Module leb128 

Unsigned LEB128 (varint) — 7 payload bits per byte, low group first, high bit set while more bytes follow. The wire format of protobuf varint, WebAssembly, DWARF.

Generic over the field’s integer width via [Varint]: the declared field type pins the width, so the same parse/write pair serves u16 and u64 fields alike. Decoding is bounded and overflow-checked — a value too large for the field or a continuation run longer than the width allows is a clean BitError, never a panic or a silent wrap. Non-minimal encodings (e.g. 0x80 0x00 for zero) are accepted: permissive parse, canonical (minimal) write.

use bnb::bin;

#[bin(big)]
#[derive(Debug, PartialEq)]
struct Record {
    #[br(parse_with = bnb::codecs::leb128::parse)]
    #[bw(write_with = bnb::codecs::leb128::write)]
    length: u32, // width inferred from the field type
    #[br(parse_with = bnb::codecs::leb128::parse)]
    #[bw(write_with = bnb::codecs::leb128::write)]
    timestamp: u64,
}

let r = Record { length: 300, timestamp: 1 };
let bytes = r.to_bytes().unwrap();
assert_eq!(bytes, [0xAC, 0x02, 0x01]); // 300 = 0b10_0101100 → AC 02; 1 → 01
assert_eq!(Record::decode_exact(&bytes).unwrap(), r);

Traits

Varint

The integer widths readable/writable as LEB128 — u8, u16, u32, u64, u128. Sealed: LEB128 is byte-granular, so the primitive widths are the whole set (a uN field wants the next wider primitive).

Functions

parse

Reads one LEB128 value; the target width comes from the field’s declared type.

write

Writes one LEB128 value in canonical (minimal) form.