Disvortilo

Disvortilo is a simple tool that breaks Esperanto words into roots and affixes.

Getting Started

You can install Disvortilo from PyPI using pip:

pip install disvortilo

Examples

from disvortilo import Disvortilo, rank

disvortilo = Disvortilo()

print(disvortilo.parse("malliberejo"))
# > [('mal', 'liber', 'ej', 'o')]

# some have more than one possible output
print(disvortilo.parse("esperantistino"))
# > [('esper', 'ant', 'ist', 'in', 'o'), ('esperant', 'ist', 'in', 'o')]

# you can also get the morphemes along the their categories
print(disvortilo.parse_detailed("plibonigojn"))
# > [(('pli', WordPart.FULL_WORD), ('bon', WordPart.ROOT), ('ig', WordPart.SUFFIX), ('ojn', WordPart.POS))]

# sort the values from most likely to less likely
print(rank(disvortilo.parse_detailed("envias"), n=2))  # get the 2 most likely options
# > [(('envi', WordPart.ROOT), ('as', WordPart.POS)), (('en', WordPart.FULL_WORD), ('vi', WordPart.FULL_WORD), ('as', WordPart.POS))]

API Reference

Disvortilo

Parser class for splitting Esperanto words into morphemes.

Disvortilo.parse(word: str) -> list[tuple[str, ...]]

Returns all valid analyses of word. Each analysis is a tuple of morpheme strings in order.

Example return value:

[('esper', 'ant', 'ist', 'in', 'o'), ('esperant', 'ist', 'in', 'o')]

Disvortilo.parse_detailed(word: str, n: int = None) -> list[tuple[tuple[str, WordPart], ...]]

Like parse, but each morpheme is returned together with its detected category (WordPart). Each analysis is a tuple of (morpheme, WordPart) pairs. The n options limits the returned options and sorts them based on a heuristic of the most likely option.

Example return value:

[(('pli', WordPart.FULL_WORD), ('bon', WordPart.ROOT), ('ig', WordPart.SUFFIX), ('ojn', WordPart.POS))]

rank(options, n: int | None = None) -> list[tuple[tuple[str, WordPart], ...]]

Sorts options based on a heuristic of the most likely option.

Note, that this will only work with the returned values from parse_detailed.

Example return value:

[(('neĝ', WordPart.ROOT), ('is', WordPart.POS)), (('ne', WordPart.FULL_WORD), ('ĝis', WordPart.FULL_WORD))]

WordPart

Enum values used by parse_detailed:

• PREFIX
• ROOT
• SUFFIX
• FULL_WORD
• POS
• NUMBER
• NAME
• CORRELATIVE_START
• CORRELATIVE_END

split_sentence(sentence: str) -> list[str]

Splits a sentence into Esperanto word-like tokens. Supports Esperanto diacritics, optional trailing apostrophes, and forms like 3 and 3an.

Example:

from disvortilo import split_sentence

split_sentence("Mi vidas 3an domon.")
# > ['Mi', 'vidas', '3an', 'domon']