A Unicode code point identifies an abstract character or symbol with one numeric value. In Go, a rune is a Unicode code point, while UTF-8 encodes that value into one to four bytes. Confusing these levels causes incorrect assumptions about string length and indexing. section 5.1
A character can occupy several bytes
The character 汉 has code point U+6C49 and occupies three UTF-8 bytes: 0xE6, 0xB1, and 0x89. Consequently, len("汉") returns 3, because Go measures the string's bytes rather than its number of code points. section 5.1
Choose the operation according to whether the task concerns bytes or code points:
- Use range to iterate over runes; each index identifies the starting byte position of the rune's UTF-8 sequence. section 5.7
- Convert a string to []rune when accessing a specific rune position, such as the third code point. section 5.7
- Do not assume every string is UTF-8: string literals are UTF-8 encoded, but strings read from external sources are arbitrary bytes. section 5.1
s := "汉" fmt.Println(len(s)) // 3
Inspect the contrasting byte length of a string containing 汉. section 5.1
