Method #getCodePoint

xml/_tokenizer.ts:439–455

* Get the full Unicode code point at the current buffer position. * Handles surrogate pairs for astral plane characters (U+10000+). * Returns [codePoint, charCount] where charCount is 1 or 2.

(
    buffer: string,
    index: number,
  )

   * Returns [codePoint, charCount] where charCount is 1 or 2.
   */
  #getCodePoint(
    buffer: string,
    index: number,
  ): [codePoint: number, charCount: number] {
    const code = buffer.charCodeAt(index);
    // Check for high surrogate (0xD800-0xDBFF)
    if (code >= 0xD800 && code <= 0xDBFF && index + 1 < buffer.length) {
      const low = buffer.charCodeAt(index + 1);
      // Check for valid low surrogate (0xDC00-0xDFFF)
      if (low >= 0xDC00 && low <= 0xDFFF) {
        // Decode surrogate pair: ((high - 0xD800) * 0x400) + (low - 0xDC00) + 0x10000
        const codePoint = ((code - 0xD800) << 10) + (low - 0xDC00) + 0x10000;
        return [codePoint, 2];
      }
    }
    return [code, 1];
  }

  /**
   * Check if the current buffer position has a valid NameStartChar.