Bug for semantic splitting

[djs0109]

seems like for a entity that does not have a specific attribute co2 to be split, like following:

    {
        "id": "HotelRoom:fiware_room_presence_1_v2",
        "type": "HotelRoom",
        "hasLocation": {
            "type": "Relationship",
            "value": "fiware_hotel",
            "metadata": {}
        },
        "name": {
            "type": "Text",
            "value": "fiware_room_presence_1_v2",
            "metadata": {}
        },
        "temperature": {
            "type": "Number",
            "value": 0.0,
            "metadata": {}
        }
    },

The current way of writing RML (as below) will still cause the creation of the CO2 node

ex:Mappingco2_HotelRoom
    a rr:TriplesMap ;
    rml:logicalSource [
        rml:source "placeholder.json" ;
        rml:referenceFormulation ql:JSONPath ;
        rml:iterator '$[?(@.type=="HotelRoom")]' ;
    ] ;
    rr:subjectMap [
        rr:template "http://example.com/co2_HotelRoom/{id}" ;
        rr:class brick:CO2_Sensor ;
    ] ;
    rr:predicateObjectMap [
        rr:predicate brick:isPointOf ;
        rr:objectMap [
            rr:parentTriplesMap ex:MappingHotelRoom ;
            rr:joinCondition [
                rr:child "id" ;
                rr:parent "id" ;
            ] ;
        ] ;
    ] ;
    rr:predicateObjectMap [
        rr:predicate rdf:value ;
        rr:objectMap [
            rr:template "http://137.226.248.250:1026/v2/entities/{id}/attrs/co2/value" ;
        ] ;
    ] ;
    .

Need to figure out a way to avoid such dummy nodes