literal-gibberish v1.0.2

literal-gibberish

Generates a buffer of alphabetical pseudo-text. Alternative to Lorem ipsum sources.

• Random-length lines of random-length "words"
• Only letters and spaces in each line
• You control the buffer size
• Your choice of encoding*
• Your choice of end-of-line marker**
• No need to seed with input text
• No dependencies

* Only single-byte encodings in this version. See Options below.

** Of the EOL values available for single-byte encodings. Also see Options.

Install

$ npm install literal-gibberish

If you plan to use literal-gibberish for testing, you probably want to use the npm install option --save-dev.

Usage

var litGib = require('literal-gibberish');
var callback = function(error, data) {
  // Use data.lines to index into data.buffer
  // or simply use data.buffer...
}
var options = { // ...
}
litGib(options, callback);
// Or, to use all defaults:
litGib(callback);

Options

Optional. Supply options as an object containing any of the following properties.

size

The number of bytes in the resulting buffer. Must be a positive integer. Default: 16384 (16KB)

encoding

Determines what byte values are allowed in the buffer, according to the letters that are mapped in the referenced encoding. The choices are:

• 'ascii' (default)
• 'latin1' - 8-bit encoding Latin-1, a.k.a. ISO 8859-1
• 'win1252' - The notorious Windows-1252 encoding that is often mistaken to be the same as Latin-1

eol

The end-of-line marker to insert at each line end. The choices are:

• 'lf' (default) or the equivalent '\n'
• 'cr' or the equivalent '\r'
• 'crlf' or the equivalent '\r\n'
• 'nel' - the 8-bit NEXT LINE control character common to ISO 8859 encodings.
  Note: 'nel' is only valid with 'latin1' encoding.

EOL names in uppercase are also recognized.

Callback Parameters

error

Error handed down from crypto.randomBytes. Any assertion error thrown by misuse will not come this way.

data

An object with the following properties.

• buffer - Buffer populated with values for only letters, spaces, and the specified EOL marker
• lines - Array of objects, each with these two properties:
  • start - index of the start of the nth line in the buffer
  • end - index of the end of the nth line in the buffer

Examples

Basic

var gibberish = require('literal-gibberish');

var options = { size: 5000, encoding: 'win1252', eol: 'crlf' };
gibberish(options, function(error, data) {
  // ...
})
// Example extract of data.buffer.toString('binary'):
// "SëÿQeim ÿýÕyHoÚÖoË Jyàþtxxu qÊBóZcUÇÐQ àHZRyàÚüxåÊ mUDýBiâÜVuacs ..."

// No options specified - gets defaults
gibberish(function(error, data) {
  // ...
});
// Example extract of data.buffer.toString():
// "PpwjKt wjYEsnad vIhSLvJiUAAc qMtfcksRgdIV rWhiPI KtlPoWpkywnaMHB ..."

Extract Lines as Strings

gibberish(function(error, data) {
  data.lines.forEach(function(lineObj) {
    var s = data.buffer.toString('ascii', lineObj.start, lineObj.end);
    // Do something with s... maybe add punctuation?
  })
});

Convert to a Stream

var writable = getWritableStreamSomehow();
// ...
gibberish(function(error, data) {
  writable.write(data.buffer); // Or give encoding and callback args if desired
})

Questions

Who needs this gibberish?

• A developer who must push a lot of reasonably text-like data through another interface to test it.
• A web page developer who needs blocks of arbitrary filler text to help assess a layout.

How random is it?
It uses the crypto.randomBytes function that comes with node.js, so it gives "cryptographically strong pseudo-random data."

How long are the 'words'/lines?

• Pseudo-words are 1-16 letters
• Lines are 6-36 'words' (line length does not include EOL marker)

Are the 'words' pronouncable?
They're not made to be, so it's accidental if any are. Filler text is not meant to draw attention, never mind to be spoken.

Why do the lines not always start with a capital letter?
Why do the words contain mixed case?
The next version will provide case control, so that the lines look more like standard text at a glance.

Why no symbol characters?
I think they're distracting.

Why no punctuation?
Say, that's a good idea for a transform stream that you could plug this module into...

When I convert the buffer toString(), I get a lot of question mark symbols...?
You specified an encoding other than 'ascii', and then you didn't pass an encoding to toString(), so your bytes were "utf8-ized" because the default encoding for toString() is 'utf8'. Try toString('binary').

License: MIT