Remove ByteOrder generic methods from Buf and BufMut (#187)

* make Buf and BufMut usable as trait objects

- All the `get_*` and `put_*` methods that take `T: ByteOrder` have
  a `where Self: Sized` bound added, so that they are only usable from
  sized types. It was impossible to make `Buf` or `BufMut` into trait
  objects before, so this change doesn't break anyone.
- Add `get_n_be`/`get_n_le`/`put_n_be`/`put_n_le` methods that can be
  used on trait objects.
- Deprecate the export of `ByteOrder` and methods generic on it.

* remove deprecated ByteOrder methods

Removes the `_be` suffix from all methods, implying that the default
people should use is network endian.
This commit is contained in:
Sean McArthur
2018-03-12 09:27:09 -07:00
committed by Carl Lerche
parent ff7c0a1d90
commit 025d53342e
6 changed files with 596 additions and 116 deletions
+3 -3
View File
@@ -33,15 +33,15 @@ fn test_get_u8() {
#[test]
fn test_get_u16() {
let buf = b"\x21\x54zomg";
assert_eq!(0x2154, Cursor::new(buf).get_u16::<byteorder::BigEndian>());
assert_eq!(0x5421, Cursor::new(buf).get_u16::<byteorder::LittleEndian>());
assert_eq!(0x2154, Cursor::new(buf).get_u16());
assert_eq!(0x5421, Cursor::new(buf).get_u16_le());
}
#[test]
#[should_panic]
fn test_get_u16_buffer_underflow() {
let mut buf = Cursor::new(b"\x21");
buf.get_u16::<byteorder::BigEndian>();
buf.get_u16();
}
#[test]