How to Transliterate Sanskrit

This guide demonstrates how to use vyasa-lipi—both as a standalone CLI tool and as a Rust library—to perform lossless transliteration of Vedic and Classical Sanskrit texts across 11 supported formats.

---

1. Installing the CLI

From the root of the vyutils workspace:

# Build release binary
cargo build --release -p vyasa-lipi

# Install directly to your PATH (~/.cargo/bin)
cargo install --path crates/vyasa-lipi

Verify the installation:

vyasa-lipi --version

---

2. Basic Transliteration

Indic Abugida Conversion

Transliterate a verse from the Bhagavad Gita (1.1) from Devanagari to Telugu:

vyasa-lipi --from devanagari --to telugu "धर्मक्षेत्रे कुरुक्षेत्रे समवेता युयुत्सवः"

Output:

ధర్మక్షేత్రే కురుక్షేత్రే సమవేతా యుయుత్సవః

Transliterate to Kannada:

vyasa-lipi --from devanagari --to kannada "धर्मक्षेत्रे कुरुक्षेत्रे समवेता युयुत्सवः"

Output:

ಧರ್ಮಕ್ಷೇತ್ರೇ ಕುರುಕ್ಷೇತ್ರೇ ಸಮವೇತಾ ಯುಯುತ್ಸವಃ

---

3. Vedic Pitch Accent Preservation

By default, vyasa-lipi preserves all Vedic pitch accents (Udātta, Anudātta, Svarita, DīrghaSvarita).

Transliterate Ṛgveda 1.1.1 to Telugu:

vyasa-lipi --from devanagari --to telugu "अ॒ग्निमी॑ळे पु॒रोहि॑तम्"

Output:

అ॒గ్నిమీ॑ళే పు॒రోహి॑తమ్

Notice:

1. The Anudātta under-bar (\u0952) on a and ro is preserved.
2. The Svarita vertical stroke (\u0951) on mī is preserved.
3. The Vedic flap ळ (\u0933) is mapped to Telugu ళ (\u0C33).

Stripping Accents for Search Indexing

If you are preparing text for non-Vedic readers or generating search indexes where accents interfere with matching, use --accents strip:

vyasa-lipi --from devanagari --to devanagari --accents strip "अ॒ग्निमी॑ळे पु॒रोहि॑तम्"

Output:

अग्निमीळे पुरोहितम्

---

4. Transliterating to Grantha & Malayalam

Grantha is the canonical Tamil-region Sanskrit script:

vyasa-lipi --from devanagari --to grantha "धर्मक्षेत्रे कुरुक्षेत्रे"

Output:

𑌧𑌰𑍍𑌮𑌕𑍍𑌷𑍇𑌤𑍍𑌰𑍇 𑌕𑍁𑌰𑍁𑌕𑍍𑌷𑍇𑌤𑍍𑌰𑍇

Malayalam:

vyasa-lipi --from devanagari --to malayalam "धर्मक्षेत्रे कुरुक्षेत्रे"

Output:

ധര്മക്ഷേത്രേ കുരുക്ഷേത്രേ

---

5. Automatic Script Detection

If you omit the --from flag, vyasa-lipi automatically inspects the Unicode codepoints of the input:

# Detected as Telugu -> Transliterated to Devanagari
vyasa-lipi --to devanagari "ధర్మక్షేత్రే కురుక్షేత్రే"

Output:

धर्मक्षेत्रे कुरुक्षेत्रे

# Detected as Kannada -> Transliterated to IAST
vyasa-lipi --to iast "ಧರ್ಮಕ್ಷೇತ್ರೇ ಕುರುಕ್ಷೇತ್ರೇ"

Output:

dharmakṣētrē kurukṣētrē

---

6. Strict ISO 15919 vs Academic IAST

Depending on whether you need strict international library standards or standard Indological transcription:

# ISO 15919 (under-ring for r̥, macron for ē)
vyasa-lipi --to iso15919 "कृष्णः"
# kr̥ṣṇaḥ

# Standard IAST (under-dot for ṛ)
vyasa-lipi --to iast "कृष्णः"
# kṛṣṇaḥ

---

7. Computational ASCII Formats (SLP1, HK, WX)

When feeding Sanskrit into computational pipelines, parsers, or search engines:

# Convert Devanagari to SLP1 (Cologne Sanskrit dictionary standard)
vyasa-lipi --from devanagari --to slp1 "धर्मक्षेत्रे कुरुक्षेत्रे"
# Darmakzetre kurukzetre

# Convert back from SLP1 to Devanagari
vyasa-lipi --from slp1 --to devanagari "Darmakzetre kurukzetre"
# धर्मक्षेत्रे कुरुक्षेत्रे

---

8. Piped Input & Batch Processing

vyasa-lipi accepts input from standard input (stdin), making it ideal for shell scripts and corpus pipelines:

# Process a full chapter from file
cat gita_chapter1_deva.txt | vyasa-lipi --to telugu > gita_chapter1_telu.txt

# Inspect difference after round-trip
diff <(cat text.txt) <(cat text.txt | vyasa-lipi --to telugu | vyasa-lipi --to devanagari)
# Zero output confirms byte-identical round-trip fidelity!

---

9. Rust API Integration

Add vyasa-lipi to your project’s Cargo.toml:

[dependencies]
vyasa-lipi = { path = "../crates/vyasa-lipi" }

Example Rust Program

use vyasa_lipi::{detect_script, transliterate, Script};
use vyasa_lipi::model::AccentMode;

fn main() {
    let input = "अ॒ग्निमी॑ळे पु॒रोहि॑तम्";

    // Auto-detect source script
    let src = detect_script(input).unwrap_or(Script::Devanagari);

    // Transliterate to Telugu with accents preserved
    let telugu = transliterate(input, src, Script::Telugu, AccentMode::Preserve);
    println!("Telugu: {}", telugu);

    // Transliterate to IAST with accents stripped
    let iast = transliterate(input, src, Script::Iast, AccentMode::Strip);
    println!("IAST (stripped): {}", iast);

    // Verify round-trip back to Devanagari
    let roundtrip = transliterate(&telugu, Script::Telugu, Script::Devanagari, AccentMode::Preserve);
    assert_eq!(roundtrip, input);
}

---

Next Steps

• Consult the complete flag and script reference: vyasa-lipi Reference
• Learn about the underlying phonetic models: Sanskrit Phonetics