A Python package for bi-directional transliteration of Cyrillic script to Latin script and vice versa.
By default, transliterates for the Serbian language. A language flag can be set in order to transliterate to and from Bulgarian, Montenegrin, Macedonian, Mongolian, Russian, Serbian, Tajik, and Ukrainian.
Transliteration is the conversion of a text from one script to another. For instance, a Latin alphabet transliteration of the Serbian phrase "Мој ховеркрафт је пун јегуља" is "Moj hoverkraft je pun jegulja".
A citation would be much appreciated if you use CyrTranslit in a research publication:
Georges Labrèche. (2023). CyrTranslit (v1.1.1). Zenodo. https://doi.org/10.5281/zenodo.7734906
BibTex entry:
@software{georges_labreche_2023_7734906,
author = {Georges Labrèche},
title = {CyrTranslit},
month = mar,
year = 2023,
note = {{A Python package for bi-directional
transliteration of Cyrillic script to Latin script
and vice versa. Supports transliteration for
Bulgarian, Montenegrin, Macedonian, Mongolian,
Russian, Serbian, Tajik, and Ukrainian.}},
publisher = {Zenodo},
version = {v1.1.1},
doi = {10.5281/zenodo.7734906},
url = {https://doi.org/10.5281/zenodo.7734906}
}
CyrTranslit is actively used as a reliable tool to advance research! Here's an incomplete list of publications for research projects that have relied on CyrTranslit:
- Ljajić, Adela & Prodanović, Nikola & Medvecki, Darija & Bašaragin, Bojana & Mitrović, Jelena. (2022). "Topic Modeling Technique on Covid19 Tweets in Serbian," in 12th International Conference on Information Society and Technology (ICIST), Kopaonik, Serbia.
- Mussylmanbay, Meiirgali. (2022). "Addresses Standardization and Geocoding using Natural Language Processing," Nazarbayev University, Kazakhstan.
- Jokic, Danka & Stanković, Ranka & Krstev, Cvetana & Šandrih Todorović, Branislava. (2021). "A Twitter Corpus and Lexicon for Abusive Speech Detection in Serbian," in 3rd Conference on Language, Data and Knowledge (LDK 2021). 10.4230/OASIcs.LDK.2021.13.
- Lakew, Surafel Melaku (2020). "Thesis Multilingual Neural Machine Translation for Low Resource Languages," University of Trento, Italy.
- Filo, Denis. (2020). "Neuronový strojový překlad pro jazykové páry s malým množstvím trénovacích dat: Low-Resource Neural Machine Translation," Brno University of Technology, Brno, Czechia.
- Batanović, Vuk & Nikolic, Bosko. (2019). "Using Language Technologies to Automate the UNDP Rapid Integrated Assessment Mechanism in Serbian," in International Conference on Language Technologies for All: Enabling Linguistic Diversity and Multilingualism Worldwide (LT4All), Paris, France.
- Brown, J. M. M. & Schmidt, Andreas & Wierzba, Marta (Eds.). (2019). "Of trees and birds: A Festschrift for Gisbert Fanselow," Universitätsverlag Potsdam, Potsdam.
- Lakew, Surafel Melaku & Erofeeva, Aliia & Federico, Marcello. (2018). "Neural Machine Translation into Language Varieties," in 3rd Conference on Machine Translation: Research Papers, Brussels, Belgium.
- Ljajić, Adela & Marovac, Ulfeta. (2018). "Improving sentiment analysis for twitter data by handling negation rules in the Serbian language," Computer Science and Information Systems. 16. 13-13. 10.2298/CSIS180122013L.
- Жабран, И., Кикоть, А., Гафияк, А., Бородина, Е., & Алёшин, С. (2017). "Developing Q-Orca site backend using various Python programming language libraries," Modern Engineering and Innovative Technologies, 3(07-03), 48–53.
CyrTranslit is hosted in the Python Package Index (PyPI) so it can be installed using pip:
python3 -m pip install cyrtranslit # latest version
python3 -m pip install cyrtranslit==1.1.2 # specific version
python3 -m pip install cyrtranslit>=1.1.2 # minimum version
CyrTranslit currently supports bi-directional transliteration of Bulgarian, Montenegrin, Macedonian, Mongolian, Russian, Serbian, Tajik, and Ukrainian:
>>> import cyrtranslit
>>> cyrtranslit.supported()
['bg', 'me', 'mk', 'mn', 'ru', 'sr', 'tj', 'ua']
CyrTranslit can be used both programatically and via command line interface.
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Съединението прави силата!", "bg")
"Săedinenieto pravi silata!"
>>> cyrtranslit.to_cyrillic("Săedinenieto pravi silata!", "bg")
"Съединението прави силата!"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Република", "me")
"Republika"
>>> cyrtranslit.to_cyrillic("Republika", "me")
"Република"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Моето летачко возило е полно со јагули", "mk")
"Moeto letačko vozilo e polno so jaguli"
>>> cyrtranslit.to_cyrillic("Moeto letačko vozilo e polno so jaguli", "mk")
"Моето летачко возило е полно со јагули"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Амрагаа Сүнжидмаагаа гэсээр ирлээ дээ хө-хө-хө", "mn")
"Amragaa Sünjidmaagaa geseer irlee dee khö-khö-khö"
>>> cyrtranslit.to_cyrillic("Amragaa Sünjidmaagaa geseer irlee dee khö-khö-khö", "mn")
"Амрагаа Сүнжидмаагаа гэсээр ирлээ дээ хө-хө-хө"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Моё судно на воздушной подушке полно угрей", "ru")
"Moyo sudno na vozdushnoj podushke polno ugrej"
>>> cyrtranslit.to_cyrillic("Moyo sudno na vozdushnoj podushke polno ugrej", "ru")
"Моё судно на воздушной подушке полно угрей"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Мој ховеркрафт је пун јегуља")
"Moj hoverkraft je pun jegulja"
>>> cyrtranslit.to_cyrillic("Moj hoverkraft je pun jegulja")
"Мој ховеркрафт је пун јегуља"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Ман мактуб навишта истодам", "tj")
"Man maktub navišta istodam"
>>> cyrtranslit.to_cyrillic("Man maktub navišta istodam", "tj")
"Ман мактуб навишта истодам"
>>> import cyrtranslit
>>> cyrtranslit.to_latin("Під лежачий камінь вода не тече", "ua")
"Pid ležačyj kamin' voda ne teče"
>>> cyrtranslit.to_cyrillic("Pid ležačyj kamin' voda ne teče", "ua")
"Під лежачий камінь вода не тече"
Sample command line call to transliterate a Russian text file:
$ cyrtranslit -l RU -i tests/ru.txt -o tests/output.txt
Use the -c argument to accomplish the reverse, that is to input latin characters and output cyrillic.
Use the -h argument for help.
You can also omit the input and output files and use standard input/output
$ echo 'Мој ховеркрафт је пун јегуља' | cyrtranslit -l sr
Moj hoverkraft je pun jegulja
$ echo 'Moj hoverkraft je pun jegulja' | cyrtranslit -l sr
Мој ховеркрафт је пун јегуља
You can test the "script" by running it directly on the Python command line interface, e.g.:
>>> import sys
>>> import cyrtranslit.cyrtranslit
>>> sys.argv.extend(['-l', 'UA'])
>>> sys.argv.extend(['-i', 'tests/ua.txt'])
>>> sys.argv.extend(['-o', 'tests/output.txt'])
>>> cyrtranslit.cyrtranslit.main()
>>> exit()
You can include support for other Cyrillic script alphabets. Follow these steps in order to do so:
- Create a new transliteration dictionary in the mapping.py file and reference to it in the TRANSLIT_DICT dictionary.
- Watch out for cases where two consecutive Latin alphabet letters are meant to transliterate into a single Cyrillic script letter. These cases need to be explicitly checked for inside the to_cyrillic() function in __init__.py.
- Add test cases inside of tests.py.
- Add test CLI input files in the tests directory.
- Update the documentation in the README.md.
- List yourself as one of the contributors.
Before tagging a release version and deploying to PyPI:
- Update the
version
anddownload_url
properties in setup.py. - Reserve a Zenodo DOI for the release and update this readme's Zenodo badge and citation instructions.
A big thank you to everyone who contributed:
- Bulgarian 🇧🇬: @Syndamia and @Sparkycz.
- Russian 🇷🇺: @ratijas and @rominf.
- Tajik 🇹🇯: @diejani.
- Ukrainian 🇺🇦: @AnonymousVoice1.
- Mongolian 🇲🇳: @Serbipunk.
- Command Line Interface (CLI): @ZJaume.