A simple script to transliterate Devanagari text with Latin alphabet.

readme.md

A simple script to transliterate words in Devanagari with Latin alphabet.

I wrote this as a sort of proof of concept and might fail with some cases. Consonants with a nuqta are not supported but should be fairly easy to extend this code to do so. (I am lazy, idk)

Why?

I was really bored.

Why does it add an "a" at the end of my name?

You might want to read about schwa deletion in Indo-Aryan Languages

Will you ever finish this?

Probably not.

Can I rely on this code to handle my-

NO. Please don't.

translitrate.py

VYANJANA_MAP = {
    "क": "k", "ख": "kh",
    "ग": "g", "घ": "gh",
    "ङ": "n",

    "च": "ch", "छ": "chh",
    "ज": "j", "झ": "jh",
    "ञ": "n",

    "ट": "t", "ठ": "th",
    "ड": "d", "ढ": "dh",
    "ण": "n",

    "त": "t", "थ": "th",
    "द": "d", "ध": "dh",
    "न": "n",

    "प": "p", "फ": "ph",
    "ब": "b", "भ": "bh",
    "म": "m",

    "य": "y",
    "र": "r",
    "ल": "l",
    "व": "v",
    "श": "sh",
    "ष": "sh",
    "स": "s",
    "ह": "h",
    "ळ": "l",
}

OSHTHYA_VANJAN = {"प", "फ", "ब", "भ", "म"}

SVAR_MAP = {
    "अ": "a", "आ": "aa",
    "इ": "i", "ई": "ee",
    "उ": "u", "ऊ": "oo",
    "ए": "e", "ऐ": "ai",
    "ओ": "o", "औ": "au",
    "ऋ": "ri"
}

MATRA_MAP = {
    "ा": "aa",
    "ि": "i", "ी": "ee",
    "ु": "u", "ू": "oo",
    "े": "e", "ै": "ai",
    "ो": "o", "ौ": "au",
    "ः": "ah", "ृ": "ri"
}

ANUSVARA = "ं"
HALANTA = "्"
VISARGA = "ः"

def _get_nasalizer_for_vyanjana(vyanyana):
    # If next vyanjana is a oshthya vyanjana, use "m" else "n"
    if vyanyana in OSHTHYA_VANJAN:
        return "m"
    else:
        return "n"

def _get_next(chars: list[str]) -> str:
    """Consume next syllable and return it's latin translitration"""
    first = chars.pop(0)
    if first in VYANJANA_MAP:

        if not chars:
            # We exhausted the characters, return the current vyanjana with the inherent "a"
            return VYANJANA_MAP[first] + "a"

        # Peek next
        next = chars[0]

        if next == HALANTA:
            # Discard halant
            chars.pop(0)

            # If this is begining of a consonant cluster
            if chars:
                return VYANJANA_MAP[first] + _get_next(chars)

            # The word ends with an halanta, simply return the consonant
            else:
                return VYANJANA_MAP[first]
        
        # If this is a consonant + vowel
        elif next in MATRA_MAP:
            return VYANJANA_MAP[first] + MATRA_MAP[chars.pop(0)]
        
        # If it's the anusvar
        elif next == ANUSVARA:
            # Discard anusavar
            chars.pop(0)

            if chars and chars[0] in VYANJANA_MAP:
                return VYANJANA_MAP[first] + "a" + _get_nasalizer_for_vyanjana(chars[0])
            else:
                raise ValueError("Excpected a vyanjana after anusavara")
        else:
            return VYANJANA_MAP[first] + "a"

    elif first in SVAR_MAP:
        if not chars:
            # Simply return the swar mapping
            return SVAR_MAP[first]
        
        next = chars[0]

        if next == ANUSVARA:
            # Discard anusvara
            chars.pop(0)
            if chars and chars[0] in VYANJANA_MAP:
                return SVAR_MAP[first] + _get_nasalizer_for_vyanjana(chars[0])
            else:
                raise ValueError("Expected a vyanjana after anusvara")
        elif next == VISARGA:
            chars.pop(0)
            return SVAR_MAP[first] + "h"
        else:
            return SVAR_MAP[first]
    else:
        raise ValueError(f"Invalid devanagari text, expected start of a syllable (a vyanjana or swara, got {first})")

def transliterate(devanagari_text: str) -> str:
    """Transforms devanagari text to it's transliterated equivalent in Latin alphabet"""

    result = []
    chars = list(devanagari_text)
    while chars:
        result.append(_get_next(chars))
    
    return "".join(result)


def test_transliterate():
    assert transliterate("अभिजीत").title() == "Abhijeeta"
    assert transliterate("संस्कृत").title() == "Sanskrita"
    assert transliterate("राणा").title() == "Raanaa"
    assert transliterate("प्रकृति").title() == "Prakriti"
    assert transliterate("फुल").title() == "Phula"

    assert transliterate("अंबर").title() == "Ambara"
    assert transliterate("अंक").title() == "Anka"
    assert transliterate("आंसू").title() == "Aansoo"

    assert transliterate("हिमः").title() == "Himah"

    assert transliterate("इंजिन्").title() == "Injin"
    assert transliterate("मोटर्").title() == "Motar"
    assert transliterate("लक्षमन").title() == "Lakshamana"

if __name__ == "__main__":
    test_transliterate()