Amazon Polly Snippet JS

polly.ts

import 'dotenv/config'
import {
  PollyClient,
  SynthesizeSpeechCommand,
} from '@aws-sdk/client-polly'
import fs from 'fs'
import read from '@nerdbond/read'
import TERM_MAP from '~/configurations/terms'

// a client can be shared by different commands.
const client = new PollyClient({
  credentials: {
    accessKeyId: process.env.AWS_ACCESS_KEY,
    secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY,
  },
  region: 'us-east-1',
})

;(async () => {
  for (const term in TERM_MAP) {
    const path = `public/words/${term.replace(/C/g, '$')}.mp3`
    if (fs.existsSync(path)) {
      continue
    }
    console.log(term)
    const vowelMatch = term.match(/[aeiou]/g)
    const vowelAtEnd = term.match(/[aeiou]$/)
    let text = term
    if (vowelMatch?.length == 3) {
      text = term.replace(
        /([aeiou])(.+[aeiou])(.+[aeiou])/,
        (_, $1, $2, $3) => `${$1}${$2}^${$3}`,
      )
    } else if (vowelMatch?.length == 2) {
      if (vowelAtEnd) {
        text = term.replace(
          /([aeiou])(.+[aeiou])/,
          (_, $1, $2) => `${$1}^${$2}`,
        )
      } else {
        text = term.replace(
          /([aeiou])(.+[aeiou])/,
          (_, $1, $2) => `${$1}${$2}^`,
        )
      }
    }
    text = `<phoneme alphabet="ipa" ph="${
      read.talk.toIPA(text).replace(/ɔ/g, 'a')
      // .replace(/r/, 'ɾ')
    }">${text}</phoneme>`

    const input = {
      Engine: 'neural',
      // LanguageCode: 'en-ZA',
      LanguageCode: 'en-US',
      OutputFormat: 'mp3', // required
      Text: `<speak>${text}</speak>`,
      TextType: 'ssml',
      // VoiceId: 'Arlet',
      // VoiceId: 'Ayanda',
      // VoiceId: 'Elin',
      VoiceId: 'Salli',
      // VoiceId: 'Salli',
    }
    const command = new SynthesizeSpeechCommand(input)
    const response = await client.send(command)
    const bytes = await response.AudioStream?.transformToByteArray()
    if (bytes) {
      fs.writeFileSync(path, bytes)
    }
  }
})()

const keys = fs
  .readdirSync('public/words')
  .filter(x => x.endsWith('.mp3'))
  .map(x => `  ${x.split('.')[0]}: '/words/${x}',`)
  .sort()

fs.writeFileSync(
  'data/termSound.ts',
  `const SOUND_MAP = {
${keys.join('\n')}
}

export default SOUND_MAP`,
)