vyasa-lipi CLI Reference

vyasa-lipi is the command-line interface for the universal Sanskrit transliteration engine. It provides high-throughput, lossless transliteration of Vedic and Classical Sanskrit across Indic, Roman, and computational ASCII scripts.

Installation

From the vyutils repository root:

cargo build --release -p vyasa-lipi
cargo install --path crates/vyasa-lipi

Binary location: ~/.cargo/bin/vyasa-lipi

Synopsis

vyasa-lipi [OPTIONS] [TEXT]
cat <file> | vyasa-lipi [OPTIONS]

If [TEXT] is provided as an argument, it is transliterated directly. If [TEXT] is omitted, vyasa-lipi reads from stdin until EOF.

Options & Flags

Flag	Long Option	Values	Default	Description
`-f`	`--from`	`<SCRIPT>`	Auto-detect	Source script format. If omitted, script is inferred from Unicode codepoints.
`-t`	`--to`	`<SCRIPT>`	`devanagari`	Target script format.
`-a`	`--accents`	`preserve` \| `strip` \| `warn`	`preserve`	Handling of Vedic pitch accents (udātta, anudātta, svarita).
`-h`	`--help`	—	—	Display help information and exit.
`-V`	`--version`	—	—	Display version information and exit.

Supported Script Identifiers

The --from and --to arguments accept case-insensitive script names and standardized short aliases:

Script Name	Recognized Identifiers / Aliases	Script Category	Primary Domain
Devanagari	`devanagari`, `deva`	Indic Abugida	Standard Sanskrit, Northern India, Vedic editions
Telugu	`telugu`, `telu`	Indic Abugida	Andhra Pradesh, Telangana, Krishna Yajurveda tradition
Kannada	`kannada`, `knda`	Indic Abugida	Karnataka, Rigveda & Sama Veda traditions
Grantha	`grantha`, `gran`	Indic Abugida	Traditional Sanskrit script of Tamil Nadu
Malayalam	`malayalam`, `mlym`	Indic Abugida	Kerala, Rigvedic Nambudiri tradition
Bengali	`bengali`, `beng`	Indic Abugida	Bengal, Assam, Eastern Indian traditions
Strict ISO 15919	`iso15919`, `iso`	Roman Diacritic	International Library standard (under-rings for liquids, macrons)
Standard IAST	`iast`	Roman Diacritic	Western academic standard (under-dots)
SLP1	`slp1`	Computational ASCII	Cologne Digital Sanskrit Dictionaries (1:1 ASCII mapping)
Harvard-Kyoto	`harvard-kyoto`, `hk`	Computational ASCII	Case-sensitive ASCII transliteration
WX	`wx`	Computational ASCII	Indian computational linguistics standard (IIT-K / UoH)

Accent Handling Modes

Vedic texts use combining pitch accents. The --accents flag controls how these are emitted:

preserve (Default): Retains all pitch accents. In Indic abugidas, combining Vedic marks \u0951 (svarita), \u0952 (anudātta), and \u1CDA (dīrghasvarita) are correctly positioned over the target script’s aksharas.
strip: Completely strips all Vedic pitch accent marks, leaving clean classical Sanskrit text suitable for search indexing or general readership.
warn: Preserves accents where supported, but outputs a warning on stderr when targeting formats with limited native diacritic representation.

Exit Codes

Code	Meaning
`0`	Success: Transliteration completed without errors.
`1`	Error: Unknown script identifier, invalid CLI argument, or I/O failure.

Examples

1. Direct Command Line Transliteration

# Transliterate Rigveda 1.1.1 to Telugu
vyasa-lipi --from devanagari --to telugu "अ॒ग्निमी॑ळे पु॒रोहि॑तम्"
# అ॒గ్నిమీ॑ళె పు॒రోహి॑తమ్

# Auto-detect script and convert to IAST
vyasa-lipi --to iast "ధర్మక్షేత్రే కురుక్షేత్రే"
# dharmakṣētrē kurukṣētrē

2. Unix Pipeline Processing

# Transliterate whole files via stdin/stdout
cat input_devanagari.txt | vyasa-lipi --to kannada > output_kannada.txt

# Strip accents during stream
cat vedic_corpus.txt | vyasa-lipi --to devanagari --accents strip > corpus_clean.txt

3. Computational ASCII Transformations

# Round-trip through Harvard-Kyoto
vyasa-lipi --from devanagari --to hk "धर्मक्षेत्रे" | vyasa-lipi --from hk --to devanagari
# धर्मक्षेत्रे