apf:strSplit if used first, causes CSV rows to be skipped despite OPTIONAL

Query test1.fx works ok. You may notice that each CSV field is fetched with optional {?ROW then processed in the same OPTIONAL block:

base            <https://graph.swissgrid.ch/>
prefix apf: <http://jena.apache.org/ARQ/property#>
prefix fx: <http://sparql.xyz/facade-x/ns/>
prefix xyz: <http://sparql.xyz/facade-x/data/>
construct {
  graph ?System_URL                {?IRI_URL meta:systemReferenz ?Referenz}
  graph ?MastersystemPrimary_URL   {?IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/primaer>}
  graph ?MastersystemSecondary_URL {?IRI_URL meta:systemPraeferenz <enum/enum/systemPraeferenz/sekundaer>}
  graph ?MastersystemFuture_URL    {?IRI_URL meta:systemPraeferenz <enum/enum/systemPraeferenz/zukuenftigePrimaer>}
} where {
  service <x-sparql-anything:> {
    fx:properties
      fx:location $_location ;
      fx:csv.headers "true" ;
      fx:csv.headers.sanitize "true" ;
      fx:csv.null-string "" ;
    .
    optional {?ROW xyz:IRI|xyz:Attribut_IRI ?IRI
      bind(iri(concat("otl/",?IRI)) as ?IRI_URL)
    }
    optional {?ROW xyz:Objekt_System_Referenz|xyz:Attribut_System_Referenz ?SystemReferenzen.
      ?SystemReferenz apf:strSplit (?SystemReferenzen ";")
      bind(iri(concat("profile/",replace(?SystemReferenz,"(.+):.*","$1"),"/graph")) as ?System_URL)
      bind(if(regex(?SystemReferenz,".+:(.+)"),replace(?SystemReferenz,".+:(.+)","$1"),?UNDEF) as ?Referenz)
    }
    optional {?ROW xyz:IRI_Mastersystem ?Mastersystem
      bind(iri(concat("profile/",if(?Mastersystem="engineeringbaseSap","engineeringbase",?Mastersystem),"/graph")) as ?MastersystemPrimary_URL)
      bind(iri(concat("profile/",if(?Mastersystem="engineeringbaseSap","sappm",          ?UNDEF),       "/graph")) as ?MastersystemSecondary_URL)
    }
    optional {?ROW xyz:IRI_Zukuenftiges_Mastersystem ?MastersystemFuture
      bind(iri(concat("profile/",?MastersystemFuture,"/graph")) as ?MastersystemFuture_URL)
    }
  }
}

Query test2.fx is very similar in structure, but it first fetches all fields (each in optional {?ROW) then processes them.
Notice that apf:strSplit is applied in OPTIONAL block after the argument ?Attribut_System_Referenz is fetched, not in the same block:

base            <https://graph.swissgrid.ch/>
prefix apf: <http://jena.apache.org/ARQ/property#>
prefix fx: <http://sparql.xyz/facade-x/ns/>
prefix xyz: <http://sparql.xyz/facade-x/data/>
construct {
  graph ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemReferenz ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID}
  graph ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/primaer>}
  graph ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/sekundaer>}
  graph ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/zukuenftigePrimaer>}
}
where {
  service <x-sparql-anything:> {
    fx:properties
      fx:location $_location ;
      fx:csv.headers "true" ;
      fx:csv.headers.sanitize "true" ;
      fx:csv.null-string "" ;
    .
    optional {?ROW xyz:Attribut_System_Referenz ?Attribut_System_Referenz}
    optional {?ROW xyz:Attribut_IRI ?Attribut_IRI}
    optional {?ROW xyz:IRI_Mastersystem ?IRI_Mastersystem}
    optional {?ROW xyz:IRI_Zukuenftiges_Mastersystem ?IRI_Zukuenftiges_Mastersystem}
    optional {?Attribut_System_Referenz_SPLIT_SEMI apf:strSplit (?Attribut_System_Referenz ";")}
    bind(replace(?Attribut_System_Referenz_SPLIT_SEMI,"(.+):.*","$1") as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS)
    bind(iri(concat("profile/",?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS,"/graph")) as ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL)
    bind(if(regex(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)"),replace(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)","$1"),?UNDEF) as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID)
    bind(iri(concat("otl/",?Attribut_IRI)) as ?otl_Attribut_IRI_URL)
    bind(if(?IRI_Mastersystem="engineeringbaseSap","engineeringbase",?IRI_Mastersystem) as ?IRI_Mastersystem_SYS_PRIMARY)
    bind(iri(concat("profile/",?IRI_Mastersystem_SYS_PRIMARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL)
    bind(if(?IRI_Mastersystem="engineeringbaseSap","sappm",?UNDEF) as ?IRI_Mastersystem_SYS_SECONDARY)
    bind(iri(concat("profile/",?IRI_Mastersystem_SYS_SECONDARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL)
    bind(iri(concat("profile/",?IRI_Zukuenftiges_Mastersystem,"/graph")) as ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL)
  }
}

Something strange happens with test2.fx:
it outputs quads only for rows that have xyz:Attribut_System_Referenz.

These CSV are progressively increasing subsets: test1.csv, test13.csv

Try the two scripts on these examples to see the problem. Eg

sparql-anything -q test1.fx -v location=test13.csv -f trig|grep -v PREFIX > out1.trig
sparql-anything -q test2.fx -v location=test13.csv -f trig|grep -v PREFIX > out2.trig

Update: the version is 1.2.0

I have reworked test2.fx to have the same structure as test1.fx: test3.fx:

base         <https://graph.swissgrid.ch/>
prefix meta: <meta/>
prefix apf:  <http://jena.apache.org/ARQ/property#>
prefix fx:   <http://sparql.xyz/facade-x/ns/>
prefix xyz:  <http://sparql.xyz/facade-x/data/>
construct {
  graph ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemReferenz ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID}
  graph ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/primaer>}
  graph ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/sekundaer>}
  graph ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/zukuenftigePrimaer>}
}
where {
  service <x-sparql-anything:> {
    fx:properties
      fx:location $_location ;
      fx:csv.headers "true" ;
      fx:csv.headers.sanitize "true" ;
      fx:csv.null-string "" ;
    .
    optional {?ROW xyz:Attribut_System_Referenz ?Attribut_System_Referenz.
      ?Attribut_System_Referenz_SPLIT_SEMI apf:strSplit (?Attribut_System_Referenz ";")
      bind(replace(?Attribut_System_Referenz_SPLIT_SEMI,"(.+):.*","$1") as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS)
      bind(iri(concat("profile/",?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS,"/graph")) as ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL)
      bind(if(regex(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)"),replace(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)","$1"),?UNDEF) as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID)
    }
    optional {?ROW xyz:Attribut_IRI ?Attribut_IRI
      bind(iri(concat("otl/",?Attribut_IRI)) as ?otl_Attribut_IRI_URL)
    }
    optional {?ROW xyz:IRI_Mastersystem ?IRI_Mastersystem
      bind(if(?IRI_Mastersystem="engineeringbaseSap","engineeringbase",?IRI_Mastersystem) as ?IRI_Mastersystem_SYS_PRIMARY)
      bind(iri(concat("profile/",?IRI_Mastersystem_SYS_PRIMARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL)
      bind(if(?IRI_Mastersystem="engineeringbaseSap","sappm",?UNDEF) as ?IRI_Mastersystem_SYS_SECONDARY)
      bind(iri(concat("profile/",?IRI_Mastersystem_SYS_SECONDARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL)
    }
    optional {?ROW xyz:IRI_Zukuenftiges_Mastersystem ?IRI_Zukuenftiges_Mastersystem
      bind(iri(concat("profile/",?IRI_Zukuenftiges_Mastersystem,"/graph")) as ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL)
    }
  }
}

But it still doesn't output the quads from a row that lacks xyz:Attribut_System_Referenz:

<https://graph.swissgrid.ch/profile/plscadd/graph> {
    <https://graph.swissgrid.ch/otl/leiterquerschnittKern>
            meta:systemPraeferenz  <https://graph.swissgrid.ch/enum/systemPraeferenz/zukuenftigePrimaer> , <https://graph.swissgrid.ch/enum/systemPraeferenz/primaer> .
}

If I swap the order of the first 2 OPTIONAL blocks, it works ok.
But they are absolutely independent of each other ?!?

test4.fx does the apf:strSplit last, and it works ok:

base         <https://graph.swissgrid.ch/>
prefix meta: <meta/>
prefix apf:  <http://jena.apache.org/ARQ/property#>
prefix fx:   <http://sparql.xyz/facade-x/ns/>
prefix xyz:  <http://sparql.xyz/facade-x/data/>
construct {
  graph ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemReferenz ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID}
  graph ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/primaer>}
  graph ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/sekundaer>}
  graph ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL {
    ?otl_Attribut_IRI_URL meta:systemPraeferenz <enum/systemPraeferenz/zukuenftigePrimaer>}
}
where {
  service <x-sparql-anything:> {
    fx:properties
      fx:location $_location ;
      fx:csv.headers "true" ;
      fx:csv.headers.sanitize "true" ;
      fx:csv.null-string "" ;
    .
    optional {?ROW xyz:Attribut_IRI ?Attribut_IRI}
    optional {?ROW xyz:IRI_Mastersystem ?IRI_Mastersystem}
    optional {?ROW xyz:IRI_Zukuenftiges_Mastersystem ?IRI_Zukuenftiges_Mastersystem}
    optional {?ROW xyz:Attribut_System_Referenz ?Attribut_System_Referenz}
    optional {?Attribut_System_Referenz_SPLIT_SEMI apf:strSplit (?Attribut_System_Referenz ";")}
    bind(replace(?Attribut_System_Referenz_SPLIT_SEMI,"(.+):.*","$1") as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS)
    bind(iri(concat("profile/",?Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS,"/graph")) as ?profile_Attribut_System_Referenz_SPLIT_SEMI_SYSREFSYS_graph_URL)
    bind(if(regex(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)"),replace(?Attribut_System_Referenz_SPLIT_SEMI,".+:(.+)","$1"),?UNDEF) as ?Attribut_System_Referenz_SPLIT_SEMI_SYSREFID)
    bind(iri(concat("otl/",?Attribut_IRI)) as ?otl_Attribut_IRI_URL)
    bind(if(?IRI_Mastersystem="engineeringbaseSap","engineeringbase",?IRI_Mastersystem) as ?IRI_Mastersystem_SYS_PRIMARY)
    bind(iri(concat("profile/",?IRI_Mastersystem_SYS_PRIMARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_PRIMARY_graph_URL)
    bind(if(?IRI_Mastersystem="engineeringbaseSap","sappm",?UNDEF) as ?IRI_Mastersystem_SYS_SECONDARY)
    bind(iri(concat("profile/",?IRI_Mastersystem_SYS_SECONDARY,"/graph")) as ?profile_IRI_Mastersystem_SYS_SECONDARY_graph_URL)
    bind(iri(concat("profile/",?IRI_Zukuenftiges_Mastersystem,"/graph")) as ?profile_IRI_Zukuenftiges_Mastersystem_graph_URL)
  }
}

Navigation Menu

Repository navigation

apf:strSplit if used first, causes CSV rows to be skipped despite OPTIONAL #654

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

apf:strSplit if used first, causes CSV rows to be skipped despite OPTIONAL #654

Description

Activity

VladimirAlexiev commented on Aug 31, 2026

VladimirAlexiev commented on Aug 31, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions