CMUSphinxDict

Wrapper for CMU Sphinx Pronouncing Dictionary. It's a hashtable that maps words to phonemes.

Install

npm install cmusphinxdict

Usage

Single word

CMUSphinxDict.get('hello', function(word, pronouncings) {
  console.log(word); // 'HELLO'
  console.log(pronouncings); // ['HH AH L OW', 'HH EH L OW']
});

Multiple words

CMUSphinxDict.get(['hello', 'world'], function(words, pronouncings) {
  console.log(words); // ['HELLO', 'WORLD']
  console.log(pronouncings); // [['HH AH L OW', 'HH EH L OW'], ['W ER L D']]
});

Special character lookup

CMUSphinxDict.get('.', function(word, pronouncings) {
  console.log(word); // '.'
  console.log(pronouncings); // [ 'D EH S AH M AH L', 'D AA T', 'P IH R IY AH D', 'P OY N T' ]
});

Add additional pronouncings

additionalPronouncings.json

[
  {
    "AMALTHEA": [
      "AH M AA L TH IY AH"
    ]
  }
]

var file = fs.createReadStream(__dirname + '/additionalPronouncings.json');

CMUSphinxDict.addPronouncings(file, function(error, words, pronouncings) {
  console.log(words); // ['AMALTHEA']
  console.log(pronouncings); // [['AH M AA L TH IY AH']]
});

Command line

$ cmusphinxdict hello world

HELLO HH AH L OW
HELLO(2) HH EH L OW
WORLD W ER L D

JSON format:

$ cmusphinxdict hello world --format=json

{
  "HELLO": [
    "HH AH L OW",
    "HH EH L OW"
  ],
  "WORLD": [
    "W ER L D"
  ]
}

Aliases for format: -f=[type], --format=[type], --json

Format types: json, text (default)

License

CMU Pronouncing Dictionary is release under the Public Domain.

Everything else is released under the MIT License.