biguint-format

Big UInteger Formatter

An arbitrary length unsigned integer formatter library for Node.js.

JavaScript uses IEEE 754 double-precision floats to represents numbers. That works perfectly fine for small numbers, however, it is an issue for big integers. This means they lose integer precision for values beyond +/- 2 pow 53

Problem

Presentation of small integer in decimal format works fine (e.g. 0x1FF). However, we can see an issue when we try to convert big integers like 0x1234567890abcdeffedcba908765421 to string decimal.

(0x1FF).toString(10) // returns '511' - correct
(0x1234567890abcdeffedcba908765421).toString(10) 

// output is '1.5123660750094533e+36' - incorrect - lose integer precision

Solution

Node.js biguint-format module has been built in order to help display very large (arbitrary lengh) unsigned integers without any integer precision lose.

Example:

var format  = require('biguint-format');

// 0x1234567890abcdeffedcba908765421 split into bytes
format([
  0x1, 0x23, 0x45, 0x67, 0x89, 0x0a, 0xbc, 0xde, 
  0xff, 0xed, 0xcb, 0xa9, 0x08, 0x76, 0x54, 0x21], 'dec')

// output value is '1512366075009453296626403467035300897' - no integer precision lose

Instalation

$ npm install biguint-format

API

The biguint-format module is a function (fn(number, format [, options])) which performs number conversion to the required string format.

The number argument represents an arbitrary lenght unsigned integer number to be converted to string. It can be provided in one of the following formats:

• Node.js Buffer e.g. new Buffer([0x1, 0xFF])
• An array of bytes (values from 0x00 to 0xFF) e.g. [0x1, 0xFF].
• A string with a number in a hexadecimal format e.g. 0x1FF0A or 1FF0A

The format argument represents output string format and it can be one of the following options:

• dec - convertion to decimal format e.g. 123324884
• bin - conversion to binary format e.g. 1100101010
• hex - conversion to hexadecimal format e.g. 0xADFFAA11
• oct - conversion to octet format e.g. 07771

If format argument is missing, dec format is used as a default option. Note that you will have to specify format if you would like to use options argument.

The options argument (optional) is an object which provides some additional conversion details:

• format - specifies format of the input number. It can be either BE for Big Endian or LE for Little Endian. BE is a default value. Check wikipedia for more details.
• prefix - output string prefix. Note that this option is not supported by dec conversion.
• groupsize - splits output string into groups of groupsize lenght characters.
• delimiter - specifes delimiter string to be inserted in between character groups. Default value is space. It is quite handy option when dealing with large numbers.
• trim - (works only with bin formatting) specifies if the leading 0's should be trimmed.
• padstr - string used for right-padding of the fomratted string if its length (inlucing prefix and grouping) is less than value of size parameter.
• size - determines formatted string size. That opiton has effect only with padstr option. Note that the formatted string is not trimmed if its length is longer than value of size parameter.

Examples

var format = require('biguint-format');

var buffer1 = new Buffer([0x63, 0xA7, 0x27]);
var buffer2 = new Buffer([0x27, 0xA7, 0x63]);

format(buffer1, 'dec', {format:'LE'})   // returns '2598755'
format(buffer2, 'dec', {format:'BE'})   // returns '2598755'
format(buffer2, 'dec')                  // returns '2598755'

format(buffer1, 'hex', {format:'LE'})   // returns '27a763'
format(buffer2, 'hex', {format:'BE'})   // returns '27a763'
format(buffer2, 'hex', {prefix:'0x'})   // returns '0x27a763'

format(buffer2, 'bin')                  // 001001111010011101100011
format(buffer2, 'bin', {groupsize:8})   // 00100111 10100111 01100011
format(buffer2, 'oct')                  // 11723543
format(buffer2, 'oct', {prefix:'0'})    // 011723543

Use of delimiter option which helps with large numbers e.g.

var format = require('biguint-format');

format([0x2A, 0xFF, 0x1E, 0x22, 0x11, 0x30, 0x12, 0x2F], 'bin')
format([0x2A, 0xFF, 0x1E, 0x22, 0x11, 0x30, 0x12, 0x2F], 'bin', {groupsize:8})

// returned values
0010101011111111000111100010001000010001001100000001001000101111        // no delimiter
00101010 11111111 00011110 00100010 00010001 00110000 00010010 00101111 // with delimiter

Example of trim option which works only with binary formatter

var format = require('biguint-format');
var buffer = new Buffer([0x1, 0xA7, 0x63]);

format(buffer, 'bin');              // returns 000000011010011101100011
format(buffer, 'bin', {trim:true}); // returns 11010011101100011

Example of padstr and size options

var format = require('biguint-format');
var buffer = new Buffer([0x1, 0xA7, 0x63]);

format(buffer, 'dec'); // returns 108387
format(buffer, 'oct'); // returns 323543
format(buffer, 'hex'); // returns 1a763

format(buffer, 'dec', {padstr:'0', size:6});  // returns 108387 - no padding effect
format(buffer, 'oct', {padstr:'0', size:10}); // returns 0000323543
format(buffer, 'hex', {padstr:'0', size:6});  // returns 01a763

Author

Writen by Tom Pawlak - Blog

License

Copyright (c) 2014 Tom Pawlak

MIT License : http://blog.tompawlak.org/mit-license