Events handler

Handler is a constructor for the user input images processing rules. Handler has five processing steps:

1. Detect faces. User can set a list attributes for estimation and head angle thresholds.

2. Extract attributes. User can define which attributes should be extracted (basic attributes, face or body descriptors or both).

3. Match extracted descriptors by lists. User can define luna lists for matching and rules for matching.

4. Create objects using the extracted data and store them. User can define rules for creating samples, attributes, faces or even events. Also, see link policy inside the face storage policy - linking created faces to lists.

For detail see spec_create_new_handler.

Module contains schemas for handler

class luna_handlers.classes.schemas.handler.CachedHandlerModel(**data)

Cached handler model | contains minimal required fields

classmethod postInit(values)

Post init

static postInitMatchPolicy(matchPolicies, accountId)

Post init matching policy - apply handler account id for matching candidates :param matchPolicies: matching policies :param accountId: handler account id

Return type:

None

class luna_handlers.classes.schemas.handler.CreateHandlerModel(**data)

Handler model for handler creation

classmethod validateNPostInit(values)

Validate handler, post init

class luna_handlers.classes.schemas.handler.FaceInputEstimationsModel(**data)

Model for incoming face estimations: urls, image or samples

classmethod validateHandler(values)

Validate handler model

Return type:

dict

class luna_handlers.classes.schemas.handler.HandlerInputEstimationsModel(*, image: ~typing.Optional[str] = None, mimetype: ~typing.Optional[str] = None, face_bounding_boxes: ~types.ConstrainedListValue[~classes.schemas.simple_schemas.BoundingBoxSchema] = None, body_bounding_boxes: ~types.ConstrainedListValue[~classes.schemas.simple_schemas.BoundingBoxSchema] = None, urls: ~classes.schemas.types.userConstrainedList.<locals>.UserConstrainedList[~luna_handlers.classes.schemas.handler.UrlForInputEvent] = None, samples: ~typing.Union[~classes.schemas.types.userConstrainedList.<locals>.UserConstrainedList[~luna_handlers.classes.schemas.handler.SampleForInputEvent], ~classes.schemas.types.userConstrainedList.<locals>.UserConstrainedList[~uuid.UUID]] = None, detect_time: ~classes.schemas.types.CustomDatetime = None, detect_ts: ~typing.Optional[~classes.schemas.types.DetectTs] = None, image_origin: ~classes.schemas.types.Str256 = None)

Handler model for incoming estimations: urls, image(with face and/or body bounding boxes) or samples

classmethod validateHandler(values)

Validate handler model

Return type:

dict

class luna_handlers.classes.schemas.handler.SampleForInputEvent(**data)

Sample schema

class luna_handlers.classes.schemas.handler.UrlForInputEvent(**data)

Url schema with detection time and image origin

class luna_handlers.classes.schemas.handler.ValidatorHandlerModel(**data)

Handler validator model

Module contains event structure.

class luna_handlers.classes.event.BodyExtractedAttribute(descriptor=None, sampleIds=None)

Body extracted attribute.

asDict()

Body attributes in api format.

class luna_handlers.classes.event.Event(sdkEstimations=None, aggregatedEstimations=None, faceAttributes=None, bodyAttributes=None, sampleIds=None, meta=None)

Structure for storage an event

eventId

event id

Type:

str

eventUrl

(Optional[str]) event url

sdkEstimations

face & body estimations

Type:

list[ImageDetection]

aggregateEstimations

event aggregate estimations

Type:

AggregateEstimations

faceAttributes

face extracted attribute

Type:

FaceExtractedAttribute

bodyAttributes

body extracted attribute

Type:

BodyExtractedAttribute

sampleIds

event attribute sample ids

Type:

list[str]

meta

user defined metadata

Type:

EventMetadata

tags

event tags (user defined & conditional ones)

Type:

list[str]

avatar

event avatar

faceId

face id associated with the event

Type:

Optional[str]

faceUrl

face url

Type:

Optional[str]

linkedLists

list that contains the face

Type:

Optional[List[str]]

matches

list of match results

Type:

List[EventMatchResult]

asDict()

Prepares dict representation.

Return type:

dict

class luna_handlers.classes.event.EventMetadata(accountId, handlerId, createEventTime, endEventTime, source=None, tags=None, userData=None, externalId=None, location=None, trackId=None)

User defined event metadata.

class luna_handlers.classes.event.ExtractedAttribute(descriptor=None, sampleIds=None)

Attribute.

attributeId

attribute id

Type:

Optional[str]

descriptor

descriptor

Type:

RawDescriptorData

sampleIds

sample ids

Type:

list(str)

class luna_handlers.classes.event.FaceExtractedAttribute(descriptor=None, basicAttributes=None, sampleIds=None)

Face extracted attribute.

asDict()

Face attributes in api format.

class luna_handlers.classes.event.GeoPosition(latitude=None, longitude=None)

Event coordinates

latitude

longitude coordinate

Type:

Optional[float]

longitude

longitude coordinate

Type:

Optional[float]

asDict()

Convert coordinates to dict.

Returns:

dict. Keys with non None value.

Return type:

Dict[str, Optional[float]]

copy()

Deep copy coordinates.

Returns:

event coordinates

Return type:

GeoPosition

isNotEmpty()

Check geo position is not empty :returns: true if not empty else false

Return type:

bool

isValid()

Check that both of coordinates are specified :returns: true if both of coordinates are set or no one otherwise false

Return type:

bool

class luna_handlers.classes.event.Location(street, houseNumber, district, area, city, geoPosition)

Event location

street

street

Type:

Optional[str]

houseNumber

house number

Type:

Optional[str]

district

district

Type:

Optional[str]

area

area

Type:

Optional[str]

city

city

Type:

Optional[str]

geoPosition

geo coordinates

Type:

luna_handlers.classes.event.GeoPosition

asDict()

Convert location to dict.

Returns:

dict. Keys with non None value.

Return type:

Dict[str, str]

copy()

Deep copy location.

Returns:

event location

Return type:

Location