Skip to main content

yiddish

A Python library for processing Yiddish text

by Isaac L. Bleaman (bleaman@berkeley.edu)

See the commit history for additional contributors.

What is this for?

This library includes functions to carry out common tasks when dealing with Yiddish text. For example, you might wish to replace precomposed Unicode characters (such as אַ, U+FB2E) with their decomposed versions (אַ, which is U+05D0 followed by U+05B7). Or you might wish to transliterate YIVO Yiddish text (איבער to iber) or render it in the orthography used more commonly in the Hasidic community (שנײיִק to שנייאיג).

See Functions below for the full list.

How to install

pip install yiddish

The package requires Python 3.9 or later and has no dependencies.

Note: The wordlists the package relies on are bundled into the releases on PyPI, but they are not stored in this repository. If you wish to install from the source, clone the repository, run bash fetch_data.sh to download the wordlists, and then run pip install .. (Installing directly with pip install git+https://github.com/ibleaman/yiddish will not include them, and the package will fail to import.)

License and data

The library is released under the MIT License (see LICENSE.md).

The wordlists come from two separate repositories, loshn-koydesh-pronunciation and hasidify_lexicon, and are bundled into each release. The loshn-koydesh lexicon was inspired by the work of lexicographer Yitskhok Niborski and expanded and adapted from a wordlist put together by Eliezer Niborski; see its repository for the sources. The hasidify list was initially created by me for personal use during my doctoral research, based on my own intuitions about common spelling correspondences.

How to cite

If you'd like to cite yiddish in a publication, please include a link to https://github.com/ibleaman/yiddish or https://pypi.org/project/yiddish/

Example

import yiddish

output = ''

string = 'אונדזער גאַנצע משפּחה װױנט אין די פֿאַראײניקטע שטאַטן.'

output += yiddish.replace_with_precomposed(string) + '\n'
output += yiddish.respell_loshn_koydesh(string) + '\n'
output += yiddish.strip_diacritics(string) +  '\n'
output += yiddish.transliterate(string) +  '\n'
output += yiddish.transliterate(string, loshn_koydesh=True) +  '\n'
output += yiddish.hasidify(string)

output += '\n\n'

string_two = 'shloymele hot khasene gehat mit rokhls tokhter leye.'

output += yiddish.detransliterate(string_two) + '\n'
output += yiddish.detransliterate(string_two, loshn_koydesh=True)

print(output)

Output:

אונדזער גאַנצע משפּחה װױנט אין די פֿאַראײניקטע שטאַטן.
אונדזער גאַנצע מישפּאָכע װױנט אין די פֿאַראײניקטע שטאַטן.
אונדזער גאנצע משפחה וווינט אין די פאראייניקטע שטאטן.
undzer gantse mshpkhh voynt in di fareynikte shtatn.
undzer gantse mishpokhe voynt in di fareynikte shtatn.
אונזער גאנצע משפחה וואוינט אין די פאראייניגטע שטאטן.

שלױמעלע האָט כאַסענע געהאַט מיט ראָכלס טאָכטער לײע.
שלמהלע האָט חתונה געהאַט מיט רחלס טאָכטער לאה.

Functions

All functions take a string and return a string. Input may be in precomposed (e.g., אַ, U+FB2E) or decomposed (אַ, U+05D0 + U+05B7) form, except where noted.

Unicode and punctuation

replace_with_precomposed(string): replace decomposed characters with precomposed ones. (replace_with_precombined is an older name for the same function and still works.) Note that characters such as אַ (U+FB2E) belong to Unicode's Alphabetic Presentation Forms block and are excluded from standard NFC normalization, so unicodedata.normalize('NFC', ...) will not produce them; this function will.

>>> yiddish.replace_with_precomposed('\u05d0\u05b7')
'אַ'

>>> len(yiddish.replace_with_precomposed('\u05d0\u05b7'))
1

replace_with_decomposed(string, vov_yud=False): replace precomposed characters with decomposed ones. With vov_yud=True, the ligatures װ, ױ, ײ are left alone.

>>> yiddish.replace_with_decomposed('\ufb2e')
'אַ'

>>> len(yiddish.replace_with_decomposed('\ufb2e'))
2
>>> yiddish.replace_with_decomposed('װױנט')
'וווינט'

replace_punctuation(string): use the punctuation conventional for (at least much of) Yiddish text on the web: ASCII hyphens become makef, and geresh/gershayim/prime variants become ' and ".

>>> yiddish.replace_punctuation('ר׳ יוסף ז״ל')
'ר\' יוסף ז"ל'

strip_diacritics(string): remove nekudes/pointing and ligatures from the text.

>>> yiddish.strip_diacritics('די פֿאַראײניקטע שטאַטן')
'די פאראייניקטע שטאטן'

Romanization

transliterate(string, loshn_koydesh=False, loc=False): romanize Yiddish text using the YIVO system. With loc=True, a Library of Congress-style system is used instead. With loshn_koydesh=True, Hebrew- and Aramaic-origin words are first respelled as they are pronounced (see respell_loshn_koydesh).

>>> yiddish.transliterate('איבער')
'iber'
>>> yiddish.transliterate('פֿאַראײניקטע', loc=True)
'fareyniḳṭe'
>>> yiddish.transliterate('די משפּחה מאַשבער', loshn_koydesh=True)
'di mishpokhe mashber'

detransliterate(string, loshn_koydesh=False): convert YIVO-style romanization back to Yiddish script. With loshn_koydesh=True, Hebrew- and Aramaic-origin words are given their traditional spelling. The output uses precomposed Unicode characters (but you can run replace_with_decomposed on the output if desired).

>>> yiddish.detransliterate('shloymele hot khasene gehat')
'שלױמעלע האָט כאַסענע געהאַט'
>>> yiddish.detransliterate('shloymele hot khasene gehat', loshn_koydesh=True)
'שלמהלע האָט חתונה געהאַט'

romanize_german(text): (also available as romanise_german) a German-style romanization (originally designed for use with the aeneas audio-text synchronizer). Unlike the other functions, this expects precomposed input.

>>> yiddish.romanize_german('װען װעלן מיר זיך שפּילן צוזאַמען?')
'wen welen mir sich spilen zusamen?'

Loshn-koydesh

These use the wordlists from loshn-koydesh-pronunciation.

respell_loshn_koydesh(text): respell Hebrew- and Aramaic-origin words as they are pronounced. The output uses precomposed Unicode.

>>> yiddish.respell_loshn_koydesh('משפּחה')
'מישפּאָכע'

spell_loshn_koydesh(text): the reverse: give Hebrew- and Aramaic-origin words their traditional spelling.

>>> yiddish.spell_loshn_koydesh('מישפּאָכע')
'משפּחה'

Orthographic variants

hasidify(text): render text in the spelling used more commonly in the Hasidic community.

>>> yiddish.hasidify('אונדזער הײליקע ייִדישע שפּראַך')
'אונזער הייליגע אידישע שפראך'

desovietify(text): convert Soviet Yiddish orthography to standard (YIVO) orthography.

>>> yiddish.desovietify('דאָס לאָשנ פונ דעמ באלעגאָלע')
'דאָס לשון פֿון דעם בעל־עגלה'

Metadata

Release files for yiddish 0.0.25

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for yiddish 0.0.25
File Size Uploaded Provenance Actions
yiddish-0.0.25.tar.gz 177.9 kB With provenance Details for yiddish-0.0.25.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for yiddish 0.0.25
File Interpreter ABI Platform Actions
yiddish-0.0.25-py3-none-any.whl Python 3 none any Details for yiddish-0.0.25-py3-none-any.whl

Total release size: 355.0 kB

Release files / yiddish-0.0.25.tar.gz

Download URL yiddish-0.0.25.tar.gz
Size 177.9 kB
Tags Source
SHA-256 checksum
How to use checksums
209795bf81d317956ab0144c68d58baedab41c8a7731b8b2a6a80736f4b01119
BLAKE2b-256 checksum
How to use checksums
9e83c46bd6d52bb02552dbeeca0154550470d97e192631f0a5452bcb2ce5a749
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 20, 2026.

Transparency log

Release files / yiddish-0.0.25-py3-none-any.whl

Download URL yiddish-0.0.25-py3-none-any.whl
Size 177.1 kB
Tags Python 3
SHA-256 checksum
How to use checksums
89c0e7515d217ced64c72a5f697c7370455f3ea46dc053b9b54919b315cbebb0
BLAKE2b-256 checksum
How to use checksums
da60ae892cb18699a1672b4827079939acf0eb5a617172475065748d971af2d5
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 20, 2026.

Transparency log

Security for yiddish 0.0.25

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for yiddish-0.0.25.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 20, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page