pytrovich

pytrovich: a Python port of an inflector for Russian anthroponyms developed by petrovich team: https://github.com/petrovich


Keywords
nlp, morphology, russian, language, inflection, natural-language-processing, python, russian-language
License
MIT
Install
pip install pytrovich==0.0.2

Documentation

Pytrovich

pytrovich is a Python 3.6+ port of petrovich library which inflects Russian names to a given grammatical case. It supports first names, last names and middle names inflections. Since version 0.0.2, gender detection is also available.

petrovich-java was the main inspiration.

The alternative (earlier) port: Petrovich (@alexeyev was not aware of it at the time of porting petrovich to Python). The only meaningful difference we have found is that it does not support gender detection.

Python 3x PyPI version Downloads

Installation

Should be as simple as that

pip install pytrovich

Usage

Inflection

from pytrovich.enums import NamePart, Gender, Case
from pytrovich.maker import PetrovichDeclinationMaker

maker = PetrovichDeclinationMaker()
print(maker.make(NamePart.FIRSTNAME, Gender.MALE, Case.GENITIVE, "Иван"))  # Ивана
print(maker.make(NamePart.LASTNAME, Gender.MALE, Case.INSTRUMENTAL, "Иванов"))  # Ивановым
print(maker.make(NamePart.MIDDLENAME, Gender.FEMALE, Case.DATIVE, "Ивановна"))  # Ивановне

Gender detection

from pytrovich.detector import PetrovichGenderDetector

detector = PetrovichGenderDetector()
print(detector.detect(firstname="Иван"))  # Gender.MALE
print(detector.detect(firstname="Иван", middlename="Семёнович"))  # Gender.MALE
print(detector.detect(firstname="Арзу", middlename="Лутфияр кызы"))  # Gender.FEMALE

Custom rule file

You can replace default rules file with some custom one. Only JSON format is supported.

maker = PetrovichDeclinationMaker("/path/to/custom/rules.file.json")

E.g. if pytrovich fails on PetrovichDeclinationMaker creation, one may consider downloading rules.json directly from petrovich-rules repo as a fix (please create an issue if that actually happens).

How to cite

Not neccessary, but greatly appreciated, if you use this work.

@misc{Pytrovich,
  title     = {{petrovich/pytrovich: Python3 port of Petrovich, an inflector for Russian anthroponyms}},
  year      = {2020},
  url       = {https://github.com/petrovich/pytrovich},
  language  = {english},
}

More info

For more information on the project please refer to other petrovich repos.

TODO

  • efficiency was not a top priority, the time has come for faster algorithms, RegEx and data structures
  • evaluation based on petrovich-eval

License

This project is available under MIT license.