Lightweight International Phonetic Alphabet (IPA) phonemizer that uses libespeak-ng
Project description
eSpeak Phonemizer
Uses ctypes and libespeak-ng to transform text into IPA phonemes.
Installation
First, install libespeak-ng:
sudo apt-get install libespeak-ng1
Next, install espeak_phonemizer:
pip install espeak_phonemizer
If installation was successful, you should be able to run:
espeak-phonemizer --version
Basic Phonemization
Simply pass your text into the standard input of espeak-phonemizer
:
echo 'This is a test.' | espeak-phonemizer -v en-us
ðɪs ɪz ɐ tˈɛst
Separators
Phoneme and word separators can be changed:
echo 'This is a test.' | espeak-phonemizer -v en-us -p '_' -w '#'
ð_ɪ_s#ɪ_z#ɐ#t_ˈɛ_s_t
Punctuation and Stress
Some punctuation can be kept (.,;:!?) in the output:
echo 'This: is, a, test.' | espeak-phonemizer -v en-us --keep-punctuation
ðˈɪs: ˈɪz, ˈeɪ, tˈɛst.
Stress markers can also be dropped:
echo 'This is a test.' | espeak-phonemizer -v en-us --no-stress
ðɪs ɪz ɐ tɛst
Delimited Input
The --csv
flag enables delimited input with fields separated by a '|' (change with --csv-delimiter
):
echo 's1|This is a test.' | espeak-phonemizer -v en-us --csv
s1|This is a test.|ðɪs ɪz ɐ tˈɛst
Phonemes are added as a final column, allowing you to pass arbitrary metadata through to the output.
Parallelize with GNU Parallel
parallel -a /path/to/input.csv --pipepart \
espeak-phonemizer -v en-us --csv \
> /path/to/output.csv
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.