SiMin - NFHN Reader

A lightweight format for writing numbers using scientific notation.

I wrote this to make it easy to allow applications to accept values with units gracefully. Although many languages implement scientific notation, I never found it as ergonomic as just using SI prefixes. Would you rather a user input a unitless 1.024e+6 or 1_000KiB?

How does it work?

A SiMin number has three parts: a value, prefix, and unit. Let's look at an example:

tokenize('-120mT') === [ '-120', 'm', 'T' ]

'120' is the value. It can optionally start with a + or -, and consists of digits, decimals, and underscores. All of the following are valid values:
- 9_800
- +1.20030
- -2300.123_456
'm' is the prefix. It is the longest valid SI prefix that the parser can find directly after the value.
'T' is the unit. It's just the rest of the string after the prefix.

More examples

// parsing numbers
console.log(parseUnit("10MB"));     // [ 10000000, 'B' ]
console.log(parseUnit("100mV"));    // [ 0.1, 'V' ]
console.log(parseUnit("-1.2uT"));   // [ -0.0000012, 'T' ]
console.log(parseUnit("123.45A"));  // [ 123.45, 'A' ]

// you can use underscores to make things more readable
console.log(parseUnit("100_000_000"));  // [ 100000000, '' ]

// automatic BigInt for really large prefixes.
console.log(parseUnit("10_000EiB"));    // [ 115292150460684697600000n, 'B' ]

// use parse to ignore the unit
console.log(parse("920d"));         // 92

// tokenizing numbers (if you want to parse it yourself)
console.log(tokenize("50Mibps"));   // [ '50', 'Mi', 'bps' ]

// you can also get the info of any SI prefix
console.log(prefixData("Mi"));      // [ 2, 20, [Function: parseFloat] ]

Prefixes

Important

SiMin numbers that use the prefixes marked with 🔢 will parse to BigNums

Prefix	Base	Exponent
E 🔢	10	18
P 🔢	10	15
T	10	12
G	10	9
M	10	6
k	10	3
h	10	2
da	10	1
d	10	-1
c	10	-2
m	10	-3
u, μ	10	-6
n	10	-9
p	10	-12
f	10	-15
a	10	-18

Prefix	Base	Exponent
Ei 🔢	2	60
Pi 🔢	2	50
Ti	2	40
Gi	2	30
Mi	2	20
Ki	2	10