@storyteller-platform/align 0.0.1

package/dist/process/__tests__/processAudiobook.test.js

@@ -0,0 +1,209 @@
+import assert from "node:assert";
+import { randomUUID } from "node:crypto";
+import { mkdir, readdir } from "node:fs/promises";
+import { availableParallelism } from "node:os";
+import { join } from "node:path";
+import { describe, it } from "node:test";
+import { AsyncSemaphore } from "@esfx/async-semaphore";
+import { createLogger } from "../../common/logging.js";
+import { processFile } from "../processAudiobook.js";
+function createTestLogger() {
+  return createLogger(process.env["CI"] ? "silent" : "info");
+}
+void describe("processFile", () => {
+  void it("can process mpeg4 files", async () => {
+    const input = join(
+      "src",
+      "__fixtures__",
+      "mpeg4",
+      "MobyDickOrTheWhalePart1_librivox.m4b"
+    );
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp4",
+      "00000-00002.mp4",
+      "00000-00003.mp4",
+      "00000-00004.mp4",
+      "00000-00005.mp4",
+      "00000-00006.mp4",
+      "00000-00007.mp4",
+      "00000-00008.mp4",
+      "00000-00009.mp4",
+      "00000-00010.mp4",
+      "00000-00011.mp4"
+    ]);
+  });
+  void it("can process mp3 files", async () => {
+    const input = join(
+      "src",
+      "__fixtures__",
+      "mp3",
+      "MobyDickOrTheWhalePart1_librivox.mp3"
+    );
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp3",
+      "00000-00002.mp3",
+      "00000-00003.mp3",
+      "00000-00004.mp3",
+      "00000-00005.mp3",
+      "00000-00006.mp3",
+      "00000-00007.mp3",
+      "00000-00008.mp3",
+      "00000-00009.mp3",
+      "00000-00010.mp3",
+      "00000-00011.mp3"
+    ]);
+  });
+  void it("can process opus files", async () => {
+    const input = join(
+      "src",
+      "__fixtures__",
+      "opus",
+      "MobyDickOrTheWhalePart1_librivox.opus"
+    );
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp4",
+      "00000-00002.mp4",
+      "00000-00003.mp4",
+      "00000-00004.mp4",
+      "00000-00005.mp4",
+      "00000-00006.mp4",
+      "00000-00007.mp4",
+      "00000-00008.mp4",
+      "00000-00009.mp4",
+      "00000-00010.mp4",
+      "00000-00011.mp4"
+    ]);
+  });
+  void it("can handle nonstandard audio file", async () => {
+    const input = join("src", "__fixtures__", "flac", "mobydick.flac");
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, ["00000-00001.flac"]);
+  });
+  void it("can transcode nonstandard audio file", async () => {
+    const input = join("src", "__fixtures__", "flac", "mobydick.flac");
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      encoding: { codec: "libopus" },
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, ["00000-00001.mp4"]);
+  });
+  void it("can process zip files", async () => {
+    const input = join("src", "__fixtures__", "zip", "moby-dick-1-7-audio.zip");
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp3",
+      "00000-00002.mp3",
+      "00000-00003.mp3"
+    ]);
+  });
+  void it("splits files longer than the max length", async () => {
+    const input = join("src", "__fixtures__", "zip", "moby-dick-1-7-audio.zip");
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      maxLength: 0.1,
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp3",
+      "00000-00002.mp3",
+      "00000-00003.mp3",
+      "00000-00004.mp3",
+      "00000-00005.mp3",
+      "00000-00006.mp3",
+      "00000-00007.mp3",
+      "00000-00008.mp3",
+      "00000-00009.mp3",
+      "00000-00010.mp3",
+      "00000-00011.mp3",
+      "00000-00012.mp3",
+      "00000-00013.mp3",
+      "00000-00014.mp3",
+      "00000-00015.mp3",
+      "00000-00016.mp3",
+      "00000-00017.mp3"
+    ]);
+  });
+  void it("splits files longer than the max length, even if they have chapters", async () => {
+    const input = join(
+      "src",
+      "__fixtures__",
+      "mpeg4",
+      "MobyDickOrTheWhalePart1_librivox.m4b"
+    );
+    const outDir = join("src", "__fixtures__", "__output__", randomUUID());
+    await mkdir(outDir, { recursive: true });
+    await processFile(input, outDir, "00000-", {
+      maxLength: 0.25,
+      lock: new AsyncSemaphore(availableParallelism() - 1),
+      logger: createTestLogger()
+    });
+    const outFiles = await readdir(outDir);
+    assert.deepStrictEqual(outFiles, [
+      "00000-00001.mp4",
+      "00000-00002.mp4",
+      "00000-00003.mp4",
+      "00000-00004.mp4",
+      "00000-00005.mp4",
+      "00000-00006.mp4",
+      "00000-00007.mp4",
+      "00000-00008.mp4",
+      "00000-00009.mp4",
+      "00000-00010.mp4",
+      "00000-00011.mp4",
+      "00000-00012.mp4",
+      "00000-00013.mp4",
+      "00000-00014.mp4",
+      "00000-00015.mp4",
+      "00000-00016.mp4",
+      "00000-00017.mp4",
+      "00000-00018.mp4",
+      "00000-00019.mp4",
+      "00000-00020.mp4",
+      "00000-00021.mp4",
+      "00000-00022.mp4",
+      "00000-00023.mp4",
+      "00000-00024.mp4",
+      "00000-00025.mp4"
+    ]);
+  });
+});

package/dist/process/mime.cjs

@@ -0,0 +1,43 @@
+"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 mime_exports = {};
+__export(mime_exports, {
+  areSameType: () => areSameType
+});
+module.exports = __toCommonJS(mime_exports);
+var import_audiobook = require("@storyteller-platform/audiobook");
+function areSameType(filepathA, filepathB) {
+  if (filepathA === filepathB) {
+    return true;
+  }
+  if (import_audiobook.MPEG4_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && import_audiobook.MPEG4_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  if (import_audiobook.MP3_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && import_audiobook.MP3_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  if (import_audiobook.OPUS_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && import_audiobook.OPUS_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  return false;
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  areSameType
+});

package/dist/process/mime.d.cts

@@ -0,0 +1,3 @@
+declare function areSameType(filepathA: string, filepathB: string): boolean;
+
+export { areSameType };

package/dist/process/mime.d.ts

@@ -0,0 +1,3 @@
+declare function areSameType(filepathA: string, filepathB: string): boolean;
+
+export { areSameType };

package/dist/process/mime.js

@@ -0,0 +1,24 @@
+import "../chunk-BIEQXUOY.js";
+import {
+  MP3_FILE_EXTENSIONS,
+  MPEG4_FILE_EXTENSIONS,
+  OPUS_FILE_EXTENSIONS
+} from "@storyteller-platform/audiobook";
+function areSameType(filepathA, filepathB) {
+  if (filepathA === filepathB) {
+    return true;
+  }
+  if (MPEG4_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && MPEG4_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  if (MP3_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && MP3_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  if (OPUS_FILE_EXTENSIONS.some((ext) => filepathA.endsWith(ext)) && OPUS_FILE_EXTENSIONS.some((ext) => filepathB.endsWith(ext))) {
+    return true;
+  }
+  return false;
+}
+export {
+  areSameType
+};

package/dist/process/parse.cjs

@@ -0,0 +1,84 @@
+"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, {
+  processCommand: () => processCommand,
+  processParser: () => processParser
+});
+module.exports = __toCommonJS(parse_exports);
+var import_core = require("@optique/core");
+var import_valueparser = require("@optique/run/valueparser");
+var import_parse = require("../common/parse.cjs");
+const codecParser = (0, import_core.dependency)((0, import_core.choice)(["libopus", "lipmp3lame", "aac", "copy"]));
+const bitrateParser = codecParser.derive({
+  metavar: "BITRATE",
+  factory: (codec) => {
+    switch (codec) {
+      case "libopus": {
+        return (0, import_core.choice)(["16K", "32K", "64K", "96K"]);
+      }
+      case "lipmp3lame": {
+        return (0, import_core.choice)(["0", "1", "2", "3", "4", "5", "6", "7", "8", "9"]);
+      }
+      default:
+        return (0, import_core.choice)([]);
+    }
+  },
+  defaultValue: () => "copy"
+});
+const processParser = (0, import_core.object)("Audio processing", {
+  codec: (0, import_core.withDefault)(
+    (0, import_core.option)("--codec", codecParser, {
+      description: import_core.message`The audio codec to transcode to. If unspecified, will copy audio data without transcoding.`
+    }),
+    "copy"
+  ),
+  bitrate: (0, import_core.optional)(
+    (0, import_core.option)("--bitrate", bitrateParser, {
+      description: import_core.message`The audio bitrate to transcode to ()`
+    })
+  ),
+  maxLength: (0, import_core.withDefault)(
+    (0, import_core.option)("--max-length", (0, import_core.integer)(), {
+      description: import_core.message`The maximum allowed length of a processed audio track, in minutes`
+    }),
+    120
+  )
+});
+const processCommand = (0, import_core.command)(
+  "process",
+  (0, import_core.merge)(
+    (0, import_core.object)({
+      action: (0, import_core.constant)("process"),
+      input: (0, import_core.argument)(
+        (0, import_valueparser.path)({ type: "directory", mustExist: true, metavar: "INPUT" })
+      ),
+      output: (0, import_core.argument)((0, import_valueparser.path)({ metavar: "OUTPUT", type: "directory" }))
+    }),
+    import_parse.parallelismParser,
+    processParser,
+    import_parse.loggingParser
+  ),
+  { description: import_core.message`Process audiobook files for transcription.` }
+);
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  processCommand,
+  processParser
+});

package/dist/process/parse.d.cts

@@ -0,0 +1,28 @@
+import * as _optique_core from '@optique/core';
+
+declare const processParser: _optique_core.Parser<"sync", {
+    readonly codec: "libopus" | "lipmp3lame" | "aac" | "copy";
+    readonly bitrate: "0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9" | undefined;
+    readonly maxLength: number;
+}, {
+    readonly codec: [_optique_core.ValueParserResult<"libopus" | "lipmp3lame" | "aac" | "copy"> | undefined] | undefined;
+    readonly bitrate: [_optique_core.ValueParserResult<"0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9"> | undefined] | undefined;
+    readonly maxLength: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+}>;
+declare const processCommand: _optique_core.Parser<"sync", {
+    readonly action: "process";
+    readonly input: string;
+    readonly output: string;
+} & {
+    readonly parallelism: number;
+} & {
+    readonly codec: "libopus" | "lipmp3lame" | "aac" | "copy";
+    readonly bitrate: "0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9" | undefined;
+    readonly maxLength: number;
+} & {
+    readonly noProgress: boolean;
+    readonly logLevel: "silent" | "debug" | "info" | "warn" | "error";
+    readonly time: boolean;
+}, ["matched", string] | ["parsing", Record<string | symbol, unknown>] | undefined>;
+
+export { processCommand, processParser };

package/dist/process/parse.d.ts

@@ -0,0 +1,28 @@
+import * as _optique_core from '@optique/core';
+
+declare const processParser: _optique_core.Parser<"sync", {
+    readonly codec: "libopus" | "lipmp3lame" | "aac" | "copy";
+    readonly bitrate: "0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9" | undefined;
+    readonly maxLength: number;
+}, {
+    readonly codec: [_optique_core.ValueParserResult<"libopus" | "lipmp3lame" | "aac" | "copy"> | undefined] | undefined;
+    readonly bitrate: [_optique_core.ValueParserResult<"0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9"> | undefined] | undefined;
+    readonly maxLength: [_optique_core.ValueParserResult<number> | undefined] | undefined;
+}>;
+declare const processCommand: _optique_core.Parser<"sync", {
+    readonly action: "process";
+    readonly input: string;
+    readonly output: string;
+} & {
+    readonly parallelism: number;
+} & {
+    readonly codec: "libopus" | "lipmp3lame" | "aac" | "copy";
+    readonly bitrate: "0" | "1" | "2" | "3" | "16K" | "32K" | "64K" | "96K" | "4" | "5" | "6" | "7" | "8" | "9" | undefined;
+    readonly maxLength: number;
+} & {
+    readonly noProgress: boolean;
+    readonly logLevel: "silent" | "debug" | "info" | "warn" | "error";
+    readonly time: boolean;
+}, ["matched", string] | ["parsing", Record<string | symbol, unknown>] | undefined>;
+
+export { processCommand, processParser };

package/dist/process/parse.js

@@ -0,0 +1,73 @@
+import "../chunk-BIEQXUOY.js";
+import {
+  argument,
+  choice,
+  command,
+  constant,
+  dependency,
+  integer,
+  merge,
+  message,
+  object,
+  option,
+  optional,
+  withDefault
+} from "@optique/core";
+import { path } from "@optique/run/valueparser";
+import { loggingParser, parallelismParser } from "../common/parse.js";
+const codecParser = dependency(choice(["libopus", "lipmp3lame", "aac", "copy"]));
+const bitrateParser = codecParser.derive({
+  metavar: "BITRATE",
+  factory: (codec) => {
+    switch (codec) {
+      case "libopus": {
+        return choice(["16K", "32K", "64K", "96K"]);
+      }
+      case "lipmp3lame": {
+        return choice(["0", "1", "2", "3", "4", "5", "6", "7", "8", "9"]);
+      }
+      default:
+        return choice([]);
+    }
+  },
+  defaultValue: () => "copy"
+});
+const processParser = object("Audio processing", {
+  codec: withDefault(
+    option("--codec", codecParser, {
+      description: message`The audio codec to transcode to. If unspecified, will copy audio data without transcoding.`
+    }),
+    "copy"
+  ),
+  bitrate: optional(
+    option("--bitrate", bitrateParser, {
+      description: message`The audio bitrate to transcode to ()`
+    })
+  ),
+  maxLength: withDefault(
+    option("--max-length", integer(), {
+      description: message`The maximum allowed length of a processed audio track, in minutes`
+    }),
+    120
+  )
+});
+const processCommand = command(
+  "process",
+  merge(
+    object({
+      action: constant("process"),
+      input: argument(
+        path({ type: "directory", mustExist: true, metavar: "INPUT" })
+      ),
+      output: argument(path({ metavar: "OUTPUT", type: "directory" }))
+    }),
+    parallelismParser,
+    processParser,
+    loggingParser
+  ),
+  { description: message`Process audiobook files for transcription.` }
+);
+export {
+  processCommand,
+  processParser
+};