Function lex

lex(input: string): Token[]
Tokenizes a dice notation string. The first stage of the pipeline — parse calls it for you; reach for lex directly only to build a syntax highlighter or an editor integration.

Notation is case-insensitive and whitespace-tolerant: 2D20 + 5 and 2d20+5 produce the same tokens, and identifier tokens carry a lowercased value. The one exception is @name, whose case is preserved.
Parameters
- input: string
  The dice notation to tokenize
Returns Token[]
Every token in source order, always ending with one TokenType.EOF token
Throws
If an invalid character or unknown identifier is found

Throws
INVALID_NOTATION_TYPE when input is not a string — raised before scanning, so it carries no position
Example
```
import { lex, TokenType } from 'roll-parser';

const tokens = lex('2d20+5');
tokens.length; // 6 — NUMBER DICE NUMBER PLUS NUMBER EOF
tokens[0]; // { type: TokenType.NUMBER, value: '2', position: 0, end: 1 }
tokens[1].type === TokenType.DICE; // true
tokens.at(-1)?.type === TokenType.EOF; // true
```
- Defined in lexer/lexer.ts:505