LIG CRPGL queue interface

Queue element structure

Shot and head active learning annotation scenarios should assume elements in queues are structured as follows:

Input element for the shot use case:

{

	"fragment": {
		"start": NUM,
		"end": NUM,
		"context": {
			"id_layer": STRING
		}
	},

	"data": [
		STRING1,
		...
	],

	"_id": STRING,
	"id_corpus": STRING,
	"id_media": STRING,
	"id_layer": STRING,
	
	
}

Output element for the shot use case:

{

	"fragment": {
		"start": NUM,
		"end": NUM,
		"context": {
			"id_layer": STRING
		}
	},

	"data": {
		"status": STRING in [VALIDATED | SKIP | MODIFIED],
		"inData": {
			"date": DATE,
			"data": [
				STRING1,
				...
			]
		},
		"outData": {
			"date": DATE,
			"data"[
				STRING1,
				...
			],
			"user": STRING,
			"duration": NUM
		},
	
	},
	"_id": STRING,
	"id_corpus": STRING,
	"id_media": STRING,
	"id_layer": STRING,
	
	
}

Input element for the head use case:

{
	
	"fragment": {
		"start": NUM,
		"end": NUM (same value as start),
		"position": {
			"top": NUM in [0,1],
			"left": NUM in [0,1],
			"width": NUM in [0,1],
			"height": NUM in [0,1]
		},
		"context": {
			"id_layer": STRING
		}
	},
	"data": STRING,

	"_id": STRING,
	"id_corpus": STRING,
	"id_media": STRING,
	"id_layer": STRING
	
}

Output element for the head use case:

{
	
	"fragment": {
		"start": NUM,
		"end": NUM (same value as start),
		"position": {
			"top": NUM in [0,1],
			"left": NUM in [0,1],
			"width": NUM in [0,1],
			"height": NUM in [0,1]
		},
		"context": {
			"id_layer": STRING
		}
	},
	"data": {
		"status": STRING in [VALIDATED | SKIP | MODIFIED],
		"inData": {
			"date": DATE,
			"data": STRING
		},

		"outData": {
			"date": DATE,
			"data"[
				STRING1,
				...
			],
			"user": STRING,
			"duration": NUM
		},	

	}	


	"_id": STRING,
	"id_corpus": STRING,
	"id_media": STRING,
	"id_layer": STRING
	
}

Names of the queues

The queues used should have the following names:

"shotIncoming", "shotOutcoming", "headIncoming", "headOutcoming".