API Reference

This document provides technical details about the classes and methods available in the Nepali Unicoder package.

Converter Class

The Converter class is a simplified wrapper around the Engine for easy usage.

__init__(self, mode: str = "roman")

Initializes a new Converter. - mode: Either "roman" (default) or "preeti".

convert(self, text: str) -> str

Translates the input text to Unicode Devanagari. - text: The input string (Romanized Nepali or Preeti characters).

---

Engine Class

The core conversion logic is implemented in the Engine class.

__init__(self, trie: Optional[Trie] = None, tokenizer: Optional[Tokenizer] = None, mode: str = "roman")

Initializes the conversion engine. - Loads the appropriate character mappings into a Trie based on the mode. - Loads post_rules for contextual transformations in Preeti mode.

---

Tokenizer Class

The Tokenizer splits the input text into meaningful chunks (Tokens).

tokenize(self, text: str, use_blocks: bool = True) -> List[Token]

Splits text into tokens based on types: ROMAN, BLOCK, LITERAL, and NUMBER. - use_blocks: If True (default), recognizes {...} as "as-is" blocks. In Preeti mode, this is typically set to False by the Engine.