About

Symmetric provides observable model and collection classes with support for field specifications that mirror backend sources with encoding functions and validation rules.

Requirements

The built-in sync function requires a fetch compatible function on window/self/ or global.

• https://github.com/github/fetch
• https://github.com/bitinn/node-fetch
• https://github.com/matthew-andrews/isomorphic-fetch

NOTE: When using create-react-app, fetch is already included. No need to add an additional dependency.

Configuration

• autoValidate - auto validate model rules after X ms from a change - use -1 (default) to disable or 0 to run immediate
• encoders - dictionary of encoding (converting to a string) functions keyed by types
• decoders - dictionary of decoding (converting from a string) functions keyed by types
• inputEncodings - dictionary of field types that are be used ONLY with encoding/decoding user input
• formats - dictionary of field types to RegExp objects or functions that return a boolean
• formatErrorMessage - function to format errors
• batch - function to convert an array of fetch init objects into a single batched fetch init
• unbatch - function to convert a single Response object to many Response objects
• sync - function that wraps the call to fetch()
• syncConfig - additional settings for the built-in sync function, see sync below

sync function

Basic options are:

• url - required url to fetch
• params - will be added to url as query string args
• data - will set body with JSON.stringify(data) and Content-Type to application/json
• auth - any authorization object that implements a prepare(request, opts) function
• meta - provide an object here to receive with meta.set(key, value) the Link and X-Headers from the response as well as any meta data from unwrapping

Standard fetch options are:

https://developer.mozilla.org/en-US/docs/Web/API/WindowOrWorkerGlobalScope/fetch

• method - defaults to 'GET'
• headers - a dictionary of http headers to use, if setting 'Content-Type' note it is case-sensitive to avoid conflicts
• body - the body of the request
• mode - defaults to 'same-origin'
• credentials - defaults to 'same-origin'
• cache - defaults to 'no-cache'
• redirect - defaults to 'follow'

Additional options with defaults provided by config.syncConfig are:

• syncCamelCase - converts responses to camelCase keys before parsing. defaults to true
• saveSnakeCase - defaults to true
• querySnakeCase - defaults to true
• saveEncoding - one of the following strings json, form, or form-json, defaults to json
• saveArrayName - when saveEncoding is form or form-json, the form name to use when data is just an array
• csrfCookieName - defaults to csrftoken
• csrfHeaderName - defaults to X-CSRFToken
• requestedWith - value of the X-Requested-With header, defaults to XMLHttpRequest
• batchTimeout - will queue requests that happen within this timeframe into a single request, -1 (default) to disable or >= 0 to enable
• batchUrl - the backend url that accepts batched requests
• auth - an optional single auth object to use for all sync operations, defaults to null
• unwrap - function to unwrap and enveloped response and return the extracted data unwrap(data, meta) => unwrapped, defaults to null
• errorFactory - receives a Response object and should return or eventually resolve to rejected Promise
• errorHandler - a callback to globally handle errors before they are thrown to the next Promise rejection handler

Subclassing

Model

• toJSON() - return data for JSON stringification. The base method will encode each field
• defaults() - return an object of default values for the fields when creating a new instance
• field(key) - return the field specification for the given attribute key
• url(options, operation) - return the url to use with the fetch, save, or delete
• parse(data) - The base method will decode all fields that are a string and have an encoding type

Collection

The following methods should be implemented on subclasses:

• model(attrs) - return a model class used to construct with the given attributes
• url(options, operation) - return the url to use with the fetch or save
• parse(data) - parse raw data from a fetch. The base method converts each item into models using the model() method.

Parsing

The base behavior of the parse() method on the model class is to decode any string values that have a specified encoding. The base behavior of the parse() method on the collection class is to pass each item in the data array to a model class constructor and create model instances inside the returned array. This is where the model() method is used.

When subclassing and providing custom parsing it is a good idea to call the base method and use the data it returns.

Field Specifications

Each field can specify one or more of the following:

• choices - an array of valid choices the value must match
• encoding - type of encoding to use, keys will match up to config encoders and decoders
• rule - the validation rule, see below
• title - title of an input for this field
• subtitle - subtitle of an input for this field
• instructions - string of instructions to present to the user for how to enter a value. Will be returned by formatErrorMessage() if set

Validation Rules

Each rule is a dictionary with one or more of the following:

• required - if true, then the value may not be null or an empty string
• type - checks to make sure that the value is properly decoded as this type
• format - key into config formats to choose a RegExp or function to check
• equals - value should exactly equal this or if an array match one of the values
• min - values needs to pass a > test with this or the length must be longer than this if a string
• max - values needs to pass a < test with this or the length must be less than this if a string

Serialization

serialize and deserialize Functions are provided to make it easier for storing and retrieving already loaded models and collections to and from local or session storage.

License

MIT