@allstar/reql-builder

reql-builder

Dynamic filter builder for rethinkdbdash inspired by https://github.com/node-tastypie/tastypie-rethink

Paging

You can use a number of special query string params to control how data is paged on the list endpoint. Namely -

• limit - Page size ( default 25 )
• offset - The starting point in the list

limit=25&offset=50 would be the start of page 3

Sorting

sorting is handled query param orderby where you can pass it the name of a field to sort on. Sorting is descending by default. Specifying a negetive field ( - ) would flip the order

Advanced Filtering

You might have noticed the filtering field on the schema. One of the things that makes an API "Good" is the ability to use query and filter the data to get very specific subsets of data. Tastypie exposes this through the query string as field and filter combinations. By default, the resource doesn't have anything enabled, you need to specify which filters are allowed on which fields, or specify 1 to allow everything

Filter Types

Filter	function
gt	greater than
gte	greater than or equal to
lt	less than
lte	less than or equal to
in	Value in set ( [ 1,2,3 ])
nin	Value Not in set
size	Size of set ( array length )
startswith	Case Sensitive string match
istartswith	Case Insensitive string match
endswith	Case Sensitive string match
iendswith	Case Insensitive string match
contains	Case Sensitive global string match
icontains	Case Insensitive global string match
exact ( = )	Exact Case Sensitive string match
iexact	Exact Case Insensitive string match
isnull	matches null values
month	Matches date values on a specific month
day	Matches date values on a speciec day of the week
year	Matches date values on a specific year
all	Matches all elements in an array
range	Matches dates withing a specified range
ne	not equal to

Filters are added by appending a double underscore __ and the filter type to the end of a field name. Given our example, if we wanted to find people who were older than 25, we would use the following URI syntax

http://localhost:3000/api/v1/user?age__gt=25

Filters are additive for a given field. For example, if we we only wanted people where we between 25 and 45, we would just add another filter

http://localhost:3000/api/v1/user?age__gt=25&age__lt=45

The same double underscore __ syntax is used to access nested fields where the filter is always the last parameter. So we could find people whos age was greater than 25, less than 45 and whose Company Name starts with W

http://localhost:3000/api/v1/user?age__gt=25&age__lt=45&company__name__istartswith=w

expressions with no filters will use the default exact filter

http://localhost:3000/api/v1/user?age=44

API

filter(r <rethinkdbdash>, filters <Object>): Query

• r - a rethinkdb dash connection w/ a default database configured
• filters - an object to generate a query from. Typically a result from querystring.parse
• returns a rethink filter query suitable to pass to r.filter

Given a table of user records

{
  "email": "[e-mail]"
, name: {
    "first": "Billy"
  , "last": "Bob"
  }
}

const qs = require('querystring')
const {filter} = require('@allstar/reql-builder')
const r = require('rethinkdbdash')({db: 'test'})

// emails w/ .biz
// name.first = ^bi

const filters = filter(r, {
  email__icontains: '.biz'
, name__first__istartswith: 'bi'
})

// same as

const opts = qs.parse('email__contains=.biz&name__first__istartswith=bi')
const filters = filter(r, opts)


r.table('users').filter(filters).then((results) => {
  console.log(results)
})

toResponse(r <rethinkdbdash>, table <String> [,options] <Object>): Promise

returns a paginated list of records suitable for an api response directly from the database

options

• limit <number> [25] the maximum number of records to return
• offset <number> [0] The number of records to skip over
• filters `[object] Filters to apply the the dataset before pagination
• orderby [string] The filed and direction (-/+) to sort the records by
• orderby_index [string] the name of the index to use to sort by - it should include the field

• With out an index to order by performance can degrade very quickly with large datasets

const toResponse = require('@allstar/reql-builder')
const r = require('rethinkdbdash')({db: 'test'})

toResponse(r, 'users', {
  orderby: '-id'
, orderby_index: 'user_id_idx'
, limit: 5
, offset: 5 // page 2
, filters: {email__iendswith: '.biz'}
})
.then(console.log)

{
  "meta": {
    total: 7
  , limit: 5
  , offset: 5
  , next: null
  , previous: null
  }
, "data" :[
    {id: 6, email: '[e-mail]'}
  , {id: 7, email: '[e-mail]'}
  ]
}