Conditionally creating triples

[rybesh]

I have a CSV file that looks like this:

candy_id,candy_name,has_peanuts,has_caramel,has_nougat
1,Snickers,1,1,1
2,Three Musketeers,0,0,1

I want to produce triples like this:

ex:1 rdfs:label "Snickers" ; ex:hasIngredient ex:peanuts , ex:caramel , ex:nougat .
ex:2 rdfs:label "Three Musketeers" ; ex:hasIngredient ex:nougat .

Is this possible using e.g. RMLMapper?

[DylanVanAssche]

Hi!

You can achieve this in RMLMapper with the help of some FnO functions.
RMLMapper has several default functions included, an overview is available here

With the function if you can do a check if the column has 1 and output a constant like ex:peanuts:

# Output
dylan@work-laptop:~/Downloads$ java -jar rmlmapper-7.0.0-r374-all.jar -m mapping.ttl 
<http://example.org/1> <http://example.org/hasIngredient> <http://example.org/caramel> .
<http://example.org/1> <http://example.org/hasIngredient> <http://example.org/nougat> .
<http://example.org/1> <http://example.org/hasIngredient> <http://example.org/peanuts> .
<http://example.org/1> <http://www.w3.org/2000/01/rdf-schema#label> "Snickers" .
<http://example.org/2> <http://example.org/hasIngredient> <http://example.org/nougat> .
<http://example.org/2> <http://www.w3.org/2000/01/rdf-schema#label> "Three Musketeers" .

# Mapping
dylan@work-laptop:~/Downloads$ cat mapping.ttl 
@prefix rr: <http://www.w3.org/ns/r2rml#> .
@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .
@prefix rml: <http://semweb.mmlab.be/ns/rml#> .
@prefix ql: <http://semweb.mmlab.be/ns/ql#> .
@prefix fnml:   <http://semweb.mmlab.be/ns/fnml#> .
@prefix fno:    <https://w3id.org/function/ontology#> .
@prefix grel: <http://users.ugent.be/~bjdmeest/function/grel.ttl#> .
@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .
@prefix ex: <http://example.org/> .
@base <http://example.com/base/> .

<#TriplesMap> a rr:TriplesMap;
  rml:logicalSource [
    rml:source "example.csv";
    rml:referenceForumation ql:CSV;
  ];

  rr:subjectMap [
    rr:template "http://example.org/{candy_id}";
  ];

  rr:predicateObjectMap [
    rr:predicate rdfs:label;
    rr:objectMap [ rml:reference "candy_name" ];
  ];

  rr:predicateObjectMap [
    rr:predicate ex:hasIngredient;
    rr:objectMap [
      fnml:functionValue [
      	rr:predicateObjectMap [
      		rr:predicate fno:executes ;
      		rr:objectMap [ rr:constant grel:controls_if ] ] ;
      	rr:predicateObjectMap [
      		rr:predicate grel:bool_b ;
      		rr:objectMap [ rml:reference "has_peanuts"; rr:datatype xsd:boolean ] ] ;
      	rr:predicateObjectMap [
            rr:predicate grel:any_true ;
            rr:objectMap [ rr:constant ex:peanuts; ] ]
      ];
      rr:termType rr:IRI;
    ];
   ];

  rr:predicateObjectMap [
    rr:predicate ex:hasIngredient;
    rr:objectMap [
      fnml:functionValue [
      	rr:predicateObjectMap [
      		rr:predicate fno:executes ;
      		rr:objectMap [ rr:constant grel:controls_if ] ] ;
      	rr:predicateObjectMap [
      		rr:predicate grel:bool_b ;
      		rr:objectMap [ rml:reference "has_caramel"; rr:datatype xsd:boolean ] ] ;
      	rr:predicateObjectMap [
            rr:predicate grel:any_true ;
            rr:objectMap [ rr:constant ex:caramel; ] ]
      ];
      rr:termType rr:IRI;
    ];
   ];

  rr:predicateObjectMap [
    rr:predicate ex:hasIngredient;
    rr:objectMap [
      fnml:functionValue [
      	rr:predicateObjectMap [
      		rr:predicate fno:executes ;
      		rr:objectMap [ rr:constant grel:controls_if ] ] ;
      	rr:predicateObjectMap [
      		rr:predicate grel:bool_b ;
      		rr:objectMap [ rml:reference "has_nougat"; rr:datatype xsd:boolean ] ] ;
      	rr:predicateObjectMap [
            rr:predicate grel:any_true ;
            rr:objectMap [ rr:constant ex:nougat; ] ]
      ];
      rr:termType rr:IRI;
    ];
   ];
.

[rybesh]

Thank you for the clear and detailed answer!

Suppose there are several dozen ingredient binaries has_X. Is there any way to handle all of them without repeating the entire predicateObjectMap for each binary, or would I need to somehow generate the triples map programmatically to achieve that?

[DylanVanAssche]

You need to repeat the Predicate Object Map for each has_X, if you want to avoid doing it manually, you need indeed some script to generate the mapping. Luckily, you only need to do it once.