Tokenization
Text Tokenization
Splitting text into meaningful units (tokens) such as words, sentences, or subword pieces for processing.
Détail technique
Tokenization operates on sequences of Unicode code points, where each character's properties (category, script, case, directionality) are defined by the Unicode standard. Text processing in the browser uses the TextEncoder/TextDecoder APIs for encoding conversion and Intl.Segmenter for locale-aware word and sentence boundary detection. Understanding the distinction between bytes, code units, code points, and grapheme clusters is essential for correct text manipulation.
Exemple
```javascript // Tokenization: text processing example const input = 'Sample text for processing'; const result = input .trim() .split(/\s+/) .filter(Boolean); console.log(result); // ['Sample', 'text', 'for', 'processing'] ```