reversible-hexy.js -- utility to convert to and from hex dumps

This is a fork of hexy.js by a2800276 that has been ported to TypeScript. Functionality has also been extended to support converting from hexdump files back into binary Buffers (similar to the xxd -r command). All credit goes to a2800276 for the original work that went into creating this library.

reversible-hexy is an easy to use javascript library to create hex dumps. It works just as well in node as in your browser. It contains a number of options to configure how the hex dump will end up looking.

It creates a pleasant looking hex dump by default:

var hexy = require('@kongsberggeospatial/reversible-hexy'),
    b = Buffer.from("\000\001\003\005\037\012\011bcdefghijklmnopqrstuvwxyz0123456789")
        // or String or Array containing numbers ( bytes, i.e. < 0xFF )

var str = hexy.hexy(b);

// Create hex dump
console.log(str);

// Create buffer
console.log(hexy.unHexy(str));

results in this dump:

 00000000: 0001 0305 1f0a 0962 6364 6566 6768 696a  .......bcdefghij
 00000010: 6b6c 6d6e 6f70 7172 7374 7576 7778 797a  klmnopqrstuvwxyz
 00000020: 3031 3233 3435 3637 3839                 0123456789

but you can configure:

• Line numbering
• Line width
• Format of byte grouping
• The case (upper/lower) of hex decimals
• Presence of the ASCII annotation in the right column.

This means it's easy to generate exciting dumps like:

 0000000: 0001 0305 1f0a 0962  .... ...b 
 0000008: 6364 6566 6768 696a  cdef ghij 
 0000010: 6b6c 6d6e 6f70 7172  klmn opqr 
 0000018: 7374 7576 7778 797a  stuv wxyz 
 0000020: 3031 3233 3435 3637  0123 4567 
 0000028: 3839                 89

or even:

 0000000: 00 01 03 05 1f 0a 09 62   63 64 65 66 67 68 69 6a 
 0000010: 6b 6c 6d 6e 6f 70 71 72   73 74 75 76 77 78 79 7a 
 0000020: 30 31 32 33 34 35 36 37   38 39

with hexy!

Accepted Input

Currently, input should be one of the following:

• a Buffer
• a String
• an Array containing Numbers. These should fit into 8 bits, i.e. be smaller than 255. Larger values are truncated (specifically val & 0xff)

Formatting Options

Formatting options are configured by passing a format object to the hexy function:

var format = {}
    format.width = width // how many bytes per line, default 16
    format.numbering = n // ["hex_bytes" | "none"],  default "hex_bytes"
    format.radix = b     // [2, 8, 10, 16], the radix for numeral representation
                         // for the right column,    default 16
    format.format = f    // ["twos"|"fours"|"eights"|"sixteens"|"none"], number of nibbles per group
                         //                          default "fours"
    format.littleEndian = true
                         // endiannes of data,       default false
                         // counts when number of nibbles is more than "twos",
                         // i.e. displaying groups bigger than one byte
    format.extendedChs = true
                         // allow displaying more characters in the text column
                         //                          default false
    format.caps = c      // ["lower"|"upper"],       default lower
    format.annotate = a  // ["ascii"|"none"], ascii annotation at end of line?
                         //                          default "ascii"
    format.prefix = p    // <string> something pretty to put in front of each line
                         //                          default ""
    format.indent = i    // <num> number of spaces to indent
                         //                          default 0
    format.html = true   // funky html divs 'n stuff! experimental.
                         //                          default: false
    format.offset = X    // generate hexdump based on X byte offset
                         // into the provided source
                         //                          default 0
    format.length = Y    // process Y bytes of the provide source 
                         // starting at `offset`. -1 for all
                         //                          default -1
    format.display_offset = Z
                         // add Z to the address prepended to each line
                         // (note, even if `offset` is provided, addressing
                         // is started at 0)
                         //                          default 0

console.log(hexy.hexy(buffer, format))

In case you're really nerdy, you'll have noticed that the defaults correspond to how xxd formats its output.

Installing

Either use npm (or whatever compatible npm thingie people are using these days) :

$ npm install @kongsberggeospatial/reversible-hexy

This will install the lib which you'll be able to use like so:

var hexy = require("@kongsberggeospatial/reversible-hexy"),
    buf  = // get Buffer from somewhere,
    str  = hexy.hexy(buf)

// For converting hexdumps into Buffers
var hexy = require("@kongsberggeospatial/reversible-hexy"),
   str  = // get string hexdump from somewhere
   buf  = hexy.unHexy(str);

It will also install reversible-hexy into your path in case you're totally fed up with using xxd.

If you don't like npm, grab the source from github:

 https://github.com/KongsbergGeospatial/reversible-hexy.js

Typescript

import {hexy, unHexy} from "@kongsberggeospatial/reversible-hexy";
const buff = ...
const str = ...
console.log(hexy(buff));
console.log(unHexy(str));

TODOS

The current version only pretty prints node.js Buffers, and JS Strings and Arrays. This should be expanded to also do typed arrays, Streams/series of Buffers which would be nice so you don't have to collect the whole things you want to pretty print in memory, and such.

I'd like to improve html rendering, e.g. to be able to mouse over the ascii annotation and highlight the hex byte and vice versa, improve browser integration and set up a proper build & packaging system.

Deno support would also be nice.

DONE Better testing for browser use.

Thanks

• Thanks to Isaac Schlueter [isaacs] for gratiously lending a hand and cheering me up.
• a2800276 (https://github.com/a2800276)
• dodo (http://coderwall.com/dodo)
• the fine folks at Travis
• radare (https://github.com/radare)
• Michele Caini (https://github.com/skypjack)
• Koen Houtman (https://github.com/automagisch)
• Stef Levesque (https://github.com/stef-levesque)
• Abdulaziz Ghuloum (https://github.com/azizghuloum)

History

This is a fairly straightforward port of hexy.rb which does more or less the same thing. You can find it here:

 http://github.com/a2800276/hexy

in case these sorts of things interest you.

0.3.5

• Forked from original repo
• Ported to pure TypeScript
• Added unHexy function; allows for parsing of hexdump files back into binary Buffer objects.

0.3.4

• issue concerning static analysis and BigInt usage

0.3.3

• introduced the concept of endiannes (googleable and wikiable). Before this change, the code assumed that the displayed data is big-endian. However, most file formats and most CPU architectures are little-endian. So, introduced the support for it. The endiannes can be controlled by passing bool via littleEndian, which defaults to false to support the behavior of the previous versions
• introduced ability to group 8 bytes (16 nibbles). With prevalence of 64-bit computing, the 64-bit (i.e. 8-byte) data is getting more and more popular. The 8-byte grouping is enabled by passing "sixteens" into config.format
• introduced ability to display the binary data in bases (radixes) other than hexadecimal: binary, octal, decimal and hexadecimal The radix is controlled by passing 2, 8, 10 or 16 into config.radix
• introduced ability to control if non-printable characters are displayed or replaced with '.'. To display extended characters, pass config.extendedChs: true. The exact behavior of this flag depends on the output type, html or not: In config.html: true mode, all the characters can be displayed, even 0-0x20 have visual represenation. In config.html: false mode, only the extended characters beyond the end of standard ASCII are displayed.
• implemented and exported maxnumberlen() -- calculates how many characters can a number occupy given bittness and radix
• several tweaks improved performance by ~15-30%, depending on the platform (compared to v.0.3.2).
• a bit more order in the node.js tests:
  • the tests are read from an uniform table. This allows enumerating the testcases, as well as sharing them with browser tests
  • added ability to do performance tests -- just run time node test perf
• enabled browser tests:
  • visual summary with details of all the tests, collapsable and color-coded
  • same set of testcases as in node.js
  • all tests pass now. Found and fixed a bug that was present in all browsers where they handle bigger-than-byte data differently compared to node.js
• created a static html page to hex display files (view.html)
• restricted the set of node.js versions and browsers (now requires support of BigInt: Node.JS 10.4+, browsers since 2018-2020)
• the Travis-ci is passing now
• nits:
  • removed some of unused variables
  • increased formating consistency

0.3.2

• documentation typos
• 2FA for npm publish

0.3.1

• use strict
• fixes undefined var. Thanks m-kircher!

0.3.0

• adds typescript support. Thanks Abdulaziz!
• remove support for old node versions (0.6-0.12)

Mail

In case you discover bugs, spelling errors, offer suggestions for improvements or would like to help out with the project, you can contact me directly ([email protected]).