lib

README

Core lib for microservices

We provide some helpers to simplify and unify microservice structure, basic units and types

Stream

We define a basic microservice message format.

Our brokers can enhance this message but must retain the core structure for pipeline processing.

Specific functions are provided for simplifying use of google's pubsub.

Google Bigquery

Add time series queries for google big query (this is temporarily PowerUsage based but should allow generalization eventually)

Device

The framework for processing device data. Device provides a pipeline with various fields for storing and processing the data.

Typically this is taken from a store/cache and attached to the interpreted device payload. Pipeline elements can then fork, process, store based around these e.g

• a microservice that stores all device information for a location
• a microservice that sums a field over all messages in a system

etc

Documentation

Index

• Constants
• type Category
• type Device
• type DeviceGroup
• type DeviceMeta
• type DeviceSkeleton
• type DeviceType
• type Firmware
• type JWTParser
  • func NewJWTParser(secret string) JWTParser
  • func (p *JWTParser) CreateToken(claims jwt.MapClaims) (string, error)
  • func (p *JWTParser) ParseToken(tokenString string) (*jwt.MapClaims, error)
• type Processor
• type Unit

Constants

const (
	Power    DeviceType  = "power"
	Lighting DeviceType  = "lighting"
	Meter    DeviceGroup = "meter"
	DaliEL   DeviceGroup = "daliEL"
)

Types

type Category

type Category struct {
	DeviceType
	DeviceGroup
}

type Device

type Device struct {
	DeviceUID   string `firestore:"uid" json:"uid"`
	*DeviceMeta `datastore:",omitempty" firestore:",omitempty" json:",omitempty"`
}

Device minimally identifies an IoT resource a devices message typically identifies itself with a UID the Meta information can then be accessed and added to the struct as well as Groups, an open array of structural helpers

type DeviceGroup

type DeviceGroup string

type DeviceMeta

type DeviceMeta struct {
	DeviceName string     `datastore:",omitempty" firestore:"name,omitempty" json:"deviceName,omitempty"`
	DeviceTag  string     `datastore:",omitempty" firestore:"tag,omitempty" json:"deviceTag,omitempty"`
	DeviceType DeviceType `datastore:",omitempty" firestore:"type,omitempty" json:"type,omitempty"`

	CompanyUID  string `datastore:",omitempty" firestore:"companyUID,omitempty" json:"companyUID,omitempty"`
	LocationUID string `datastore:",omitempty" firestore:"locationUID,omitempty" json:"locationUID,omitempty"`

	Firmware *Firmware `datastore:",omitempty" firestore:"firmware,omitempty" json:"firmware,omitempty"`
	//Datastore can't handle map[string]interface so a datastore implementation will need to serialize the specific processors it needs
	Processors *Processor `datastore:"-" firestore:"processors,omitempty" json:"processors,omitempty"`
}

DeviceMeta helps further identify a Device. Deployed devices *must* have an UID that identifies the physical unit.

The metadata can be added to passed messages so a cache is not needed to query the structure needed for microservice

DeviceName is a human-readable identifier for the device which may or may not be unique DeviceTag identifies the function of the device in its environment and should remain constant if the device is switched out due to failure or replacement but the function of the device remains constant

so for example:

the cooler in room 4B is still the cooler in 4B even if it is upgraded from PhilipsKA4500 to PhilipsKA4501

CompanyUID and LocationUID remain as basic organizational elements

type DeviceSkeleton

type DeviceSkeleton struct {
	DeviceUID      string
	ServiceUID     string
	DeviceCategory Category
}

DeviceSkeleton is used by microservices before they have full admin details for a device - this allows a device's data to be stored and traced even if it is not currently identified in admin.

type DeviceType

type DeviceType string

type Firmware

type Firmware struct {
	FirmwareName    string `datastore:",omitempty" firestore:"name,omitempty" json:"name,omitempty"`
	FirmwareVersion string `datastore:",omitempty" firestore:"version,omitempty" json:"version,omitempty"`
}

type JWTParser

type JWTParser struct {
	// contains filtered or unexported fields
}

JWTParser: We use JWT primarily in conjunction with external Webhooks

func NewJWTParser

func NewJWTParser(secret string) JWTParser

func (*JWTParser) CreateToken

func (p *JWTParser) CreateToken(claims jwt.MapClaims) (string, error)

func (*JWTParser) ParseToken

func (p *JWTParser) ParseToken(tokenString string) (*jwt.MapClaims, error)

type Processor

type Processor map[string]interface{}

Processor allows implementations to save a number of identifiers for different processing options

type Unit

type Unit string

const (
	KWH Unit = "kWh"
)