pysle

Questions? Comments? Feedback?

Pronounced like 'p' + 'isle'.

An interface to a pronunciation dictionary with stress markings (ISLEX - the international speech lexicon), along with some tools for working with comparing and aligning pronunciations (e.g. a list of phones someone said versus a standard or canonical dictionary pronunciation).

Table of contents

1. Documentation
2. Common Use Cases
3. Version History
4. Requirements
5. Optional resources
6. Installation
7. Example usage
8. Citing psyle
9. Acknowledgements

Documentation

Automatically generated pdocs can be found here:

http://timmahrt.github.io/pysle/

The documentation is generated with the following command: pdoc ./pysle -d google -o docs

Common Use Cases

What can you do with this library?

• look up the list of phones and syllables for canonical pronunciations of a word

  isletool.LexicalTool('ISLEdict.txt').lookup('cat')

• map an actual pronunciation to a dictionary pronunciation (can be used to automatically find speech errors)

  pronunciationtools.findClosestPronunciation(isleDict, 'cat', ['k', 'æ',])

• automatically syllabify a praat textgrid containing words and phones (e.g. force-aligned text) -- requires the praatIO library

  pysle.syllabifyTextgrid(isleDict, praatioTextgrid, "words", "phones")

• search for words based on pronunciation

  isletool.LexicalTool('ISLEdict.txt').search('dVV') # Any word containing a 'd' followed by two vowels

  e.g. Words that start with a sound, or have a sound word medially, or in stressed vowel position, etc.

  see /tests/dictionary_search.py

Version History

Pysle uses semantic versioning (Major.Minor.Patch)

Please view CHANGELOG.md for version history.

Requirements

• Python 3.7.* or above (or below, probably)

Click here to visit travis-ci and see the specific versions of python that pysle is currently tested under

• The praatIO library is required IF you want to use the textgrid functionality. It is not required for normal use.

ISLE Dictionary

pysle requires the ISLEdict pronunciation dictionary (copyright Mark Hasegawa-Johnson, licensed under the MIT open source license). This is bundled with psyle. However, you may want to use a subset of the pronunciations or you may want to add your own pronunciations.

In that case, please get the original file.

ISLEX github page

Direct link to the ISLEX file used in this project (ISLEdict.txt)

See examples/isletool_examples.py for an example of how to load a custom ISLEdict file.

Installation

Pysle is on pypi and can be installed or upgraded from the command-line shell with pip like so

python -m pip install pysle --upgrade

Otherwise, to manually install, after downloading the source from github, from a command-line shell, navigate to the directory containing setup.py and type

python setup.py install

If python is not in your path, you'll need to enter the full path e.g.

C:\Python36\python.exe setup.py install

Example usage

Here is a typical usage

from pysle import isletool
isleDict = isletool.LexicalTool('C:\islev2.dict')
print(isleDict.lookup('catatonic')[0]) # Get the first pronunciation
# >> (([['k', 'ˌæ'], ['ɾ', 'ə'], ['t', 'ˈɑ'], ['n', 'ɪ', 'k']], [2, 0], [1, 1]),)

and another

from pysle import isletool
from pysle import pronunciationtools

isleDict = isletool.LexicalTool('C:\islev2.dict')

searchWord = 'another'
phoneList = ['n', '@', 'th', 'r'] # Actually produced (ASCII or IPA ok here)

returnList = pronunciationtools.findBestSyllabification(isleDict, searchWord, phoneList)
syllableList = returnList[2]
print(syllableList)
# >> [["''"], ['n', '@'], ['th', 'r']]

Please see \examples for example usage

Citing pysle

Pysle is general purpose coding and doesn't need to be cited (you should cite the ISLEX project instead) but if you would like to, it can be cited like so:

Tim Mahrt. Pysle. https://github.com/timmahrt/pysle, 2016.

Acknowledgements

Development of Pysle was possible thanks to NSF grant IIS 07-03624 to Jennifer Cole and Mark Hasegawa-Johnson, NSF grant BCS 12-51343 to Jennifer Cole, José Hualde, and Caroline Smith, and to the A*MIDEX project (n° ANR-11-IDEX-0001-02) to James Sneed German funded by the Investissements d'Avenir French Government program, managed by the French National Research Agency (ANR).