utf-8 - 请定义术语 "Multi-byte safe"

Question

很难说出这里问的是什么。这个问题是含糊的、模糊的、不完整的、过于宽泛的或修辞性的，无法以目前的形式得到合理的回答。如需帮助澄清此问题以便重新打开它， visit the help center 。

已关闭 12 年前。

我现在对 UTF-8 有点迷茫。
我正在寻找术语“多字节安全”的精确定义。

Answer 1

当您处理 unicode 字符时，假设所有字符仅采用单个字节或字符 (java) 是不安全的。因此，在读取或解析字符串时，需要考虑到这一点。

这是一个excellent article这解释了在 Java 中处理 Unicode 时的复杂性。

1. Stored characters can take up an inconsistent number of bytes. A UTF-8 encoded character might take between one (LATIN_CAPITAL_LETTER_A) and four (MATHEMATICAL_FRAKTUR_CAPITAL_G) bytes. Variable width encoding has implications for reading into and decoding from byte arrays.

2. Not all code points can be stored in a char. The MATHEMATICAL_FRAKTUR_CAPITAL_G example lies in the supplementary range of characters and cannot be stored in 16 bits. It must be represented by two sequential char values, neither of which is meaningful by itself. The Character class provides methods for working with 32-bit code points.

    // Unicode code point to char array
     char[] math_fraktur_cap_g = Character.toChars(0x1D50A);