Extract chapter marks from M4A/MP4 files as Podlove Simple Chapters

extract-chapters-from-m4a-as-psc.js

#!/usr/bin/env node
'use strict';
const cheerio = require('cheerio');
const child_process = require('child_process');
const fs = require('fs');
const meow = require('meow');
const MP4Box = require('mp4box');
const path = require('path');
const tmp = require('tmp');
const groupBy = require('lodash.groupby');
const xmlbuilder = require('xmlbuilder');
const replaceExt = require('replace-ext');

// Init cli helper
const cli = meow(`
  Usage
    $ extract-chapters-from-m4a-as-psc <input> [<output>]

  Examples
    $ extract-chapters-from-m4a-as-psc input.m4a output.psc
    🔖 23 chapters extracted to output.psc
`);

// Render help if input is missing
if (!cli.input[0]) {
  console.log(cli.help);
  process.exit(0);
}

// Input
const input = path.resolve(cli.input[0]);

// Output
const output = cli.input[1] ? path.resolve(cli.input[1]) : replaceExt(input, '.psc');

// Load m4a/mp4 file
const mp4box = new MP4Box.MP4Box();
const file = fs.readFileSync(input);
const arrayBuffer = new Uint8Array(file).buffer;
arrayBuffer.fileStart = 0;
mp4box.appendBuffer(arrayBuffer);
const info = mp4box.getInfo();

// Extract tx3g tracks
const chapterMarks = info.tracks
  // Filter by type
  .filter(t => t.codec === 'tx3g')

  // Extract XML
  .map(t => {
    // Get temp file
    const tmpFile = tmp.fileSync();

    // Run mp4box
    // TODO: Replace with JS
    child_process.execSync(`mp4box -ttxt ${t.id} ${input} -out ${tmpFile.name}`, {
      stdio: 'ignore',
    });

    // Extract XML from tmp file
    const xmlFile = fs.readFileSync(tmpFile.name);
    const $xml = cheerio.load(xmlFile.toString('utf8'), {
      decodeEntities: true,
      normalizeWhitespace: true,
      xmlMode: true,
    });

    // Chapters
    const chapters = $xml('TextSample').map((index, ts) => {
      // Get chapter object
      const $ts = $xml(ts);
      const chapter = {
        start: $ts.attr('sampleTime'),
        title: $ts.text().trim(),
      };

      // Check if href exists
      if ($ts.find('HyperLink').length > 0) {
        chapter.href = $ts.find('HyperLink').attr('URL').trim();
      }

      // Return chapter JSON
      return chapter;
    }).get().filter(c => c.title !== '');
    
    // Remove tmp file
    tmpFile.removeCallback();

    // Return chapters
    return chapters;
  })

  // Reduce to one chapter mark object
  .reduce((prevCm, cm) => {
    const chapters = [].concat(prevCm, cm);
    const groupedByTime = groupBy(chapters, o => o.start);
    const output = [];
    
    Object.keys(groupedByTime).forEach(t => {
      output.push(Object.assign.apply(Object.assign, groupedByTime[t]));
    });

    return output;
  }, []);

// Generate Podlove Simple Chapters
const psc = xmlbuilder.create('psc:chapters', { encoding: 'utf-8' })
  .att('version', '1.1')
  .att('xmlns:psc', 'http://podlove.org/simple-chapters');

// Add entries
chapterMarks.forEach(cm => psc.ele('psc:chapter', cm).up());

// Save PSC file
fs.writeFileSync(output, psc.end({ pretty: true }));

// Ready.
console.log(`🔖 ${chapterMarks.length} chapters extracted to ${output}`);

package.json

{
  "name": "extract-chapters-from-m4a-as-psc",
  "version": "0.1.0",
  "description": "Extract chapter marks from M4A/MP4 files as Podlove Simple Chapters",
  "main": "extract-chapters-from-m4a-as-psc.js",
  "author": "Dennis Morhardt <info@dennismorhardt.de>",
  "license": "MIT",
  "bin": {
    "extract-chapters-from-m4a-as-psc": "extract-chapters-from-m4a-as-psc.js"
  },
  "dependencies": {
    "cheerio": "^0.22.0",
    "lodash.groupby": "^4.6.0",
    "meow": "^3.7.0",
    "mp4box": "^0.3.15",
    "replace-ext": "^1.0.0",
    "tmp": "^0.0.31",
    "xmlbuilder": "^9.0.0"
  }
}