nxus-storage
v4.0.9
Published
Storage framework for Nxus applications
Downloads
31
Readme
nxus-storage
Storage Module
A storage framework for Nxus applications using waterline.
Configuration
"config": {
"storage": {
"adapters": {
"default": "sails-mongo"
},
"connections": {
"default": {
"adapter": "default",
"url": "mongodb://...."
}
},
}
}
Defining models
You define models by extending a base model class: BaseModel is a general-purpose base class; GeoModel is a base class for models containing a GeoJSON attribute. You can also define your own base class by extending BaseModel or GeoModel.
(The PointModel base class, for models containing a GeoJSON coordinate point attribute, is deprecated; use GeoModel instead.)
Using models
HasModels is a base class for nxus modules that define or use
models. It automatically registers all model definitions contained in
the module ./models
subdirectory. It makes model definitions
available through its HasModels#models property, which you
can configure to load models defined by your module or other modules.
The Storage model() and modelDir() methods let you explicitly register model definitions, and you can use getModel() to load definitions. However, these methods are rarely used, since using the HasModels base class is typically more concise and convenient.
Model events
The storage model emits events for create, update, and destroy. You can register a handler for all events:
storage.on('model.create', (identity, record) => {})
storage.on('model.update', (identity, record) => {})
storage.on('model.destroy', (identity, record) => {})
Or just a specific model identity:
storage.on('model.create.user', (identity, record) => {})
storage.on('model.update.user', (identity, record) => {})
storage.on('model.destroy.user', (identity, record) => {})
Lifecycle notes
load
- Models should be registered during
load
, e.g. var User = BaseModel.extend({ identity: 'user', ... }); application.get('storage').model(User)
- Models should be registered during
startup
The configured database is connected during
load.after
You can query models from
startup
and beyond, retrieve the model by the 'identity':application.get('storage').getModel('user').then((User) => { User.create(...); });
API
Storage
Extends NxusModule
Storage provides a common interface for defining models. Uses the Waterline ORM.
model
Register a model
Parameters
model
object A Waterline-compatible model class
Examples
storage.model(...)
getModel
Request a model based on its identity (name)
Parameters
Examples
storage.getModel('user')
Returns Promise The model class(es)
modelDir
Register all models in a directory
Parameters
dir
string Directory containing model files
Examples
application.get('storage').model(...)
Returns Promise Array of model identities
getWaterlineConfig
After init, get the waterline config with populated adapter modules
Returns Promise Config object for waterline
HasModels
Extends NxusModule
Base class for nxus modules that define or use models.
It automatically registers all model definitions contained in
the module ./models
subdirectory. It makes model definitions
available through its #HasModels#models property, which you
can configure to load models defined by your module or other modules.
Parameters
options
Object Configuration options. (optional, default{}
)options.modelNames
(Object | Array) Model definitions to load. If specified as an object, each object property specifies a model – its key is a model identity (the model'sidentity
property), and its value is used as the model's name in HasModels#models. If specified as an array, each array element specifies a model – the element string serves as both model identity and name. (optional, defaultnull
)
models
Collection of model definitions.
You can use the constructor modelNames
option to configure which
model definitions to load. By default, the model definitions
that were registered from the module ./models
subdirectory are
also loaded into the model definitions.
Type: Object
modelNames
Override to define the model names to access.
Examples
modelNames() {
return ['user']
}
Returns (Array | Object) Model identities to add to this.models
, or object of {identity: name}
Meta
- deprecated: Use the constructor
modelNames
option to specify names of models to be made available in thethis.models
property.
BaseModel
Extends Waterline.Collection
Base class for Waterline model definitions.
It extends the Waterline.Collection
to make it less awkward to use
and better adapted to the nxus environment.
It provides an extend() method that is
better-behaved than the Waterline extend()
method.
It emits create, update, and destroy events for model instances. The model identity and record are passed as parameters to the handler. You can register to be notified of all events of a specified type, or restrict notifications to a specific model by specifying its identity as a suffix to the event type:
model.create
,model.create.
identity
- emitted after createmodel.update
,model.update.
identity
- emitted after updatemodel.destroy
,model.destroy.
identity
- emitted after destroy
Use the extend()
method to create a Waterline model based on
BaseModel
. For example:
import {BaseModel} from 'nxus-storage'
const MyModel = BaseModel.extend({
identity: 'my-model',
attributes: {
name: 'string'
}
})
displayName
Gets a display name for the record.
Provides a consistent way to obtain a display name. The default
definition examines the record's id
, createdAt
, and
updatedAt
attributes, returning the first that has a string
value. Override the default definition to provide a more useful
display name.
Returns string display name
connection
Name of the database connection for the model.
Connections are defined in the storage configuration settings.
The default connection is default
; override to select another
configuration.
findOrCreate
Finds a record if it exists, creates it if not.
Parameters
Returns Object existing or newly created record
createOrUpdate
Creates a record if it doesn't exist, updates it if it does.
Parameters
Returns Object newly created or updated record
extend
Extends this model definition to create a new definition.
Parameters
properties
Object prototype properties for the created model; theattributes
property, if specified, augments (rather than replaces) the attributes of the model being extended
Returns Object extended Waterline model definition
GeoModel
Extends BaseModel
Base class for Waterline models containing a GeoJSON geographic attribute.
It provides methods for performing geo queries on the GeoJSON
attribute – findWithin()
selects records that lie within
specified coordinates, and findIntersects()
selects records that
intersect.
To implement the geo queries, the GeoJSON data must be indexed. And because the MongoDB 2dsphere index can handle only GeoJSON geometry features, the index is applied to a derived features attribute that contains just the geometry features from the primary GeoJSON attribute.
The GeoModel
provides the machinery for keeping the features
attribute synchronized with the primary GeoJSON attribute. It also
attempts to ensure the features attribute is well-formed and has a
consistent organization. For Polygon objects, it discards duplicate
points, closes open paths and ensures clockwise winding order. It
combines Geometry objects so there is at most one of each geometry
type: Polygon/MultiPolygon, Point/MultiPoint and
LineString/MultiLineString.
The createGeoIndex()
method should be invoked to ensure the index
is created. Typically, you do this after the startup lifecycle phase.
Configuration is through these model properties:
geometryField
(string) - Name of the primary GeoJSON attribute (default isgeo
).geometryFeatureField
(string) - Name of the geometry features attribute (default isgeoFeatures
). Both of these attributes must also be defined as model attributes with typejson
.
Use the extend()
method to create a Waterline model based on
GeoModel
. For example:
import {GeoModel} from 'nxus-storage'
const MyGeo = GeoModel.extend({
identity: 'my-geo',
attributes: {
...
location: 'json',
locationFeatures: 'json'
},
geometryField: 'location',
geometryFeatureField: 'locationFeatures'
})
createGeoIndex
Ensures index is defined for geographic attribute.
(Actually, on the attribute specified by geometryFeatureField
.)
Returns Promise A promise that resolves or rejects when index
creation completes. It's worthwhile attaching a .catch()
clause to this promise to log errors; index creation can fail
for a variety of reasons, including invalid data in the
geometry feature attribute.
findWithin
Finds records within specified geographic coordinates.
The method goes through some odd gyrations in order to mesh the MongoDB geographic query with Waterline query handling.
First, it indirectly returns a Waterline query, by returning a
promise that resolves to a function that evaluates to the query.
This due to the fact that the query has promise semantics.
Returning it indirectly defers evaluation, allowing you to refine
it (using the chainable methods such as .populate(),
.where(),
and
.sort()`) before evaluating it.
Second, it actually evaluates the MongoDB geographic query to produce a list of matching record ids, and the returned Waterline query is a second query based on these ids. Again, this to allow you to refine the query before evaluation. Be aware that this won't scale well to huge record sets, but should work for any of moderate size.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select; it must be of typePolygon
orMultiPolygon
.
Examples
```
model.findWithin({ 'type': 'Polygon', 'coordinates': ... }).then((query) => {
return query().where(...).populate(...)
}).then((records) => {
...
})
```
Returns Promise A promise that resolves to a function that returns a Waterline query for the specified coordinates.
findIntersects
Finds records intersecting specified geographic coordinates.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select.
Returns Promise A promise that resolves to a function that
returns a Waterline query for the specified coordinates. See
findWithin()
for further explanation.
getGeometry
Gets GeoJSON geometry object from the GeoJSON geographic attribute.
Typical use is to extract Polygon
geometry objects for use as
coordinates for the findWithin()
or findIntersects()
methods.
Parameters
record
ObjectGeoModel
record containing geographic attributetypes
...string geometry types to include (default is all types:Polygon
,Point
andLineString
)
Returns Object GeoJSON geometry object; undefined if no matching geometry objects were present
getCentroid
Gets the centroid of the GeoJSON geographic attribute.
Parameters
record
ObjectGeoModel
record containing geographic attribute
Returns Object GeoJSON Point object; undefined if no geometry objects were present from which to derive a centroid
PointModel
Extends BaseModel
Base collection for Waterline models containing a GeoJSON coordinate point attribute.
Meta
- deprecated: Use GeoModel.
WaterlineSessions
Extends RouterSessions
WaterlineSessions provides a nxus-router
session middleware using connect-waterline
The session model will be saved in your configured 'default' database connection.
Usage:
Application config (.nxusrc) for router:
"router": { "sessionStoreName": "waterline-session" }
findWithin
Finds records within specified geographic coordinates.
The method goes through some odd gyrations in order to mesh the MongoDB geographic query with Waterline query handling.
First, it indirectly returns a Waterline query, by returning a
promise that resolves to a function that evaluates to the query.
This due to the fact that the query has promise semantics.
Returning it indirectly defers evaluation, allowing you to refine
it (using the chainable methods such as .populate(),
.where(),
and
.sort()`) before evaluating it.
Second, it actually evaluates the MongoDB geographic query to produce a list of matching record ids, and the returned Waterline query is a second query based on these ids. Again, this to allow you to refine the query before evaluation. Be aware that this won't scale well to huge record sets, but should work for any of moderate size.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select; it must be of typePolygon
orMultiPolygon
.
Examples
```
model.findWithin({ 'type': 'Polygon', 'coordinates': ... }).then((query) => {
return query().where(...).populate(...)
}).then((records) => {
...
})
```
Returns Promise A promise that resolves to a function that returns a Waterline query for the specified coordinates.
findIntersects
Finds records intersecting specified geographic coordinates.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select.
Returns Promise A promise that resolves to a function that
returns a Waterline query for the specified coordinates. See
findWithin()
for further explanation.
Returns Object extended Waterline model definition
BaseModel
BaseModel extends Waterline.Collection to provide the following defaults and methods:
- uses the 'default' connection
- merges attributes provided by subsequent base classes to share attribute definitions
- displayName() attribute property for consistent access to an object's "name"
- emit the nxus-storage CRUD events
- findOrCreate(criteria, values) - creates the object if it does not exist
- createOrUpdate(criteria, values) - creates the object, or updates if it exists
You should almost always extend this or one of its subclasses when defining your models.
GeoModel
master
GeoModel
Extends BaseModel
Base class for Waterline models containing a GeoJSON geographic attribute.
It provides methods for performing geo queries on the GeoJSON
attribute – findWithin()
selects records that lie within
specified coordinates, and findIntersects()
selects records that
intersect.
To implement the geo queries, the GeoJSON data must be indexed. And because the MongoDB 2dsphere index can handle only GeoJSON geometry features, the index is applied to a derived features attribute that contains just the geometry features from the primary GeoJSON attribute.
The GeoModel
provides the machinery for keeping the features
attribute synchronized with the primary GeoJSON attribute. It also
attempts to ensure the features attribute is well-formed and has a
consistent organization. For Polygon objects, it discards duplicate
points, closes open paths and ensures clockwise winding order. It
combines Geometry objects so there is at most one of each geometry
type: Polygon/MultiPolygon, Point/MultiPoint and
LineString/MultiLineString.
The createGeoIndex()
method should be invoked to ensure the index
is created. Typically, you do this after the startup lifecycle phase.
Configuration is through these model properties:
geometryField
(string) - Name of the primary GeoJSON attribute (default isgeo
).geometryFeatureField
(string) - Name of the geometry features attribute (default isgeoFeatures
). Both of these attributes must also be defined as model attributes with typejson
.
Use the extend()
method to create a Waterline model based on
GeoModel
. For example:
import {GeoModel} from 'nxus-storage'
const MyGeo = GeoModel.extend({
identity: 'my-geo',
attributes: {
...
location: 'json',
locationFeatures: 'json'
},
geometryField: 'location',
geometryFeaturesField: 'locationFeatures'
})
createGeoIndex
Ensures index is defined for geographic attribute.
(Actually, on the attribute specified by geometryFeatureField
.)
Returns Promise A promise that resolves or rejects when index
creation completes. It's worthwhile attaching a .catch()
clause to this promise to log errors; index creation can fail
for a variety of reasons, including invalid data in the
geometry feature attribute.
findWithin
Finds records within specified geographic coordinates.
The method goes through some odd gyrations in order to mesh the MongoDB geographic query with Waterline query handling.
First, it indirectly returns a Waterline query, by returning a
promise that resolves to a function that evaluates to the query.
This due to the fact that the query has promise semantics.
Returning it indirectly defers evaluation, allowing you to refine
it (using the chainable methods such as .populate(),
.where(),
and
.sort()`) before evaluating it.
Second, it actually evaluates the MongoDB geographic query to produce a list of matching record ids, and the returned Waterline query is a second query based on these ids. Again, this to allow you to refine the query before evaluation. Be aware that this won't scale well to huge record sets, but should work for any of moderate size.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select; it must be of typePolygon
orMultiPolygon
.
Examples
```
model.findWithin({ 'type': 'Polygon', 'coordinates': ... }).then((query) => {
return query().where(...).populate(...)
}).then((records) => {
...
})
```
Returns Promise A promise that resolves to a function that returns a Waterline query for the specified coordinates.
findIntersects
Finds records intersecting specified geographic coordinates.
Parameters
coordinates
Object A GeoJSON geometry object specifying the geographic region to select.
Returns Promise A promise that resolves to a function that
returns a Waterline query for the specified coordinates. See
findWithin()
for further explanation.
getGeometry
Gets GeoJSON geometry object from the GeoJSON geographic attribute.
Typical use is to extract Polygon
geometry objects for use as
coordinates for the findWithin()
or findIntersects()
methods.
Parameters
record
ObjectGeoModel
record containing geographic attributetypes
...string geometry types to include (default is all types:Polygon
,Point
andLineString
)
Returns Object GeoJSON geometry object; undefined if no matching geometry objects were present
getCentroid
Gets the centroid of the GeoJSON geographic attribute.
Parameters
record
ObjectGeoModel
record containing geographic attribute
Returns Object GeoJSON Point object; undefined if no geometry objects were present from which to derive a centroid
PointModel
Extends BaseModel
Base collection for Waterline models containing a GeoJSON coordinate point attribute.
Meta
- deprecated: Use GeoModel.