@storyteller-platform/align 0.1.20 → 0.1.22

package/dist/snapshot/parse.js

@@ -0,0 +1,45 @@
+import "../chunk-BIEQXUOY.js";
+import {
+  argument,
+  command,
+  constant,
+  merge,
+  message,
+  object,
+  option
+} from "@optique/core";
+import { path } from "@optique/run";
+import { loggingParser } from "../common/parse.js";
+const snapshotParser = object("Snapshot", {
+  transcriptions: option(
+    "--transcriptions",
+    path({ mustExist: true, type: "directory" })
+  ),
+  epub: option(
+    "--epub",
+    path({ mustExist: true, type: "file", extensions: [".epub"] }),
+    {
+      description: message`Path to an EPUB file to snapshot. This EPUB must have Media Overlays and audio files corresponding to the transcription files passed to --transcriptions.`
+    }
+  ),
+  output: argument(path({ type: "file", metavar: "OUTPUT_PATH" }), {
+    description: message`Path to save the snapshot.`
+  })
+});
+const snapshotCommand = command(
+  "snapshot",
+  merge(
+    object({
+      action: constant("snapshot")
+    }),
+    snapshotParser,
+    loggingParser
+  ),
+  {
+    description: message`Print a human-readable snapshot of the EPUB’s alignment to a text file.`
+  }
+);
+export {
+  snapshotCommand,
+  snapshotParser
+};

package/dist/snapshot/snapshot.cjs

@@ -0,0 +1,224 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __knownSymbol = (name, symbol) => (symbol = Symbol[name]) ? symbol : Symbol.for("Symbol." + name);
+var __typeError = (msg) => {
+  throw TypeError(msg);
+};
+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 __using = (stack, value, async) => {
+  if (value != null) {
+    if (typeof value !== "object" && typeof value !== "function") __typeError("Object expected");
+    var dispose, inner;
+    if (async) dispose = value[__knownSymbol("asyncDispose")];
+    if (dispose === void 0) {
+      dispose = value[__knownSymbol("dispose")];
+      if (async) inner = dispose;
+    }
+    if (typeof dispose !== "function") __typeError("Object not disposable");
+    if (inner) dispose = function() {
+      try {
+        inner.call(this);
+      } catch (e) {
+        return Promise.reject(e);
+      }
+    };
+    stack.push([async, dispose, value]);
+  } else if (async) {
+    stack.push([async]);
+  }
+  return value;
+};
+var __callDispose = (stack, error, hasError) => {
+  var E = typeof SuppressedError === "function" ? SuppressedError : function(e, s, m, _) {
+    return _ = Error(m), _.name = "SuppressedError", _.error = e, _.suppressed = s, _;
+  };
+  var fail = (e) => error = hasError ? new E(e, error, "An error was suppressed during disposal") : (hasError = true, e);
+  var next = (it) => {
+    while (it = stack.pop()) {
+      try {
+        var result = it[1] && it[1].call(it[2]);
+        if (it[0]) return Promise.resolve(result).then(next, (e) => (fail(e), next()));
+      } catch (e) {
+        fail(e);
+      }
+    }
+    if (hasError) throw error;
+  };
+  return next();
+};
+var snapshot_exports = {};
+__export(snapshot_exports, {
+  createAlignmentSnapshot: () => createAlignmentSnapshot,
+  snapshotAlignment: () => snapshotAlignment
+});
+module.exports = __toCommonJS(snapshot_exports);
+var import_promises = require("node:fs/promises");
+var import_node_path = require("node:path");
+var import_posix = require("node:path/posix");
+var import_epub = require("@storyteller-platform/epub");
+var import_segmentation = require("../markup/segmentation.cjs");
+async function snapshotAlignment(epubPath, transcriptionsPath, outputPath) {
+  var _stack = [];
+  try {
+    const transcriptionFilepaths = await (0, import_promises.readdir)(transcriptionsPath).then(
+      (filenames) => filenames.filter((f) => f.endsWith(".json")).map((f) => (0, import_node_path.join)(transcriptionsPath, f))
+    );
+    const epub = __using(_stack, await import_epub.Epub.from(epubPath));
+    const snapshot = await createAlignmentSnapshot(
+      epub,
+      transcriptionFilepaths,
+      "id-fragment"
+    );
+    await (0, import_promises.writeFile)(outputPath, snapshot, { encoding: "utf-8" });
+  } catch (_) {
+    var _error = _, _hasError = true;
+  } finally {
+    __callDispose(_stack, _error, _hasError);
+  }
+}
+async function createAlignmentSnapshot(epub, transcriptionFilepaths, textRef) {
+  let newSnapshot = "";
+  const manifest = await epub.getManifest();
+  const mediaOverlayItems = Object.values(manifest).map((item) => item.mediaOverlay).filter((mediaOverlayId) => !!mediaOverlayId).map((id) => manifest[id]);
+  const mediaOverlays = [];
+  for (const item of mediaOverlayItems) {
+    const contents = await epub.readItemContents(item.id, "utf-8");
+    const parsed = import_epub.Epub.xmlParser.parse(contents);
+    mediaOverlays.push(parsed);
+    const smil = import_epub.Epub.findXmlChildByName("smil", parsed);
+    if (!smil) continue;
+    const body = import_epub.Epub.findXmlChildByName("body", import_epub.Epub.getXmlChildren(smil));
+    if (!body) continue;
+    const seq = import_epub.Epub.findXmlChildByName("seq", import_epub.Epub.getXmlChildren(body));
+    if (!seq) continue;
+    const textref = seq[":@"]?.["@_epub:textref"];
+    if (!textref) continue;
+    newSnapshot += `// ${(0, import_posix.basename)(textref)}
+
+`;
+    const chapterContents = await epub.readFileContents(
+      textref,
+      item.href,
+      "utf-8"
+    );
+    const chapterXml = import_epub.Epub.xhtmlParser.parse(chapterContents);
+    const { result: segmentation } = await (0, import_segmentation.getXhtmlSegmentation)(
+      import_epub.Epub.getXhtmlBody(chapterXml),
+      {
+        primaryLocale: new Intl.Locale("en-US")
+      }
+    );
+    let lastChapterSentence = -1;
+    const chapterSentences = segmentation.filter((s) => s.text.match(/\S/));
+    for (const par of import_epub.Epub.getXmlChildren(seq)) {
+      newSnapshot += `
+`;
+      const text = import_epub.Epub.findXmlChildByName("text", import_epub.Epub.getXmlChildren(par));
+      if (!text) continue;
+      const audio = import_epub.Epub.findXmlChildByName("audio", import_epub.Epub.getXmlChildren(par));
+      if (!audio) continue;
+      const textSrc = text[":@"]?.["@_src"];
+      if (!textSrc) continue;
+      const result = textRef === "id-fragment" ? getTextSentenceIndexByIdFragment(textSrc) : getTextSentenceIndexByTextFragment(
+        textSrc,
+        chapterSentences,
+        lastChapterSentence
+      );
+      if (result === null) continue;
+      const { fragment, sentenceId } = result;
+      const textSentence = chapterSentences[sentenceId]?.text;
+      if (!textSentence) continue;
+      lastChapterSentence = sentenceId;
+      if (textRef === "text-fragment") {
+        newSnapshot += `${fragment}
+`;
+      }
+      newSnapshot += `Text:  ${textSentence.replace(/\n/, "")}
+`;
+      const audioSrc = audio[":@"]?.["@_src"];
+      if (!audioSrc) continue;
+      const audioStart = audio[":@"]?.["@_clipBegin"];
+      const audioEnd = audio[":@"]?.["@_clipEnd"];
+      if (!audioStart || !audioEnd) continue;
+      const audioStartTime = parseFloat(audioStart.slice(0, -1));
+      const audioEndTime = parseFloat(audioEnd.slice(0, -1));
+      const audioFilename = (0, import_posix.basename)(audioSrc, (0, import_posix.extname)(audioSrc));
+      const transcriptionFilepath = transcriptionFilepaths.find(
+        (f) => (0, import_node_path.basename)(f, (0, import_node_path.extname)(f)) === audioFilename
+      );
+      if (!transcriptionFilepath) continue;
+      const transcription = JSON.parse(
+        await (0, import_promises.readFile)(transcriptionFilepath, { encoding: "utf-8" })
+      );
+      if ("wordTimeline" in transcription) {
+        transcription.timeline = transcription.wordTimeline;
+      }
+      const transcriptionWords = [];
+      let started = false;
+      let i = 0;
+      let word = transcription.timeline[i];
+      while (word && word.endTime <= audioEndTime) {
+        if (word.startTime >= audioStartTime) {
+          started = true;
+        }
+        if (started) {
+          transcriptionWords.push(word.text);
+        }
+        word = transcription.timeline[++i];
+      }
+      const transcriptionSentence = transcriptionWords.join(" ");
+      newSnapshot += `Audio: ${transcriptionSentence}
+`;
+    }
+    newSnapshot += `
+`;
+  }
+  return newSnapshot;
+}
+function getTextSentenceIndexByTextFragment(textSrc, chapterSentences, lastChapterSentence) {
+  const textSrcMatch = textSrc.match(/#:~:text=(.+)$/);
+  if (!textSrcMatch) return null;
+  const textFragment = textSrcMatch[1];
+  if (textFragment === void 0) return null;
+  const textFragmentParts = textFragment.split(",");
+  const textFragmentPrefix = textFragmentParts[0]?.endsWith("-") ? decodeURIComponent(textFragmentParts[0]).slice(0, -1) : "";
+  const textFragmentStart = decodeURIComponent(
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    textFragmentPrefix ? textFragmentParts[1] : textFragmentParts[0]
+  );
+  const textSentenceIndex = chapterSentences.slice(lastChapterSentence + 1).findIndex((s, i) => {
+    const prev = chapterSentences[lastChapterSentence + i];
+    return (!prev || prev.text.replace("\n", " ").toLowerCase().endsWith(textFragmentPrefix)) && s.text.replace("\n", " ").toLowerCase().startsWith(textFragmentStart);
+  });
+  if (textSentenceIndex === -1) return null;
+  return {
+    fragment: textSrcMatch[0],
+    sentenceId: textSentenceIndex + lastChapterSentence + 1
+  };
+}
+function getTextSentenceIndexByIdFragment(textSrc) {
+  const match = textSrc.match(/#.*s([0-9]+)$/);
+  if (!match) return null;
+  const [fragment, sentenceId] = match;
+  return { fragment, sentenceId: parseInt(sentenceId, 10) };
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  createAlignmentSnapshot,
+  snapshotAlignment
+});

package/dist/snapshot/snapshot.d.cts

@@ -0,0 +1,6 @@
+import { Epub } from '@storyteller-platform/epub';
+
+declare function snapshotAlignment(epubPath: string, transcriptionsPath: string, outputPath: string): Promise<void>;
+declare function createAlignmentSnapshot(epub: Epub, transcriptionFilepaths: string[], textRef: "id-fragment" | "text-fragment"): Promise<string>;
+
+export { createAlignmentSnapshot, snapshotAlignment };

package/dist/snapshot/snapshot.d.ts

@@ -0,0 +1,6 @@
+import { Epub } from '@storyteller-platform/epub';
+
+declare function snapshotAlignment(epubPath: string, transcriptionsPath: string, outputPath: string): Promise<void>;
+declare function createAlignmentSnapshot(epub: Epub, transcriptionFilepaths: string[], textRef: "id-fragment" | "text-fragment"): Promise<string>;
+
+export { createAlignmentSnapshot, snapshotAlignment };

package/dist/snapshot/snapshot.js

@@ -0,0 +1,161 @@
+import {
+  __callDispose,
+  __using
+} from "../chunk-BIEQXUOY.js";
+import { readFile, readdir, writeFile } from "node:fs/promises";
+import { basename, extname, join } from "node:path";
+import {
+  basename as posixBasename,
+  extname as posixExtname
+} from "node:path/posix";
+import { Epub } from "@storyteller-platform/epub";
+import { getXhtmlSegmentation } from "../markup/segmentation.js";
+async function snapshotAlignment(epubPath, transcriptionsPath, outputPath) {
+  var _stack = [];
+  try {
+    const transcriptionFilepaths = await readdir(transcriptionsPath).then(
+      (filenames) => filenames.filter((f) => f.endsWith(".json")).map((f) => join(transcriptionsPath, f))
+    );
+    const epub = __using(_stack, await Epub.from(epubPath));
+    const snapshot = await createAlignmentSnapshot(
+      epub,
+      transcriptionFilepaths,
+      "id-fragment"
+    );
+    await writeFile(outputPath, snapshot, { encoding: "utf-8" });
+  } catch (_) {
+    var _error = _, _hasError = true;
+  } finally {
+    __callDispose(_stack, _error, _hasError);
+  }
+}
+async function createAlignmentSnapshot(epub, transcriptionFilepaths, textRef) {
+  let newSnapshot = "";
+  const manifest = await epub.getManifest();
+  const mediaOverlayItems = Object.values(manifest).map((item) => item.mediaOverlay).filter((mediaOverlayId) => !!mediaOverlayId).map((id) => manifest[id]);
+  const mediaOverlays = [];
+  for (const item of mediaOverlayItems) {
+    const contents = await epub.readItemContents(item.id, "utf-8");
+    const parsed = Epub.xmlParser.parse(contents);
+    mediaOverlays.push(parsed);
+    const smil = Epub.findXmlChildByName("smil", parsed);
+    if (!smil) continue;
+    const body = Epub.findXmlChildByName("body", Epub.getXmlChildren(smil));
+    if (!body) continue;
+    const seq = Epub.findXmlChildByName("seq", Epub.getXmlChildren(body));
+    if (!seq) continue;
+    const textref = seq[":@"]?.["@_epub:textref"];
+    if (!textref) continue;
+    newSnapshot += `// ${posixBasename(textref)}
+
+`;
+    const chapterContents = await epub.readFileContents(
+      textref,
+      item.href,
+      "utf-8"
+    );
+    const chapterXml = Epub.xhtmlParser.parse(chapterContents);
+    const { result: segmentation } = await getXhtmlSegmentation(
+      Epub.getXhtmlBody(chapterXml),
+      {
+        primaryLocale: new Intl.Locale("en-US")
+      }
+    );
+    let lastChapterSentence = -1;
+    const chapterSentences = segmentation.filter((s) => s.text.match(/\S/));
+    for (const par of Epub.getXmlChildren(seq)) {
+      newSnapshot += `
+`;
+      const text = Epub.findXmlChildByName("text", Epub.getXmlChildren(par));
+      if (!text) continue;
+      const audio = Epub.findXmlChildByName("audio", Epub.getXmlChildren(par));
+      if (!audio) continue;
+      const textSrc = text[":@"]?.["@_src"];
+      if (!textSrc) continue;
+      const result = textRef === "id-fragment" ? getTextSentenceIndexByIdFragment(textSrc) : getTextSentenceIndexByTextFragment(
+        textSrc,
+        chapterSentences,
+        lastChapterSentence
+      );
+      if (result === null) continue;
+      const { fragment, sentenceId } = result;
+      const textSentence = chapterSentences[sentenceId]?.text;
+      if (!textSentence) continue;
+      lastChapterSentence = sentenceId;
+      if (textRef === "text-fragment") {
+        newSnapshot += `${fragment}
+`;
+      }
+      newSnapshot += `Text:  ${textSentence.replace(/\n/, "")}
+`;
+      const audioSrc = audio[":@"]?.["@_src"];
+      if (!audioSrc) continue;
+      const audioStart = audio[":@"]?.["@_clipBegin"];
+      const audioEnd = audio[":@"]?.["@_clipEnd"];
+      if (!audioStart || !audioEnd) continue;
+      const audioStartTime = parseFloat(audioStart.slice(0, -1));
+      const audioEndTime = parseFloat(audioEnd.slice(0, -1));
+      const audioFilename = posixBasename(audioSrc, posixExtname(audioSrc));
+      const transcriptionFilepath = transcriptionFilepaths.find(
+        (f) => basename(f, extname(f)) === audioFilename
+      );
+      if (!transcriptionFilepath) continue;
+      const transcription = JSON.parse(
+        await readFile(transcriptionFilepath, { encoding: "utf-8" })
+      );
+      if ("wordTimeline" in transcription) {
+        transcription.timeline = transcription.wordTimeline;
+      }
+      const transcriptionWords = [];
+      let started = false;
+      let i = 0;
+      let word = transcription.timeline[i];
+      while (word && word.endTime <= audioEndTime) {
+        if (word.startTime >= audioStartTime) {
+          started = true;
+        }
+        if (started) {
+          transcriptionWords.push(word.text);
+        }
+        word = transcription.timeline[++i];
+      }
+      const transcriptionSentence = transcriptionWords.join(" ");
+      newSnapshot += `Audio: ${transcriptionSentence}
+`;
+    }
+    newSnapshot += `
+`;
+  }
+  return newSnapshot;
+}
+function getTextSentenceIndexByTextFragment(textSrc, chapterSentences, lastChapterSentence) {
+  const textSrcMatch = textSrc.match(/#:~:text=(.+)$/);
+  if (!textSrcMatch) return null;
+  const textFragment = textSrcMatch[1];
+  if (textFragment === void 0) return null;
+  const textFragmentParts = textFragment.split(",");
+  const textFragmentPrefix = textFragmentParts[0]?.endsWith("-") ? decodeURIComponent(textFragmentParts[0]).slice(0, -1) : "";
+  const textFragmentStart = decodeURIComponent(
+    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion
+    textFragmentPrefix ? textFragmentParts[1] : textFragmentParts[0]
+  );
+  const textSentenceIndex = chapterSentences.slice(lastChapterSentence + 1).findIndex((s, i) => {
+    const prev = chapterSentences[lastChapterSentence + i];
+    return (!prev || prev.text.replace("\n", " ").toLowerCase().endsWith(textFragmentPrefix)) && s.text.replace("\n", " ").toLowerCase().startsWith(textFragmentStart);
+  });
+  if (textSentenceIndex === -1) return null;
+  return {
+    fragment: textSrcMatch[0],
+    sentenceId: textSentenceIndex + lastChapterSentence + 1
+  };
+}
+function getTextSentenceIndexByIdFragment(textSrc) {
+  const match = textSrc.match(/#.*s([0-9]+)$/);
+  if (!match) return null;
+  const [fragment, sentenceId] = match;
+  return { fragment, sentenceId: parseInt(sentenceId, 10) };
+}
+export {
+  createAlignmentSnapshot,
+  snapshotAlignment
+};

package/dist/transcribe/parse.cjs

@@ -25,7 +25,7 @@ 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_constants = require("@storyteller-platform/ghost-story/constants");
 var import_parse = require("../common/parse.cjs");
 const transcribeParser = (0, import_core.or)(
   (0, import_core.object)("whisper.cpp", {
@@ -35,7 +35,7 @@ const transcribeParser = (0, import_core.or)(
       (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" }), {
+      (0, import_core.option)("--model", "-m", (0, import_core.choice)(import_constants.WHISPER_MODELS, { metavar: "MODEL" }), {
         description: import_core.message`The whisper model to use`
       }),
       "tiny.en"

package/dist/transcribe/parse.js

@@ -16,7 +16,7 @@ import {
 } 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 { WHISPER_MODELS } from "@storyteller-platform/ghost-story/constants";
 import {
   languageParser,
   loggingParser,

package/dist/transcribe/transcribe.cjs

@@ -140,6 +140,8 @@ async function transcribe(input, output, locale, options) {
           });
           options.logger?.info(`Found existing transcription for ${filepath}`);
           transcriptions.push(transcriptionFilepath);
+          options.onProgress?.((transcriptions.length + 1) / filenames.length);
+          return;
         } catch {
         }
         if (aborted()) throw new Error("Aborted");

package/dist/transcribe/transcribe.d.cts

@@ -1,6 +1,7 @@
 import * as _storyteller_platform_ghost_story from '@storyteller-platform/ghost-story';
-import { RecognitionEngine, WhisperModel, TimingAggregator } from '@storyteller-platform/ghost-story';
+import { TimingAggregator } from '@storyteller-platform/ghost-story';
 import { Logger } from 'pino';
+import { RecognitionEngine, WhisperModel } from '@storyteller-platform/ghost-story/constants';
 
 type WhisperCpuOverride = "blas" | "cpu" | null;
 interface TranscribeOptions {

package/dist/transcribe/transcribe.d.ts

@@ -1,6 +1,7 @@
 import * as _storyteller_platform_ghost_story from '@storyteller-platform/ghost-story';
-import { RecognitionEngine, WhisperModel, TimingAggregator } from '@storyteller-platform/ghost-story';
+import { TimingAggregator } from '@storyteller-platform/ghost-story';
 import { Logger } from 'pino';
+import { RecognitionEngine, WhisperModel } from '@storyteller-platform/ghost-story/constants';
 
 type WhisperCpuOverride = "blas" | "cpu" | null;
 interface TranscribeOptions {

package/dist/transcribe/transcribe.js

@@ -71,6 +71,8 @@ async function transcribe(input, output, locale, options) {
           });
           options.logger?.info(`Found existing transcription for ${filepath}`);
           transcriptions.push(transcriptionFilepath);
+          options.onProgress?.((transcriptions.length + 1) / filenames.length);
+          return;
         } catch {
         }
         if (aborted()) throw new Error("Aborted");

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@storyteller-platform/align",
-  "version": "0.1.20",
+  "version": "0.1.22",
   "description": "A library and CLI for automatically aligning audiobooks and EPUBs to produce Media Overlays",
   "author": "Shane Friedman",
   "license": "MIT",
@@ -60,8 +60,8 @@
     "@optique/core": "^0.10.7",
     "@optique/run": "^0.10.7",
     "@storyteller-platform/audiobook": "^0.3.10",
-    "@storyteller-platform/epub": "^0.4.8",
-    "@storyteller-platform/ghost-story": "^0.1.7",
+    "@storyteller-platform/epub": "^0.4.9",
+    "@storyteller-platform/ghost-story": "^0.1.8",
     "@storyteller-platform/transliteration": "^3.1.2",
     "chalk": "^5.4.1",
     "cli-progress": "^3.12.0",