signal,
+        logger
+      );
+      const vadTimeline = await detectVoiceActivity(tmpFilepath, {
+        engine: "active-gate-og"
+      });
+      if (vadTimeline.length === 0) {
+        break;
+      }
+      const silenceTimeline = vadTimeline.reduce((acc, entry) => {
+        const lastEntry = acc.at(-1);
+        if (!lastEntry) {
+          return [
+            ...acc,
+            {
+              start: entry.endTime + searchStart,
+              end: entry.endTime + searchStart
+            }
+          ];
+        }
+        return [
+          ...acc.slice(0, -1),
+          {
+            ...lastEntry,
+            end: entry.startTime + searchStart
+          },
+          {
+            start: entry.endTime + searchStart,
+            end: entry.endTime + searchStart
+          }
+        ];
+      }, []);
+      const nearestLikelySentenceBreak = silenceTimeline.reduce((acc, entry) => {
+        const currLength = acc.end - acc.start;
+        const entryLength = entry.end - entry.start;
+        return currLength > entryLength ? acc : entry;
+      });
+      const lastRange = ranges.at(-1);
+      lastRange.end = nearestLikelySentenceBreak.start;
+      ranges.push({
+        start: nearestLikelySentenceBreak.end,
+        end: duration + start
+      });
+    }
+    return ranges;
+  } catch (_) {
+    var _error = _, _hasError = true;
+  } finally {
+    var _promise = __callDispose(_stack, _error, _hasError);
+    _promise && await _promise;
+  }
+}
+export {
+  getSafeChapterRanges,
+  getSafeRanges
+};

package/dist/transcribe/parse.cjs

@@ -0,0 +1,149 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var parse_exports = {};
+__export(parse_exports, {
+  transcribeCommand: () => transcribeCommand,
+  transcribeParser: () => transcribeParser
+});
+module.exports = __toCommonJS(parse_exports);
+var import_core = require("@optique/core");
+var import_valueparser = require("@optique/core/valueparser");
+var import_valueparser2 = require("@optique/run/valueparser");
+var import_ghost_story = require("@storyteller-platform/ghost-story");
+var import_parse = require("../common/parse.cjs");
+const transcribeParser = (0, import_core.or)(
+  (0, import_core.object)("whisper.cpp", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["whisper.cpp"], { metavar: "whisper.cpp" })
+    ),
+    model: (0, import_core.withDefault)(
+      (0, import_core.option)("--model", "-m", (0, import_core.choice)(import_ghost_story.WHISPER_MODELS, { metavar: "MODEL" }), {
+        description: import_core.message`The whisper model to use`
+      }),
+      "tiny.en"
+    ),
+    threads: (0, import_core.withDefault)((0, import_core.option)("--threads", (0, import_core.integer)()), 4),
+    processors: (0, import_core.withDefault)(
+      (0, import_core.option)("--processors", (0, import_core.integer)(), {
+        description: import_core.message`The number of processors to use (values greater than 1 may affect timing accuracy)`
+      }),
+      1
+    ),
+    cpuOverride: (0, import_core.optional)(
+      (0, import_core.option)(
+        "--cpu-override",
+        (0, import_core.choice)(["blas", "cpu"], { metavar: "CPU_ENGINE" }),
+        {
+          description: import_core.message`When provided, will use this whisper variant even if another is available`
+        }
+      )
+    )
+  }),
+  (0, import_core.object)("whisper-server", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["whisper-server"], { metavar: "whisper-server" })
+    ),
+    whisperServerUrl: (0, import_core.map)(
+      (0, import_core.option)("--whisper-server-url", (0, import_valueparser.url)()),
+      (url2) => url2.toString()
+    ),
+    whisperServerApiKey: (0, import_core.optional)((0, import_core.option)("--whisper-server-api-key", (0, import_valueparser.string)()))
+  }),
+  (0, import_core.object)("openai-cloud", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["openai-cloud"], { metavar: "openai-cloud" })
+    ),
+    openaiModelName: (0, import_core.optional)((0, import_core.option)("--openai-model", (0, import_valueparser.string)())),
+    openAiApiKey: (0, import_core.optional)((0, import_core.option)("--openai-api-key", (0, import_valueparser.string)())),
+    openAiOrganization: (0, import_core.optional)((0, import_core.option)("--openai-organization", (0, import_valueparser.string)())),
+    openAiBaseUrl: (0, import_core.optional)(
+      (0, import_core.map)((0, import_core.option)("--openai-base-url", (0, import_valueparser.url)()), (url2) => url2.toString())
+    )
+  }),
+  (0, import_core.object)("google-cloud", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["google-cloud"], { metavar: "google-cloud" })
+    ),
+    googleCloudApiKey: (0, import_core.option)("--google-cloud-api-key", (0, import_valueparser.string)())
+  }),
+  (0, import_core.object)("microsoft-azure", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["microsoft-azure"], { metavar: "microsoft-azure" })
+    ),
+    azureServiceRegion: (0, import_core.option)("--azure-service-region", (0, import_valueparser.string)()),
+    azureSubscriptionKey: (0, import_core.option)("--azure-subscription-key", (0, import_valueparser.string)())
+  }),
+  (0, import_core.object)("amazon-transcribe", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["amazon-transcribe"], { metavar: "amazon-transcribe" })
+    ),
+    amazonTranscribeRegion: (0, import_core.option)("--amazon-transcribe-region", (0, import_valueparser.string)()),
+    amazonTranscribeAccessKeyId: (0, import_core.option)(
+      "--amazon-transcribe-access-key-id",
+      (0, import_valueparser.string)()
+    ),
+    amazonTranscribeSecretAccessKey: (0, import_core.option)(
+      "--amazon-transcribe-secret-access-key",
+      (0, import_valueparser.string)()
+    )
+  }),
+  (0, import_core.object)("deepgram", {
+    engine: (0, import_core.option)(
+      "--engine",
+      "-e",
+      (0, import_core.choice)(["deepgram"], { metavar: "deepgram" })
+    ),
+    deepgramApiKey: (0, import_core.option)("--deepgram-api-key", (0, import_valueparser.string)()),
+    deepgramModel: (0, import_core.withDefault)((0, import_core.option)("--deepgram-model", (0, import_valueparser.string)()), "nova-3")
+  })
+);
+const transcribeCommand = (0, import_core.command)(
+  "transcribe",
+  (0, import_core.merge)(
+    (0, import_core.object)({
+      action: (0, import_core.constant)("transcribe"),
+      input: (0, import_core.argument)(
+        (0, import_valueparser2.path)({ type: "directory", mustExist: true, metavar: "INPUT" })
+      ),
+      output: (0, import_core.argument)((0, import_valueparser2.path)({ metavar: "OUTPUT", type: "directory" }))
+    }),
+    import_parse.parallelismParser,
+    import_parse.languageParser,
+    transcribeParser,
+    import_parse.loggingParser
+  ),
+  { description: import_core.message`Transcribe a directory of audiobook files.` }
+);
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  transcribeCommand,
+  transcribeParser
+});

package/dist/transcribe/parse.d.cts

@@ -0,0 +1,114 @@
+import * as _optique_core from '@optique/core';
+
+declare const transcribeParser: _optique_core.Parser<"sync", {
+    readonly engine: "whisper.cpp";
+    readonly model: "tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0";
+    readonly threads: number;
+    readonly processors: number;
+    readonly cpuOverride: "blas" | "cpu" | undefined;
+} | {
+    readonly engine: "whisper-server";
+    readonly whisperServerUrl: string;
+    readonly whisperServerApiKey: string | undefined;
+} | {
+    readonly engine: "openai-cloud";
+    readonly openaiModelName: string | undefined;
+    readonly openAiApiKey: string | undefined;
+    readonly openAiOrganization: string | undefined;
+    readonly openAiBaseUrl: string | undefined;
+} | {
+    readonly engine: "google-cloud";
+    readonly googleCloudApiKey: string;
+} | {
+    readonly engine: "microsoft-azure";
+    readonly azureServiceRegion: string;
+    readonly azureSubscriptionKey: string;
+} | {
+    readonly engine: "amazon-transcribe";
+    readonly amazonTranscribeRegion: string;
+    readonly amazonTranscribeAccessKeyId: string;
+    readonly amazonTranscribeSecretAccessKey: string;
+} | {
+    readonly engine: "deepgram";
+    readonly deepgramApiKey: string;
+    readonly deepgramModel: string;
+}, [0, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"whisper.cpp"> | undefined;
+    readonly model: [_optique_core.ValueParserResult<"tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0"> | undefined] | undefined;
+    readonly threads: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+    readonly processors: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+    readonly cpuOverride: [_optique_core.ValueParserResult<"blas" | "cpu"> | undefined] | undefined;
+}>] | [1, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"whisper-server"> | undefined;
+    readonly whisperServerUrl: _optique_core.ValueParserResult<URL> | undefined;
+    readonly whisperServerApiKey: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+}>] | [2, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"openai-cloud"> | undefined;
+    readonly openaiModelName: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiApiKey: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiOrganization: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiBaseUrl: [_optique_core.ValueParserResult<URL> | undefined] | undefined;
+}>] | [3, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"google-cloud"> | undefined;
+    readonly googleCloudApiKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [4, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"microsoft-azure"> | undefined;
+    readonly azureServiceRegion: _optique_core.ValueParserResult<string> | undefined;
+    readonly azureSubscriptionKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [5, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"amazon-transcribe"> | undefined;
+    readonly amazonTranscribeRegion: _optique_core.ValueParserResult<string> | undefined;
+    readonly amazonTranscribeAccessKeyId: _optique_core.ValueParserResult<string> | undefined;
+    readonly amazonTranscribeSecretAccessKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [6, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"deepgram"> | undefined;
+    readonly deepgramApiKey: _optique_core.ValueParserResult<string> | undefined;
+    readonly deepgramModel: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+}>] | undefined>;
+declare const transcribeCommand: _optique_core.Parser<"sync", {
+    readonly action: "transcribe";
+    readonly input: string;
+    readonly output: string;
+} & {
+    readonly parallelism: number;
+} & ({
+    readonly language: Intl.Locale | undefined;
+} & (({
+    readonly engine: "whisper.cpp";
+    readonly model: "tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0";
+    readonly threads: number;
+    readonly processors: number;
+    readonly cpuOverride: "blas" | "cpu" | undefined;
+} | {
+    readonly engine: "whisper-server";
+    readonly whisperServerUrl: string;
+    readonly whisperServerApiKey: string | undefined;
+} | {
+    readonly engine: "openai-cloud";
+    readonly openaiModelName: string | undefined;
+    readonly openAiApiKey: string | undefined;
+    readonly openAiOrganization: string | undefined;
+    readonly openAiBaseUrl: string | undefined;
+} | {
+    readonly engine: "google-cloud";
+    readonly googleCloudApiKey: string;
+} | {
+    readonly engine: "microsoft-azure";
+    readonly azureServiceRegion: string;
+    readonly azureSubscriptionKey: string;
+} | {
+    readonly engine: "amazon-transcribe";
+    readonly amazonTranscribeRegion: string;
+    readonly amazonTranscribeAccessKeyId: string;
+    readonly amazonTranscribeSecretAccessKey: string;
+} | {
+    readonly engine: "deepgram";
+    readonly deepgramApiKey: string;
+    readonly deepgramModel: string;
+}) & {
+    readonly noProgress: boolean;
+    readonly logLevel: "silent" | "debug" | "info" | "warn" | "error";
+    readonly time: boolean;
+})), ["matched", string] | ["parsing", Record<string | symbol, unknown>] | undefined>;
+
+export { transcribeCommand, transcribeParser };

package/dist/transcribe/parse.d.ts

@@ -0,0 +1,114 @@
+import * as _optique_core from '@optique/core';
+
+declare const transcribeParser: _optique_core.Parser<"sync", {
+    readonly engine: "whisper.cpp";
+    readonly model: "tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0";
+    readonly threads: number;
+    readonly processors: number;
+    readonly cpuOverride: "blas" | "cpu" | undefined;
+} | {
+    readonly engine: "whisper-server";
+    readonly whisperServerUrl: string;
+    readonly whisperServerApiKey: string | undefined;
+} | {
+    readonly engine: "openai-cloud";
+    readonly openaiModelName: string | undefined;
+    readonly openAiApiKey: string | undefined;
+    readonly openAiOrganization: string | undefined;
+    readonly openAiBaseUrl: string | undefined;
+} | {
+    readonly engine: "google-cloud";
+    readonly googleCloudApiKey: string;
+} | {
+    readonly engine: "microsoft-azure";
+    readonly azureServiceRegion: string;
+    readonly azureSubscriptionKey: string;
+} | {
+    readonly engine: "amazon-transcribe";
+    readonly amazonTranscribeRegion: string;
+    readonly amazonTranscribeAccessKeyId: string;
+    readonly amazonTranscribeSecretAccessKey: string;
+} | {
+    readonly engine: "deepgram";
+    readonly deepgramApiKey: string;
+    readonly deepgramModel: string;
+}, [0, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"whisper.cpp"> | undefined;
+    readonly model: [_optique_core.ValueParserResult<"tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0"> | undefined] | undefined;
+    readonly threads: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+    readonly processors: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+    readonly cpuOverride: [_optique_core.ValueParserResult<"blas" | "cpu"> | undefined] | undefined;
+}>] | [1, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"whisper-server"> | undefined;
+    readonly whisperServerUrl: _optique_core.ValueParserResult<URL> | undefined;
+    readonly whisperServerApiKey: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+}>] | [2, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"openai-cloud"> | undefined;
+    readonly openaiModelName: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiApiKey: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiOrganization: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+    readonly openAiBaseUrl: [_optique_core.ValueParserResult<URL> | undefined] | undefined;
+}>] | [3, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"google-cloud"> | undefined;
+    readonly googleCloudApiKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [4, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"microsoft-azure"> | undefined;
+    readonly azureServiceRegion: _optique_core.ValueParserResult<string> | undefined;
+    readonly azureSubscriptionKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [5, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"amazon-transcribe"> | undefined;
+    readonly amazonTranscribeRegion: _optique_core.ValueParserResult<string> | undefined;
+    readonly amazonTranscribeAccessKeyId: _optique_core.ValueParserResult<string> | undefined;
+    readonly amazonTranscribeSecretAccessKey: _optique_core.ValueParserResult<string> | undefined;
+}>] | [6, _optique_core.ParserResult<{
+    readonly engine: _optique_core.ValueParserResult<"deepgram"> | undefined;
+    readonly deepgramApiKey: _optique_core.ValueParserResult<string> | undefined;
+    readonly deepgramModel: [_optique_core.ValueParserResult<string> | undefined] | undefined;
+}>] | undefined>;
+declare const transcribeCommand: _optique_core.Parser<"sync", {
+    readonly action: "transcribe";
+    readonly input: string;
+    readonly output: string;
+} & {
+    readonly parallelism: number;
+} & ({
+    readonly language: Intl.Locale | undefined;
+} & (({
+    readonly engine: "whisper.cpp";
+    readonly model: "tiny" | "tiny.en" | "tiny-q5_1" | "tiny.en-q5_1" | "tiny-q8_0" | "base" | "base.en" | "base-q5_1" | "base.en-q5_1" | "base-q8_0" | "small" | "small.en" | "small-q5_1" | "small.en-q5_1" | "small-q8_0" | "medium" | "medium.en" | "medium-q5_0" | "medium.en-q5_0" | "medium-q8_0" | "large-v1" | "large-v2" | "large-v2-q5_0" | "large-v2-q8_0" | "large-v3" | "large-v3-q5_0" | "large-v3-turbo" | "large-v3-turbo-q5_0" | "large-v3-turbo-q8_0";
+    readonly threads: number;
+    readonly processors: number;
+    readonly cpuOverride: "blas" | "cpu" | undefined;
+} | {
+    readonly engine: "whisper-server";
+    readonly whisperServerUrl: string;
+    readonly whisperServerApiKey: string | undefined;
+} | {
+    readonly engine: "openai-cloud";
+    readonly openaiModelName: string | undefined;
+    readonly openAiApiKey: string | undefined;
+    readonly openAiOrganization: string | undefined;
+    readonly openAiBaseUrl: string | undefined;
+} | {
+    readonly engine: "google-cloud";
+    readonly googleCloudApiKey: string;
+} | {
+    readonly engine: "microsoft-azure";
+    readonly azureServiceRegion: string;
+    readonly azureSubscriptionKey: string;
+} | {
+    readonly engine: "amazon-transcribe";
+    readonly amazonTranscribeRegion: string;
+    readonly amazonTranscribeAccessKeyId: string;
+    readonly amazonTranscribeSecretAccessKey: string;
+} | {
+    readonly engine: "deepgram";
+    readonly deepgramApiKey: string;
+    readonly deepgramModel: string;
+}) & {
+    readonly noProgress: boolean;
+    readonly logLevel: "silent" | "debug" | "info" | "warn" | "error";
+    readonly time: boolean;
+})), ["matched", string] | ["parsing", Record<string | symbol, unknown>] | undefined>;
+
+export { transcribeCommand, transcribeParser };

package/dist/transcribe/parse.js

@@ -0,0 +1,143 @@
+import "../chunk-BIEQXUOY.js";
+import {
+  argument,
+  choice,
+  command,
+  constant,
+  integer,
+  map,
+  merge,
+  message,
+  object,
+  option,
+  optional,
+  or,
+  withDefault
+} from "@optique/core";
+import { string, url } from "@optique/core/valueparser";
+import { path } from "@optique/run/valueparser";
+import { WHISPER_MODELS } from "@storyteller-platform/ghost-story";
+import {
+  languageParser,
+  loggingParser,
+  parallelismParser
+} from "../common/parse.js";
+const transcribeParser = or(
+  object("whisper.cpp", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["whisper.cpp"], { metavar: "whisper.cpp" })
+    ),
+    model: withDefault(
+      option("--model", "-m", choice(WHISPER_MODELS, { metavar: "MODEL" }), {
+        description: message`The whisper model to use`
+      }),
+      "tiny.en"
+    ),
+    threads: withDefault(option("--threads", integer()), 4),
+    processors: withDefault(
+      option("--processors", integer(), {
+        description: message`The number of processors to use (values greater than 1 may affect timing accuracy)`
+      }),
+      1
+    ),
+    cpuOverride: optional(
+      option(
+        "--cpu-override",
+        choice(["blas", "cpu"], { metavar: "CPU_ENGINE" }),
+        {
+          description: message`When provided, will use this whisper variant even if another is available`
+        }
+      )
+    )
+  }),
+  object("whisper-server", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["whisper-server"], { metavar: "whisper-server" })
+    ),
+    whisperServerUrl: map(
+      option("--whisper-server-url", url()),
+      (url2) => url2.toString()
+    ),
+    whisperServerApiKey: optional(option("--whisper-server-api-key", string()))
+  }),
+  object("openai-cloud", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["openai-cloud"], { metavar: "openai-cloud" })
+    ),
+    openaiModelName: optional(option("--openai-model", string())),
+    openAiApiKey: optional(option("--openai-api-key", string())),
+    openAiOrganization: optional(option("--openai-organization", string())),
+    openAiBaseUrl: optional(
+      map(option("--openai-base-url", url()), (url2) => url2.toString())
+    )
+  }),
+  object("google-cloud", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["google-cloud"], { metavar: "google-cloud" })
+    ),
+    googleCloudApiKey: option("--google-cloud-api-key", string())
+  }),
+  object("microsoft-azure", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["microsoft-azure"], { metavar: "microsoft-azure" })
+    ),
+    azureServiceRegion: option("--azure-service-region", string()),
+    azureSubscriptionKey: option("--azure-subscription-key", string())
+  }),
+  object("amazon-transcribe", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["amazon-transcribe"], { metavar: "amazon-transcribe" })
+    ),
+    amazonTranscribeRegion: option("--amazon-transcribe-region", string()),
+    amazonTranscribeAccessKeyId: option(
+      "--amazon-transcribe-access-key-id",
+      string()
+    ),
+    amazonTranscribeSecretAccessKey: option(
+      "--amazon-transcribe-secret-access-key",
+      string()
+    )
+  }),
+  object("deepgram", {
+    engine: option(
+      "--engine",
+      "-e",
+      choice(["deepgram"], { metavar: "deepgram" })
+    ),
+    deepgramApiKey: option("--deepgram-api-key", string()),
+    deepgramModel: withDefault(option("--deepgram-model", string()), "nova-3")
+  })
+);
+const transcribeCommand = command(
+  "transcribe",
+  merge(
+    object({
+      action: constant("transcribe"),
+      input: argument(
+        path({ type: "directory", mustExist: true, metavar: "INPUT" })
+      ),
+      output: argument(path({ metavar: "OUTPUT", type: "directory" }))
+    }),
+    parallelismParser,
+    languageParser,
+    transcribeParser,
+    loggingParser
+  ),
+  { description: message`Transcribe a directory of audiobook files.` }
+);
+export {
+  transcribeCommand,
+  transcribeParser
+};