basic: bytes, string and runes.md

go basic: bytes, string and runes

You may find bytes, string and runes about characters. So whatare the differences between them and which one to use?

bytes

Byte is actually alias for uint8, defined as type byte = uint8. And bytes or byte slice are just slice of byte like s1 := []byte("hello world"). Each element of the slice is a single byte (8 bits). We often use it with binary data, raw data or ascii characters.

string

The string is like a read only slice of bytes. This is different from string of python. Following are attentions you need to take care about string:

• string is just RO slice, then len(str) is actually length of the slice

    s1 := "你好，世界!"
    fmt.Printf("length: %d\n", len(s1))        // get 16 but not 6

• index a string is same like slice, s[i] is the i-th byte but not i-th char.

• string is just read-only slice, it's no need to be VALID string, you can store any bytes

• range loop string will decode string as utf8 and return unicode code points (rune)

    s1 := "你好!!"
    for i, ch := range s1 {
        fmt.Printf("index: %d, char: %s, code: %U\n", i, string(ch), ch)
    }
// output
    index: 0, char: 你, code: U+4F60
    index: 3, char: 好, code: U+597D
    index: 6, char: !, code: U+0021
    index: 7, char: !, code: U+0021

If you don't want this, you can loop string use `for i := 0; i < len(s1); i++ {}`
 or `for i, b := range []byte(s1) {}`.

runes

Rune is alias of int32, defined as type rune = int32. Often used to store unicode code point. So why rune?

Rune uses fixed width for each character, no matter it's ascii like a, or chinese like 你. It will use more memory than string (utf8 encoded) but the benefit is that it is O(1) to find character via index.

With utf8 encoded string, it needs to decode from beginning to find the Nth character ( not Nth byte) since each character may be 1 byte to 4 bytes. You cannot jump to the Nth character with decoding. However, with runes, just simple jump to the 4 * (N-1) bytes.

conversion

Conversion between bytes, string and runes.

    s1 := "你好!!"
    s2 := []rune(s1)     // convert string to []rune
    s3 := string(s2)     // convert rune to string