databob

(Also available in Scala and Kotlin flavours)

Given an example JS object, generates random examples of JavaScript objects for usage in tests. Think automatic builder objects where you only supply an example object.

Supports generation of object trees containing all of the primitive JS types, plus:

• NaN, ±Infinity, undefined, null
• native Dates
• arrays and nested arrays
• child objects
• GUIDs
• common timestamp formats, such as ISO8601
• pluggable custom formats
• "format safe" overriding of values

This is useful for a number of reasons:

• reduces need for boilerplate test-code/duplication
• increases resiliency of tests by enforcing explicit reliance only on important properties (rather than implicit properties of a commonly-built data object)
• simple "cut and paste" updating of data formats (which means you only need to update the example models in one place)
• strict-mode overriding will break should the template suddenly becomes inconsistent with example model

Installation

Via npm, simply run: npm install databob

Then instantiate a new databob with: var databob = require('databob')();

Features

Given an example object:

> var book = {
    name: 'lord of the rings',
    pages: 500,
    simpleNames: false,
    movie: [ 180.52, 'elijah wood', true],
    author: {
        name: 'tolkien',
        dead: true
    }
};

Make a randomised object from the passed model:

> databob.make(book)
{
     name: 'diam praesent',
     pages: 373,
     simpleNames: false,
     movie: [ 140.47, 'nunc metus', true ],
     author: { name: 'dictum in', dead: true }
}

Override any number of the values of the generated instance. By default, strict-mode is enabled so overriding non-existent values will blow up:

> databob.make(book, { name: 'Harry Potter and the English Accent' }, { pages: 999 });
{
     name: 'Harry Potter and the English Accent',
     pages: 999,
     simpleNames: false,
     movie: [ 121.42, 'lorem ipsum', true ],
     author: { name: 'ullamcorper', dead: false }
}

Merge additional values into the generated instance:

> databob.make(book, { ibsn: '978-3-16-148410-0' }, true);
 
{
     name: 'orci',
     pages: 26,
     simpleNames: true,
     movie: [ 54.55, 'purus', true ],
     author: { name: 'elementum', dead: false },
     ibsn: '978-3-16-148410-0'
}

Register a example model under a name:

> databob.register({
    Book: book
});

...then recall it repeatedly under that name to generate new instances:

> databob.Book();

...or use it as a part of another model:

> databob.register({
    Librarian: {
        FavouriteBook: databob.Book
    }
});

Override (or merge extra) properties of a generated instance using the same mechanism as above:

> databob.Book({ name: 'Harry Potter and the English Accent' }, { ibsn: '978-3-16-148410-1' } , true);

Extent of randomness

By default, generated instances will use the following rules for random values:

• Integers: The model value is used as a maximum
• Doubles: The ceil() of the model value is used as a maximum, and the number of decimal places is retained
• Strings (w/o spaces): Single words
• Strings (w spaces): A sentence with the number of words in the model value used as a maximum
• Strings (w newlines): Paragraph of text using the number of lines in the model value as a maximum
• String GUIDs: 5 Strings seperated by '-'
• Arrays: Like-for-like element generation to an identical length of array as the model value
• Dates & Timestamps: Uses current date & time

General Acks

• Datafixture.js for random value generation
• Moment.js for timestamp handling