PHONETIC

A C++ library for working with the CMU Pronouncing Dictionary

Intro

The CMU Pronouncing Dictionary (CMUdict) is

an open-source machine-readable pronunciation dictionary for North American English that contains over 134,000 words and their pronunciations.

There's various tooling built around it, e.g. this lovely Python library. But I was working in C++ and WebAssembly, and ended up writing my own interface for using CMUdict.

Build

You can include use this in your own projects: you'll need the header file (include/phonetic.hpp), the source file (src/phonetic.cpp), and the CMUdict data, which is included as a sub-module of this repo, and which phonetic.cpp expects to find at ../data/CMUdict/cmudict-0.7b, relative to itself.

This can also be compiled with CMake:

mkdir build
cd build
cmake ..
make

Which will also build a build/tests/test_phonetic Catch-2 test file.

This library can also be compiled to WebAssembly using Emscripten:

mdkir build
cd build
emcmake cmake ..
emmake make

This will generate a phonetic.js, phonetic.wasm, and phonetic.data. Using these you'll be able to call any of the Phonetic class methods straight from Javascript.

Usage

Use this library to convert English words into:

1. Possible pronunciations, as ARPABET which encodes IPA into two letter ASCII sequences.
2. Possible patterns of syllabic stress, as strings of numbers 0, 1, & 2, where 0 is unstressed, 1 is primary stress, and 2 is secondary stress.
3. Possible syllable counts (counting the number of vowel phones).

Note that if a word has multiple pronunciations, stress patterns, or syllable counts, all of these will be returned.

The tests (test/test_phonetic.cpp) offers a chance to see all of the methods in action.

Future Features

• Search by phones.
• Search by stress.
• Search by syllable count.
• More useful WebAssembly error handling.