| r[type.char] |
| # Character type |
| |
| r[type.char.intro] |
| The `char` type represents a single [Unicode scalar value] (i.e., a code point that is not a surrogate). |
| |
| > [!EXAMPLE] |
| > ```rust |
| > let c: char = 'a'; |
| > let emoji: char = '😀'; |
| > let unicode: char = '\u{1F600}'; |
| > ``` |
| |
| > [!NOTE] |
| > See [the standard library docs][`char`] for information on the impls of the `char` type. |
| |
| r[type.char.value] |
| A value of type `char` is represented as a 32-bit unsigned word in the 0x0000 to 0xD7FF or 0xE000 to 0x10FFFF range. It is immediate [undefined behavior] to create a `char` that falls outside this range. |
| |
| r[type.char.layout] |
| `char` is guaranteed to have the same size and alignment as `u32` on all platforms. |
| |
| r[type.char.validity] |
| Every byte of a `char` is guaranteed to be initialized. In other words, `transmute::<char, [u8; size_of::<char>()]>(...)` is always sound -- but since some bit patterns are invalid `char`s, the inverse is not always sound. |
| |
| [Unicode scalar value]: http://www.unicode.org/glossary/#unicode_scalar_value |
| [undefined behavior]: ../behavior-considered-undefined.md |
