This is a tool to create a pickle database (https://pythonhosted.org/pickleDB/) from CMU-IPA Dictionary.
The pickle database has key-value pairs of English words and phonetic transcriptions. A single word may have multiple transcriptions. For this reason, values are arrays.
pip install -r requirements.txt
The code assumus the CMU-IPA Dictionary is in this top directory as CMU.in.IPA.txt
.
The pickle database name can be given. If none is given, the default name is cmu_ipa.pickle
.
$ ./createdb.py -h
usage: createdb.py [-h] [-d DBNAME]
Creates picledb from CMU IPA dict
optional arguments:
-h, --help show this help message and exit
-d DBNAME, --dbname DBNAME
pickledb file. if not given, default name will be used
The dictionary is from http://people.umass.edu/nconstan/CMU-IPA/. The dictionary has around 130,000 words and those American phonetic transcription.