Skip to content

anvilresearch/modinha-redis

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

modinha-redis

Build Status

The RedisDocument mixin for Modinha defines a collection of persistence methods that map cleanly between HTTP semantics and Redis data structures.

Usage

Suppose we've defined an Account model with Modinha like so:

var Modinha = require('modinha')
  , RedisDocument = require('modinha-redis').RedisDocument

var Account = Modinha.define('accounts', {
  email: { type: 'string', required: true, unique: true },
  role:  { type: 'string', secondary: true, enum: ['admin', 'editor', 'author'] },
  hash:  { type: 'string', private: true }
});

Account.extend(RedisDocument);

RedisDocument will add the following persistence methods to the Account document.

HTTP                     MODEL METHOD

GET    /accounts         Account.list(options, callback)
GET    /accounts/id      Account.get(ids, options, callback)
POST   /accounts         Account.insert(data, options, callback)
PUT    /accounts/id      Account.replace(id, data, options, callback)
PATCH  /accounts/id      Account.patch(id, data, options, callback)
DELETE /accounts/id      Account.delete(id, callback)

Extending Account with RedisDocument will also define the following properties on Account.schema:

_id:      { type: 'string', required: true, default: Model.defaults.uuid },
created:  { type: 'number', order: true, default: Model.defaults.timestamp },
modified: { type: 'number', order: true, default: Model.defaults.timestamp }

Since we defined unique and secondary properties on email and role, respectively, the mixin will also generate property specific methods for those indexes.

Account.getByEmail(email, callback)
Account.listByRole(role, callback)

More about indexing

We can index in a variety of ways with Redis hashes and sorted sets. For example, we could explicitly define our unique email index like so:

Account.defineIndex({
  type:  'hash',
  key:   'accounts:email',
  field: 'email',
  value: '_id'
});

This tells the model to store an account's _id property in a hash named accounts:email with email as the field name. Because this is a very common use of the hash type index, the mixin also provides a helper method for defining unique indices:

Account.indexUnique('email');

This is equivalent to adding unique: true to the property definition in our schema.

Sorted set indices get a little more interesting. We have a great deal of flexibility in how we can index our models. For example, suppose we have a Video model that has a category property and a likes property. We want to retrieve a list of videos for a specific category, sorted by the number of likes.

Video.defineIndex({
  type:   'sorted',
  key:    ['videos:#:$', 'category', 'category'],
  score:  'likes',
  member: '_id'
});

When we index the following instance...

{
  _id: 'r4nd0m',
  name: 'Awesome Presentation',
  url: 'https://youtube.com/wh4t3v3r'
  category: 'conferences',
  likes: 777
}

... the object's _id will be added to a sorted set in Redis called videos:category:conferences, with a score of 777. Notice the key property of the index definition: ['videos:#:$', 'category', 'category']. The first element of this array is a template for a key name. In the template, the placeholders # and $ will be replaced in order according to the remaining elements of the array. # will be replaced literally with element and $ will be used to access a property on the object being indexed.

Like the hash-type index, there are a few very common indexing patterns for sorted sets. The mixin provides higher level methods for defining these, and in some cases, they can be created as part of a schema definition. Some examples:

Model.indexSecondary(propertyName, [score]);
Video.indexSecondary('category', 'likes');                     // Same as previous example


Model.indexReference(propertyName, ReferencedModel, [score]);
Comment.indexReference('videoId', Video);                      // multi.zadd('videos:ID:comments', comment.created, comment._id);


Model.indexOrder(propertyName);
Comment.indexOrder('likes');


// video schema
{
  name:     { type: 'string', unique: true },
  url:      { type: 'string', unique: true },
  category: { type: 'string', enum: ['tutorial', 'presentation'], secondary: true },
  likes:    { type: 'string', order: true }
}

Unique values are enforced by the insert, replace, and patch methods. If you write custom methods, you can use Account.enforceUnique(callback) (for example) to generate a UniqueValueError.

The default timestamp methods define an ordered index for created and modified. Account.list(options, callback) uses the accounts:created index by default to deliver reverse chronological account listings.

The MIT License

Copyright (c) 2015 Anvil Research, Inc. http://anvil.io

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.