lipimala for Python

IAST-style transliteration for Devanagari and Gujarati, with canonical reverse conversion, direct script conversion, and Plain-English or Hunterian output.

Install

pip install lipimala

Requires Python 3.12 or newer. The package has no runtime dependencies.

Use

from lipimala import (
    to_devanagari_from_iast,
    to_canonical_iast_from_gujarati,
    to_plain_english_from_devanagari,
    to_plain_english_from_gujarati,
    to_canonical_gujarati_from_devanagari,
)

print(to_devanagari_from_iast('Kṛṣṇa'))  # कृष्ण
print(to_canonical_iast_from_gujarati('કૃષ્ણ'))  # kṛṣṇa
print(to_plain_english_from_devanagari('कानपुर'))  # kaanpur
print(to_plain_english_from_gujarati('આજ મારે'))  # aaj maare
print(to_canonical_gujarati_from_devanagari('कृष्ण'))  # કૃષ્ણ

Use to_hunterian_from_devanagari() and to_hunterian_from_gujarati() for Hunterian output. List helpers such as to_devanagari_from_iast_list() convert sequences of strings.

Profiles and options

Forward conversion accepts strictIast, iso15919Core, or extendedIndic profiles. The default is extendedIndic. Options control normalization, unknown input, digits, punctuation, Vedic accents, whitespace, and exact-source metadata.

Direct Brahmic transcription accepts BrahmicPlainEnglishOptions for language selection, schwa deletion, vowel length, terminal anusvara, Gujarati ʿayn, and other pronunciation choices.

Preserve the original input

The rendered script may not encode the original Latin spelling, capitalization, or normalization. A result envelope keeps the source beside the rendered value:

from lipimala import to_devanagari

result = to_devanagari('Kṛṣṇa')
print(result.rendered)  # कृष्ण
print(result.restore_original())  # Kṛṣṇa

For a standalone string, set embed_exact_source_metadata=True in the conversion options and use an exact reverse function such as to_exact_iast_from_devanagari(). Unicode-tag metadata can be removed by systems that sanitize invisible characters.

License

MIT. See LICENSE.