wassname / pysle

Python interface to ISLEX, an English IPA pronunciation dictionary with syllable and stress marking.

pysle

Pronounced like 'p' + 'isle'.

An interface for the ILSEX (international speech lexicon) dictionary, 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).

Contents

• 1   Common Use Cases
• 2   Major revisions
• 3   Requirements
• 4   Installation
• 5   Example usage
• 6   Citing pysle
• 7   Acknowledgements

1   Common Use Cases

What can you do with this library?

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

  pysle.isletool.LexicalTool.lookup('cat')
  

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

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

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

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

• search for words based on pronunciation:

  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
  

2   Major revisions

Ver 1.4 (July 9, 2016)

• added search functionality
• ported code to use the new unicode IPA-based isledict (the old one was ascii)

Ver 1.3 (March 15, 2016)

• added indicies for stressed vowels

Ver 1.2 (June 20, 2015)

• Python 3.x support

Ver 1.1 (January 30, 2015)

• word lookup ~65 times faster

Ver 1.0 (October 23, 2014)

• first public release.

3   Requirements

• Before you use this library (before or after installing it) you will need to download the ILSEX dictionary. It can be downloaded here under the section 'English' linked under the text 'English Pronlex' (with a file name of ISLEdict.txt):

  ISLEX project page

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

• Python 2.7.* or above

• Python 3.3.* or above

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

4   Installation

If you on Windows, you can use the installer found here (check that it is up to date though) Windows installer

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:\Python27\python.exe setup.py install

5   Example usage

Here is a typical common usage:

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

and another:

from pysle import isle
from psyle import pronunciationTools

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

returnList = pronunciationTools.findBestSyllabification(isleDict,
                                                        searchWord,
                                                        anotherPhoneList)
print syllableList
>> [["''"], ['n', '@'], ['th', 'r']]

Please see \examples for example usage

6   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.

7   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).