Typed Bytesize

The library provides two types to represent bytesize:

• ByteSizeSi for decimal prefix unit (SI)
• ByteSizeIec for binary prefix unit (IEC 80000-13)

Functions

• Bytesize types can parse each other's units (e.g. ByteSizeIec can parse SI values like 114514GB);
• Bytesize values will only be formatted as the unit has their owned prefix;
• Bytesize types can be converted to each other;
• Supporting addition, subtraction, scalar multiplication arithmetic operations;
• Optional serde support.

Example

use typed_bytesize::{ByteSizeIec, ByteSizeSi};

// SI
assert_eq!(ByteSizeSi::b(114u64), "114".parse().unwrap());
assert_eq!(ByteSizeSi::mb(114), "114MB".parse().unwrap());
print!("{}", ByteSizeSi::kb(310)); // 310.0kB

// IEC
assert_eq!(ByteSizeIec::b(514u64), "514".parse().unwrap());
assert_eq!(ByteSizeIec::mib(514), "514MiB".parse().unwrap());
print!("{}", ByteSizeIec::gib(93696)); // 91.5GiB

For more detailed examples, refer to tests.

BNF

Parsing follows the rule:

expr    ::= num | term
term    ::= decimal " "* unit
decimal ::= num | float
float   ::= num "." num
num     ::= [0-9]+

Features

• serde: enable serde on ByteSizeSi and ByteSizeIec.
• u128: use u128 instead of u64 as inner numeric type to support larger units. (TODO)