RTTC

Runtime (recursive) type-checking for JavaScript.

This package is the official SDK for working with the RTTC type system. It includes a lot of methods suitable for everyday use, as well as some lower-level methods that are intended for developers building tools which leverage the machine specification.

What is RTTC?

Throwing errors in an asynchronous callback can be dangerous, particularly in Node.js. Too often, these types of errors and crashes occur because of a trivial mistake or miscommunication about the data type of a variable; particularly when you're working on a team with other developers.

RTTC is a lightweight type system that provides a safety net for JavaScript code. It provides flexible, performant type guarantees on an as-needed basis; without messing with your development stack or build tools. Instead, RTTC builds on top of the existing data structures and programming concepts from JavaScript and Node.js to validate and coerce data at runtime. This allows you to add as much or as little type-checking as you like, in any new or existing Node.js/Sails.js application.

RTTC semantics are used by:

• the Node-Machine project's core utility packages, including the machine runner
• the Sails.js framework core
• every Waterline driver
• every machinepack published on NPM, and
• the Treeline standard library

Installation

$ npm install rttc --save

Basic Usage

var rttc = require('rttc');

The rttc package has many different methods for working with fixtures, examples, type schemas, and runtime data in JavaScript. But the most commonly-used RTTC methods are related to validation and coercion of runtime data:

// If the value is valid vs. the specified type schema, then `.validateStrict()` simply returns undefined
rttc.validateStrict('number', 999);
// => undefined

// But if the provided value is **even slightly off**, then `.validateStrict()` throws.
rttc.validateStrict('number', '999');
// throws Error

// If the provided value is close-ish, `.validate()` coerces as needed to make it fit.
rttc.validate('number', '999');
// => 999

// But when confronted with **major** differences, `.validate()` throws too.
rttc.validate('number', { x: 32, y: 79 });
// throws Error

// As long as the provided type schema is valid, `.coerce()` **never** throws
rttc.coerce('number', '999');
// => 999

// When confronted with **major** differences, `.coerce()` returns the _base value_ for the given type
rttc.coerce('number', { x: 32, y: 79 });
// => 0

Unless otherwise stated, all RTTC methods support recursive (or "deep") traversal of values. In other words, they iterate over the keys of dictionaries (aka plain old JavaScript objects) and the indices of arrays-- and if those dictionary properties and array items are themselves dictionaries or arrays, then rttc recursively dives into them too (and so on and so forth).

For example:

rttc.coerce([ { name: 'string', age: 'number', friends: [ 'string' ] } ], [
  { name: 'Karl', age: 258 },
  { name: 'Samantha', age: '937' },
  { name: 'Lupé', age: 82, friends: ['Henry', 'Mario', undefined] },
  { name: 'Andres', age: '22' },
  { age: ['nonsense!'] }
]);
// => [
//  { name: 'Karl', age: 258, friends: [] },
//  { name: 'Samantha', age: 937, friends: [] },
//  { name: 'Lupé', age: 82, friends: ['Henry', 'Mario'] },
//  { name: 'Andres', age: 22, friends: [] },
//  { name: '', age: 0, friends: [] },
// ]
//

Next Steps

For a quick rundown of common use cases, as well as some additional examples, check out the RTTC quick start guide. Keep reading for a brief overview of how RTTC works and a tour of each of its data types. Or, if this isn't your first rodeo, feel free to skip ahead to the complete reference documentation below.

   

Types & Exemplars

There are 10 different types recognized by rttc, each of which is uniquely expressed by special notation called RTTC exemplar syntax. For example, if we were to interpret 'hello world' as an exemplar, we would be able to infer that it represents a string data type. Exemplars make it easier to write out intricate data structures and validation rules, because they allow us to reason about our data types using representative examples. Plus, when working with exemplars programatically, you have access to a much richer set of information about any given API. But most importantly: writing data types as examples makes it easier for other humans to read and understand our intentions.

Exemplars can be mixed and matched with varying levels of specificity, up to any imaginable depth, by using the recursive array and faceted dictionary types. For example, we can infer from the exemplar ['Rover'] that its type schema is ['string'], indicating that it accepts any array of strings. Similarly, given the exemplar [{ name: 'Rover' }], we can infer the type schema [{name: 'string'}]. This indicates that it accepts any array of dictionaries, so long as each of those dictionaries has a key called name with any string value.

The table below gives each of the RTTC types, the exemplar notation used to describe it, as well as its base value:

type	rttc exemplar syntax	type schema	base value
string	'any string like this'	'string'	''
number	1337 (any number)	'number'	0
boolean	false (or true)	'boolean'	false
lamda (aka function)	'->'	'lamda'	function () { throw new Error('Not implemented! (this function was automatically created by rttc'); };
generic dictionary	{}	{}	{} (empty dictionary)
json	'*'	'json'	null
ref	'==='	'ref'	null
faceted dictionary (recursive, w/ keys called "facets")	{...} (i.e. w/ facet:nested-exemplar pairs)	{...} (i.e. w/ facet:nested-type-schema pairs)	{...} (w/ facet:nested-base-value pairs)
array (recursive, w/ 1 item called the "pattern")	[...] (i.e. w/ pattern exemplar)	[...] (i.e. w/ pattern type schema)	[] (empty array)

A type's "base value" is its minimum empty state. When coercing some data vs. an exemplar, if coercion fails at a particular path within that exemplar, then the "base value" for the type will be used at that path instead. For example, if you are coercing the value {name:'Lynda'} vs. the exemplar {name: 'Angela', age: 47}, then the result would be {name: 'Lynda', age: 0} (because the base value for the number type is zero).

Compatibility Note

RTTC also supports an 11th type, sometimes called the "generic array" and represented by the exemplar ([]). However, if an exemplar is specified as [], it is really just an alias for ['*'], an array exemplar with a * (json) pattern. That means it accepts any array of JSON-compatible items.

The [] alias is for consistency with the generic dictionary type ({}), as well as for backwards compatibility. While future versions of RTTC will likely continue to maintain support for the [] exemplar, for clarity, you should switch to using ['*'] in new code and documentation and migrate [] to ['*'] in existing code at your earliest convenience.

Strings

Exemplar	RTTC Display Type	Display Type Label	Base Value
'foo' (any string)	'string'	'String'	''

The string type accepts any string.

Numbers

Exemplar	RTTC Display Type	Display Type Label	Base Value
32 (any number)	'number'	'Number'	0

The number type accepts integers and decimal numbers like 0, -4, or 235.3. Number-ish properties like Infinity, -Infinity and NaN (as well as -0) are all coerced to zero.

Booleans

Exemplar	RTTC Display Type	Display Type Label	Base Value
true (or false)	'boolean'	'Boolean'	false

The boolean type accepts true or false.

Functions (aka "lamdas")

Exemplar	RTTC Display Type	Display Type Label	Base Value
'->'	'lamda'	'Function'	(see below)

The lamda type accepts any function.

Base Value

The base value for the lamda type is the following automatically-generated function:

function () {
  throw new Error('Not implemented! (this function was automatically created by `rttc`');
};

Generic dictionaries

Exemplar	Type Schema	RTTC Display Type	Display Type Label	Base Value
{}	{}	'dictionary'	'Dictionary'	{} (empty dictionary)

The generic dictionary type accepts any JSON-serializable dictionary.

Dictionaries that have been validated/coerced against the generic dictionary type:

• will have no prototypal properties, getters, or setters, as well as a complete deficit of any other sort of dark magic
• are guaranteed to be JSON-serializable, with a few additional affordances:
  • normally, Error instances get stringified into empty objects. Instead, rttc turns them into human-readable strings by reducing them to their .stack property (this includes the error message and the stack trace w/ line numbers)
  • normally, RegExp instances get stringified into empty objects. Instead, rttc turns them into human-readable strings like '/some regexp/gi'
  • normally, function() instances get stringified into empty objects. Instead, rttc turns them into human-readable strings like 'function doStuff (a,b) { console.log(\'wow I can actually read this!\'); }'
• keys with undefined values at any level will be stripped out
• undefined items in nested arrays will be stripped out
• keys with null values may be present
• null items in nested arrays may be present

What about keys with undefined values?

When validating or coercing a value vs. a generic dictionary exemplar or type schema, keys with undefined values will always be stripped out. For example, coercing { name: 'Rob', age: undefined, weight: undefined } vs. the type schema {} would result in { name: 'Rob' }. This ensures consistency with the behavior of the native JSON.stringify() and JSON.parse() methods in browser-side JavaScript and Node.js.

Note that undefined array items are stripped out even if you are using ['==='].

JSON-Compatible Values

Exemplar	Rttc Display Type	Display Type Label	Base Value
'*'	'json'	'JSON-Compatible Value'	null

This works pretty much like the generic dictionary type, with one major difference: the top-level value can be a string, boolean, number, dictionary, array, or null value. When faced with a dictionary or array that contains nested values, the generic JSON type follows the same JSON-serializability as the generic dictionary type (see above).

Mutable references (aka "ref")

Exemplar	RTTC Display Type	Display Type Label	Base Value
'==='	'ref'	'Anything'	null

This special type allows anything except undefined at the top level (undefined is permitted at any other level). It also does not rebuild objects, which means it maintains the original reference (i.e. is ===). It does not guarantee JSON-serializability.

Faceted dictionaries

Exemplar	Type Schema	RTTC Display Type	Display Type Label	Base Value
{...} (recursive)	{...} (see below)	'dictionary'	'Dictionary'	{...} (see below)

The faceted dictionary type is any dictionary type schema with at least one key. When coercing a value to a faceted dictionary, any keys in the value that are not in the type schema will be stripped out. Missing keys in the value will cause .validate() to throw.

Dictionary type schemas (i.e. plain old JavaScript objects nested like {a:{}}) can be infinitely nested. Type validation and coercion will proceed through the nested objects recursively.

{
  id: 'number',
  name: 'string',
  isAdmin: 'boolean',
  mom: {
    id: 'number',
    spouse: 'json',
    occupation: {
      title: 'string',
      workplace: 'json',
      hobbies: {},
      incomingUploads: [
        {
          fd: 'string',
          startBuffering: 'lamda',
          rawStream: 'ref'
        }
      ]
    }
  }
}

Base Value

The base value for the faceted dictionary type is a dictionary which consists of a key for every expected facet, each with its own base value (recursively deep).

For example, for the type schema described above, the base value is:

{
  id: 0,
  name: '',
  isAdmin: false,
  mom: {
    id: 0,
    spouse: null,
    hobbies: {},
    occupation: {
      title: '',
      workplace: null,
      incomingUploads: []
    }
  }
}

What about keys with undefined values?

When validating or coercing a value vs. a faceted dictionary exemplar or type schema, if a required key exists, but has an undefined value, it is considered the same thing as if the key did not exist at all. This is a deliberate decision designed to normalize the use of null vs. undefined in your application, and to avoid the pitfalls of == vs. === equality comparisons and hasOwnProperty checks. This approach prevents countless bugs and makes it much easier to hunt down the sources of problems when they occur.

Can I have optional facets?

The best way to allow dictionaries which may or may not include certain keys is to always provide those keys using the appropriate base values. For example, let's say a user in your app or script may or may not have msOutlookEmail, contactInfo, or misc keys. Regardless, you could still use the same faceted dictionary exemplar:

var USER_SCHEMA = {
  id: 38,
  name: 'Margaret Thatcher',
  email: '[email protected]',
  msOutlookEmail: '[email protected]',
  contactInfo: {},
  misc: '*'
}

Then whenever you build a dictionary that you want to validate at runtime, just coerce it first:

var alfred = rttc.cast(USER_SCHEMA, {
  id: 100,
  name: 'Alfred Roberts',
  email: '[email protected]',
  contactInfo: {
    phone: '+3 9284829424'
  }
});

This will ensure that all of the facets exist, even if it's just as base values; e.g.:

console.log(alfred);

// - - - - - - - - - - - - -
{
  id: 100,
  name: 'Alfred Roberts',
  email: '[email protected]',
  msOutlookEmail: '',
  contactInfo: {
    phone: '+3 9284829424'
  },
  misc: null
}

Can I have facets that could be vastly different types?

The best way to implement union facets, or facets that could be more than one type, is to use a more generic type (such as {}, '*', or '==='), and then add additional specificity through custom code.

What if I don't need to validate every key recursively deep?

Even if you don't need to validate every key recursively deep, to use the faceted dictionary exemplar, you still need to declare every facet you plan to use. Alternately, to indicate any dictionary of JSON-compatible values, just use the generic dictionary ({}) type / exemplar instead. Then, implement any additional validation or coercion logic you need on top of that by writing it into your code.

This is another way you can go about validating dictionaries with keys that may or may not exist, and keys that could be multiple different types. Just be careful: this approach has the problem of introducing human error into the equation. If possible, the best, safest, and most foolproof approach is to use a faceted dictionary. This ensures your runtime dictionaries always include the key/value pairs you expect them to, and it reduces the number of times you have to squint at the computer screen and read Cannot read property "foo" of undefined.

Arrays

Exemplar	Type Schema	RTTC Display Type	Display Type Label	Base Value
[...] (recursive)	[...] (see below)	'array'	'Array'	[] (empty array)

The array type accepts any array, so long as all of that array's items are also valid (recursively deep). Every array exemplar and type schema must declare a pattern: a nested exemplar or type schema which indicates the expected type of array items. This pattern is how the array type is able to validate nested values. When validating vs. an array type schema, RTTC first checks that the corresponding value is an array (a la _.isArray()), then also recursively checks each of its items vs. the expected pattern. For example, given the exemplar ['Margaret'], we can infer that the type schema is ['string'], and therefore that it would accept any array of strings. So when designing an array exemplar or type schema, make sure the array has exactly one item to serve as the pattern, which is itself another exemplar or type schema. This pattern will be used for validating/coercing array items.

An array type schema or exemplar may be infinitely nested simply by using another array or a faceted dictionary as its pattern. For example:

[
  {
    id: 'number',
    name: 'string',
    email: 'string',
    age: 'number',
    isAdmin: 'boolean',
    favoriteColors: ['string'],
    friends: [
      {
        id: 'number',
        name: 'string'
      }
    ]
  }
]

What if I don't need to validate array items recursively deep?

Even if you don't need to validate array items, you still need a pattern. But luckily, there's another RTTC type ('===') that you can easily use to accept any value. To indicate an array of anything, just use the mutable reference type as your pattern:

// Type schema that indicates an array of anything, where array items are passed by reference:
['ref']

// Exemplar that indicates an array of anything, where array items are passed by reference:
['===']

What about undefined?

When validating or coercing a value vs. an array exemplar or type schema, undefined items in the array will always be stripped out. For example, coercing ['Jerry', undefined, undefined, 'Robin'] vs. the type schema ['string'] would result in ['Jerry', 'Robbin']. This ensures consistency with the behavior of the native JSON.stringify() and JSON.parse() methods in browser-side JavaScript and Node.js.

Note that undefined array items are stripped out even if you are using ['==='].

 

Conventions and edge-cases

The following is a high-level overview of important conventions used by the rttc module. For detailed coverage of every permutation of validation and coercion, check out the declarative tests in the spec/ folder of this repository.

undefined and null values

• undefined is never valid as a top-level value against ANY type, even mutable reference (===)
• undefined IS, however, allowed as an item in a nested array or value in a nested dictionary, but only within a dictionary or array being validated against the mutable reference type (===)
• null is only valid against the JSON (*) and mutable reference (===) types.

Weird psuedo-numeric values

• NaN is only valid against the mutable reference type ('===')
• Infinity and -Infinity are only valid against the mutable reference type ('===')
• +0 and -0 are always coerced to 0 (except against the mutable reference type)

Instances of ECMAScript core classes

When coerced against the generic dictionary or generic json types, the following is true:

• Error instances are coerced to the string value of their .stack property (i.e. the message + stack trace you're used to seeing in the terminal)
• Date instances are coerced to the string value of running their .toJSON() method (a ISO-8601 timestamp, e.g. '2015-05-24T15:16:48.999Z'. This reflects the Date in GMT/UTC time, so is therefore timezone-agnostic).
• RegExp instances are coerced to the string value you get from running their .toString() method (e.g. '/foo/' or '/^bar/gi')
• Functions are coerced to the string value you get from running their .toString() method (e.g. 'function someFunction (some,args,like,this,maybe){ /* and some kind of implementation in here prbly */ }')

Instances of Node.js core classes

• Stream and Buffer instances (from Node.js) are only valid against the mutable reference type.
• Streams and Buffers are coerced to null against the generic dictionary or the generic json types.

Base values

As mentioned above, every type has a base value.

• For the "string" type, base value is ""
• For the "number" type, base value is 0
• For the "boolean" type, base value is false
• For the "lamda" type ('->'), base value is a function that uses the standard machine fn signature and triggers its "error" callback w/ a message about being the rttc default (e.g. function(inputs,exits,env) { return exits.error(new Error('not implemented')); })
• For the generic dictionary type ({}) or a faceted dictionary type (e.g. {foo:'bar'}), the base value is {}.
• For the array type (e.g. [3] or [{age:48,name: 'Nico'}]), the base value is [] (an empty array)
• For the "json" type ('*'), base value is null.
• For the "ref" type ('==='), base value is null.

Note that, for both arrays and dictionaries, any keys in the schema will get the base value for their type (and their keys for their type, etc. -- recursive)

Methods

This package exposes a number of different methods, some of which are much more likely to be relevant than others for your everyday development needs. The methods in this reference documentation are listed roughly in descending order of familiarity, starting with the most commonly-used and ending with the more bizarre.

Validation

.validateStrict(expectedTypeSchema, actualValue)

Throw an error if the provided value is not the right type (recursive).

.validate(expectedTypeSchema, actualValue)

Either return a (potentially "lightly" coerced) version of the value that was accepted, or throw an error. The "lightly" coerced value turns "3" into 3, "true" into true, -4.5 into "-4.5", etc.

.isEqual(firstValue, secondValue, [expectedTypeSchema=undefined])

Determine whether two values are equivalent using _.isEqual().

This is the method used by rttc's own tests to validate that expected values and actual values match. If the third argument is provided, .isEqual also looks for expected lamda values in the optional type schema and calls toString() on functions before comparing them.

Munging

.coerce(expectedTypeSchema, actualValue)

ALWAYS return an acceptable version of the value, even if it has to be mangled (i.e. by using the "base value" for the expected type schema).

.rebuild(value, handlePrimitive, [handleComposite])

Recursively rebuild (non-destructively) the specified value using the provided transformer function (handlePrimitive) to potentially modify each primitive (null, string, number, boolean, or function) therein. Values like JavaScript Dates, Errors, streams, etc. are coerced to strings before being passed in to handlePrimitive.

The handlePrimitive transformer function is not run for dictionaries or arrays, since they're recursed into automatically by default-- unless the handleComposite transformer is provided. If provided, the handleComposite transformer function is called once for each array and once for each dictionary in value. It is expected to return a modified dictionary or array that will then continue to be recursively iterated into by rebuild().

In any case, arrays and dictionaries end up as normal array and dictionary literals in the rebuilt value, meaning that any JavaScript-language-specific metadata such as getters/setters/non-enumerable properties like prototypal methods and constructor information are all stripped out. .rebuild() also protects against endless recursion due to circular references, whether or not the handleComposite transformer function is being used (since even if it is provided, JSON-serializability is ensured before it is called).

Both transformer functions should be written expecting the particular primitive, dictionary or array value as their first argument and an RTTC display type string as the second argument. For handlePrimitive, that second argument is either 'string', 'number', 'boolean', 'lamda', or 'null'. For handleComposite, it is either 'dictionary' or 'array'.

If you need further technical specifics, see the implementation of rebuild() in lib/rebuild.js in this repo.

Example usage:

return res.json(rttc.rebuild(someData, function handlePrimitive(val, type){
  if (type === 'string') { return val + ' (a grass-type Pokemon)'; }
  else { return val; }
}));

.dehydrate(value, [allowNull=false], [dontStringifyFunctions=false])

Prepare a value for serialization by taking care of a few edge-cases, such as:

• stringifying regexps, errors (grabs the .stack property), and functions (unless dontStringifyFunctions is set)
• replacing circular references with a string (e.g. [Circular])
• replacing -Infinity, Infinity, and NaN with 0
• stripping keys and array items with undefined or null values. If allowNull is set to true, null values will not be stripped from the encoded string.

Note that arrays, dictionaries and literals are not stringified by dehydrate. Rather, dehydrate prepares a value for stringification (see rttc.stringify() below).

.hydrate(value, [typeSchema=undefined])

Use the provided typeSchema to figure out where "lamda" values (functions) are expected, then use eval() to bring them back to life. Use with care.

.parseHuman(stringFromHuman, [typeSchema=undefined], [unsafeMode=false])

Parse a human-readable string (typically entered by a human into some kind of UI or CLI application) and return a best guess at the JavaScript value it represents.

• If provided, typeSchema will be used to make a more educated guess. If you are calling parseHuman() in order to parse a string that was generated using stringifyHuman(), then be sure to use the same type schema.
• If the unsafeMode flag is enabled, lamda functions will be hydrated.

For example:

var result;


result = rttc.parseHuman('hi');
// result === 'hi'

result = rttc.parseHuman('"hi"', 'string');
// result === '"hi"'

result = rttc.parseHuman('"hi"', 'json');
// result === 'hi'
// typeof result === 'string'




result = rttc.parseHuman('3');
// result === '3'
// typeof result === 'string'

result = rttc.parseHuman('3', 'number');
// result === 3
// typeof result === 'number'

result = rttc.parseHuman('3', 'json');
// result === 3
// typeof result === 'number'


// JSON is not parsed by default:
result = rttc.parseHuman('{"foo":"100"}')
// result === '{"foo":"100"}'
// typeof result === 'string'

// But it will be, if given the proper type schema:
// e.g. either one of these:
result = rttc.parseHuman('{"foo":"100"}', 'json')
result = rttc.parseHuman('{"foo":"100"}', {})
// assert.deepEqual(result, { foo: '100' })
// typeof result === 'object'
// typeof result.foo === 'string'

// But:
result = rttc.parseHuman('{"foo":"100"}', { foo: 'number' })
// assert.deepEqual(result, { foo: 100 })
// typeof result === 'object'
// typeof result.foo === 'number'

Another example, this time using a more complex type schema:

var result = rttc.parseHuman('{"name":"Mr. Tumnus","friends":[{"name":"Broderick","age":13},{"name":"Ashley","age":8000}]}',{
  name: 'string',
  friends: [
    {
      name: 'string',
      age: 'number'
    }
  ]
});

// Results in the following dictionary:
//
// { name: 'Mr. Tumnus',
//   friends:
//    [ { name: 'Broderick', age: 13 },
//      { name: 'Ashley', age: 8000 } ] }

.stringifyHuman(value, typeSchema)

Convert a JavaScript value into a string that can be parsed by parseHuman().

Specifically, this method is an inverse operation of .parseHuman(); that is, if you take the stringified result from this function and pass that in to .parseHuman() using the same type schema, you'll end up back where you started: with the original JavaScript value you passed in to rttc.stringifyHuman().

This losslessness is guaranteed by two factors: that stringifyHuman() (1) enforces strict RTTC validation rules (i.e. rttc.validateStrict(typeSchema, value)) and (2) the fact that it rejects values which cannot be safely stringified in a reversible way (e.g. JavaScript Dates, Errors, streams, prototypal objects, dictionaries and arrays with circular references, etc.). If either of these checks fails, stringifyHuman() throws an error.

So even though parseHuman() is quite forgiving (it uses RTTC loose validation), you can rest assured that any string you generate using stringifyHuman() will be properly deserialized by parseHuman(), provided it is passed in with the same type schema.

For example:

var result;

// Basic usage:
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
result = rttc.stringifyHuman(100, 'number');
// result === 100
// typeof result === 'number'


// The method performs strict validation, so the value must be compatible with the provided type schema:
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
result = rttc.stringifyHuman('100', 'number');
// Error: rttc.stringifyHuman() failed: the provided value does not match the expected type schema.
// Details:
// Error: 1 error validating value:
//  • Specified value (a string: '100') doesn't match the expected type: 'number'
//     at consolidateErrors (/Users/mikermcneil/code/rttc/lib/helpers/consolidate-errors.js:45:13)
//     ...


// And even if you specify the `ref` type, non-serializable things are not allowed:
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
result = rttc.stringifyHuman(new Date(), 'ref');
// Error: rttc.stringifyHuman() failed: the provided value cannot be safely stringified in a reversible way.
//    at Object.stringifyHuman (/Users/mikermcneil/code/rttc/lib/stringify-human.js:49:11)
//    at repl:1:15
//    ...


// One more normal-case usage scenario, this time using the same more complex value and type schema
// from the `parseHuman()` example above:
// - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
var result = rttc.stringifyHuman({
  name: 'Mr. Tumnus',
  friends: [
    { name: 'Broderick', age: 13 },
    { name: 'Ashley', age: 8000 }
  ]
},
{
  name: 'string',
  friends: [
    {
      name: 'string',
      age: 'number'
    }
  ]
});
// result === '{"name":"Mr. Tumnus","friends":[{"name":"Broderick","age":13},{"name":"Ashley","age":8000}]}'
// typeof result === 'string'

.compile(value)

Given a value, return a human-readable string which represents it. This string is equivalent to a JavaScript code snippet which would accurately represent the value in code.

This is a lot like util.inspect(val, {depth: null}) in the Node core util package. But there are a few differences. rttc.compile() also has special handling for Errors, Dates, and RegExps (using dehydrate() with allowNull enabled), as well as for Functions (making them eval()-ready.) The biggest difference is that the string you get back from rttc.compile() is ready for use as the right hand side of a variable initialization statement in JavaSript.

Useful for:

• bootstrapping data on server-rendered views for access by client-side JavaScript
  • see this gist for a comprehensive example
• generating code samples
• error messages
• debugging
• user interfaces

Finally, here's a table listing notable differences between util.inspect() and rttc.compile() for reference:

value	util.inspect()	rttc.compile()
a function	[Function: foo]	function foo (){}
a Date	Tue May 26 2015 20:05:37 GMT-0500 (CDT)	'2015-05-27T01:06:37.072Z'
a RegExp	/foo/gi	'/foo/gi/'
an Error	[Error]	'Error\n at repl:1:24\n...'
a deeply nested thing	{ a: { b: { c: [Object] } } }	{ a: { b: { c: { d: {} } } } }
a circular thing	{ y: { z: [Circular] } }	{ y: { z: '[Circular ~]' } }
undefined	undefined	null
[undefined]	[undefined]	[]
{foo: undefined}	{foo: undefined}	{}
Infinity	Infinity	0
-Infinity	-Infinity	0
NaN	NaN	0
Readable (Node stream)	{ _readableState: { highWaterMar..}}	null
Buffer (Node bytestring)	<Buffer 61 62 63>	null

Note that undefined values in arrays and undefined values of keys in dictionaries will be stripped out, and circular references will be handled as they are in util.inspect(val, {depth: null}).

.parse(stringifiedValue, [typeSchema=undefined], [unsafeMode=false])

Parse a stringified value back into a usable value.

This is basically just a variation on JSON.parse that calls rttc.hydrate() first if unsafeMode is enabled.

.stringify(value, [allowNull=false])

Encode a value into a string.

This is basically just a variation on JSON.stringify that calls rttc.dehydrate() first.

Meta

Methods for working w/ exemplars, type schemas, and display types

.infer(exemplar)

Infer the type schema from the given exemplar.

rttc.infer([
  {
    name: 'Rachael',
    age: 27,
    filesBeingUploaded: ['==='],
    friends: [
      {
        name: 'Mr. Bailey',
        species: 'cat',
        getClawSharpness: '->'
      }
    ]
  }
]);

/* =>
[
  {
    name: 'string'
    age: 'number',
    filesBeingUploaded: [
      'ref'
    ],
    friends: [
      {
        name: 'string',
        species: 'string',
        getClawSharpness: 'lamda'
      }
    ]
  }
]
*/

.inferDisplayType(exemplar)

Compute the display type (aka "typeclass") for an RTTC exemplar.

Always returns one of the standard RTTC display types:

• string
• number
• boolean
• lamda
• dictionary
• array
• json
• ref

Or '' (empty string) if the exemplar is unrecognized or invalid; e.g. null.

rttc.inferDisplayType({foo: 'bar'});
// => 'dictionary'


rttc.inferDisplayType('->');
// => 'lamda'

.getDisplayTypeLabel(displayType)

Get the appropriate human-readable label for a given RTTC "display type" (aka "typeclass") string.

Useful for error messages, user interfaces, etc.

rttc.getDisplayTypeLabel('ref');
//   => 'Anything'

rttc.getDisplayTypeLabel('string');
//   => 'String'

rttc.getDisplayTypeLabel('dictionary');
//   => 'Dictionary'

.coerceExemplar(value, [allowSpecialSyntax=false])

Build a reasonable-looking exemplar from a normal value-- specifically, the most specific exemplar which would accept that value.

Note: This is particularly useful for inferring RTTC exemplar schemas from fixture data.

In most cases, this leaves the value untouched-- however it does take care of a few special cases:

• Empty dictionaries become generic dictionaries ({}). The most specific exemplar which can accept an empty dictionary is the generic dictionary.
• Empty arrays become generic arrays ([]). Since we don't know the contents, we have to assume this array could be heterogeneous (i.e. have items with different types).
• Multi-item arrays become pattern arrays, and any extra items (other than the first one) are lopped off.
• Functions become '->'.
• null becomes '*'.
• If the top-level value is undefined, it becomes '==='. (however this behavior is subject to change in an upcoming release; since undefined is not supported by any exemplar)
• '->' becomes the string: 'an arrow symbol'.
• '*' becomes the string: 'a star symbol'.
• '===' becomes the string: '3 equal signs'.
• NaN, Infinity, -Infinity, and -0 become 0.
• Nested array items and keys with undefined values are stripped.
• Other than the exceptions mentioned above, non-JSON-serializable things (like circular references) are boiled away when this calls dehydrate internally.

If the allowSpecialSyntax flag is enabled, then *, ->, and === will be left untouched (allowing them to be intperpreted as special rttc exemplar syntax) instead of being replaced with string samples (e.g. "a star symbol" or "an arrow symbol").

rttc.coerceExemplar([{a:null}, {b: [[74,39,'surprise string!']] }])
//   =>   [ {} ]

rttc.coerceExemplar([74,39,'surprise string!'])
//   =>   [ 'surprise string!' ]

rttc.coerceExemplar({x:'*'})
//   =>   { x: 'a star symbol' }

rttc.coerceExemplar({x:'*'}, true)
//   =>   { x: '*' }

.getPathInfo(exemplar, path)

Given an exemplar schema and a keypath, return information about the specified segment. If the path is inside of a generic, then the exemplar is '*', and this path is optional. If the path is inside of a ref, then the exemplar is '===', and this path is optional. If the path is not reachable (i.e. inside of a string, or lamda... or something) then throw an error.

WARNING: Since hops in keypaths are represented by . (dots), this method is not safe to use on exemplars which contain any keys which contain dots. This may be improved in future versions.

var SOME_EXEMPLAR = {
  salutation: 'Mr.',
  hobbies: ['knitting'],
  medicalInfo: {
    numYearsBlueberryAbuse: 12.5,
    latestBloodWork: {}
  }
};

rttc.getPathInfo(SOME_EXEMPLAR, 'hobbies.238');
// =>
//     {
//       exemplar: 'knitting',
//       optional: false
//     }


rttc.getPathInfo(SOME_EXEMPLAR, 'medicalInfo.latestBloodWork.whiteBloodCellCount');
// =>
//     {
//       exemplar: '*',
//       optional: true
//     }

.union(schema0, schema1, [isExemplar=false], [isStrict=false])

Given two rttc schemas (e.g. A and B), return the most specific schema that would accept the superset of what both schemas accept normally (A ∪ B).

• schema0 - the first schema
• schema1 - the second schema (order doesn't matter)
• isExemplar - if set, the schemas will be treated as exemplars (rather than type schemas)
• isStrict - if set, the schemas will be unioned using strict validation rules (i.e. like validateStrict())

.intersection(schema0, schema1, [isExemplar=false], [isStrict=false])

Given two rttc schemas, return the most specific schema that accepts the shared subset of values accepted by both. Formally, this subset is the intersection of A and B (A ∩ B), where A is the set of values accepted by schema0 and B is the set of values accepted by schema1. If A ∩ B is the empty set, then this function will return null. Otherwise it will return the schema that precisely accepts A ∩ B.

• schema0 - the first schema
• schema1 - the second schema (order doesn't matter)
• isExemplar - if set, the schemas will be treated as exemplars (rather than type schemas)
• isStrict - if set, the schemas will be intersected using strict validation rules (i.e. like validateStrict())

Convenience

Simple convenience methods that wrap up one or more of the other rttc methods for a particular use case.

.getBaseVal(exemplar)

A convenience method to return the base value for the given exemplar. This is effectively the same thing as calling rttc.infer() to get the exemplar's type schema, then coercing undefined to match it (i.e. passing the type schema to rttc.coerce() without a second argument).

rttc.getBaseVal(exemplar);
// ... is just a shorcut for doing:
rttc.coerce(rttc.infer(exemplar), undefined);

.cast(exemplar, actualValue)

A convenience method that calls rttc.infer() on the provided exemplar to get the type schema, then uses it to rttc.coerce() the actualValue provided.

rttc.cast(exemplar, actualValue);

// ... is just a shorcut for doing:
rttc.coerce(rttc.infer(exemplar), actualValue);

Experimental

The following functions are newly implemented, experimental, and tend to be a bit more advanced. They may undergo frequent changes over the coming months, so use with care. You have been warned!

.sample(typeSchema, [n=2])

Given a type schema, return an array of up to n unique sample values that would validate against it (in random order). n defaults to 2 if left undefined.

.isSpecific(typeSchemaOrExemplar, [recursive=false], [isExemplar=false])

Determine whether the given type schema is "specific". String, number, boolean, lamda, faceted dictionary, or patterned array types are "specific". Everything else is "generic".

If the second argument (recursive) is set to true, then also recursively check the subkeys of faceted dictionaries and patterns of arrays in the type schema.

If the third argument (isExemplar) is set to true, then treat the provided schema as an rttc example rather than a type schema.

For reference

type	is specific?
string	yes (always)
number	yes (always)
boolean	yes (always)
lamda	yes (always)
{} (generic)	no
[] (generic)	no
{...} (faceted)	yes (maybe recursively)
[...] (patterned)	yes (maybe recursively)
json	no
ref	no

.getDefaultExemplar(typeSchema)

Given a type schema, return a random exemplar which accepts precisely the same set of values.

License

MIT

© 2014 Mike McNeil, Cody Stoltman; © 2015-2016 The Treeline Company