SPARQL Updates

A SPARQL update is used to edit RDF data. There are several types of updates, which follow under two categories:

• Graph Update
  • :insert-data
  • :delete-data
  • :delete/:insert
  • :delete-where
  • :load
  • :clear
• Graph Management
  • :create
  • :drop
  • :copy
  • :move
  • :add

Graph Update updates change existing RDF graphs without adding or deleting them, while Graph Management updates are free to add, delete, or move graphs.

All Graph Management update clauses, as well as :load and :clear, have -silent versions, e.g. :load-silent, which will return success in all cases and only silently fail.

Each SPARQL update in Flint is a map that includes one of the aforementioned clauses (or two in the case of having both :delete and :insert clauses), as well as one or more of the following clauses:

• Prologue clauses
  • :base
  • :prefixes
• :delete/:insert-specific clauses:
  • :where (required)
  • :with
  • :using
• :load-specific clauses:
  • :into
• :add, :move, and :copy-specific clauses:
  • :to

The triple insertion and deletion clauses accept both triples and quads, which have the form [:graph iri triples]. This is similar to the :graph clause in graph patterns, except that variables cannot be substituted in the graph IRI position.

NOTE: Any key other than the above keywords is not allowed in a SPARQL update map.

Update clauses

:insert-data

Reference: 3.1.1 INSERT DATA

The :insert-data clause inserts triples in an RDF graph. Syntactically, it consists of triples or quads.

The example:

{:prefixes    {:foaf "<http://xmlns.com/foaf/0.1/>"}
 :insert-data [[:graph "<http://census.marley/data>"
                       [["<http://census.marley/entry#211402>" :foaf/givenName "Reiner"]
                        ["<http://census.marley/entry/211402>" :foaf/familyName "Braun"]]]]}

becomes:

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
INSERT DATA {
    GRAPH <http://census.marley/data> {
        <http://census.marley/entry#211402> foaf:givenName "Reiner" .
        <http://census.marley/entry#211402> foaf:familyName "Braun" .
    }
}

NOTE: Property paths and variables are not allowed in an :insert-data clause.

:delete-data

Reference: 3.1.2 DELETE DATA

The :delete-data clause deletes triples in an RDF graph. Syntactically, it consists of triples or quads.

The example:

{:prefixes    {:foaf "<http://xmlns.com/foaf/0.1/>"}
 :delete-data [[:graph "<http://census.marley/data>"
                       [["<http://census.marley/entry#211397>" :foaf/givenName "Bertolt"]
                        ["<http://census.marley/entry#211397>" :foaf/familyName "Hoover"]]]]}

becomes:

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
DELETE DATA {
    GRAPH <http://census.marley/data> {
        <http://census.marley/entry#211397> foaf:givenName "Bertlot" .
        <http://census.marley/entry#211397> foaf:familyName "Hoover" .
    }
}

NOTE: Property paths, variables, and blank nodes are not allowed in a :delete-data clause.

:delete and :insert

Reference: 3.1.3 DELETE/INSERT

A :delete or :insert clause deletes or inserts triples, respectively, in an RDF graph with variables specified by a :where clause. Syntactically, both clauses consists of triples or quads. An update may contain either a :delete clause, an :insert clause, or both.

The example:

{:prefixes {:foaf "<http://xmlns.com/foaf/0.1/>"}
 :delete   [[:graph "<http://census.marley/data>"
                    [[?x :foaf/familyName "Brown"]]]]
 :insert   [[:graph "<http://census.marley/data>"
                    [[?x :foaf/familyName "Braun"]]]]
 :where    [[:graph "<http://census.marley/data>"
                    [[?x :foaf/familyName "Brown"]]]]}

becomes:

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
DELETE {
    GRAPH <http://census.marley/data> {
        ?x foaf:familyName "Brown" .
    }
}
INSERT {
    GRAPH <http://census.marley/data> {
        ?x foaf:familyName "Braun" .
    }
}
WHERE {
    GRAPH <http://census.marley/data> {
        ?x foaf:familyName "Brown" .
    }
}

NOTE: Blank nodes are not allowed in the :delete clause.

For information about :using and :with clauses, see Graph IRIs.

:delete-where

Reference: 3.1.3.3 DELETE WHERE

The :delete-where clause is a shorthand for the combination :delete and :where, without :insert in between. Syntactically, the :delete-where clause consists of triples or quads.

The example:

{:prefixes     {:foaf "<http://xmlns.com/foaf/0.1/>"}
 :delete-where [[:graph "<http://census.marley/data>"
                       [[?x :foaf/givenName "Annie"]
                        [?x :foaf/familyName "Leonhart"]]]]}

becomes

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
DELETE WHERE {
    GRAPH <http://census.marley/data> {
        ?x foaf:givenName "Annie" .
        ?x foaf:familyName "Leonhart" .
    }
}

NOTE: Blank nodes are not allowed in the :delete-where clause.

:load/:load-silent and :into

Reference: 3.1.4 LOAD

The :load update loads triples from a source specified by an IRI. The :load clause is followed by an optional :into clause (if the latter is omitted the data is loaded into the default graph). Syntactically, both the :load and :into clauses consist of an IRI or prefixed IRI.

The example:

{:load "<file:marleycensus/data.rdf>"
 :into "<http://census.marley/data>"}

becomes:

LOAD <file:marleycensus/data.rdf>
INTO <http://census.marley/data>

Example of silent mode:

{:load-silent "<file:marleycensus/data.rdf>"
 :into        "<http://census.marley/data>"}

becomes:

LOAD SILENT <file:marleycensus/data.rdf>
INTO <http://census.marley/data>

:clear/:clear-silent

Reference: 3.1.5 CLEAR

The :clear update is used to clear all data from one or more RDF graphs. Syntactically, the :clear clause consists of either an explicit IRI or prefixed IRI, :default, :named, or :all.

The example:

{:clear :default}

becomes:

CLEAR DEFAULT

:create/:create-silent

Reference: 3.2.1 CREATE

The :create update creates a new empty graph in an RDF store. The :create clause consists of the IRI or prefixed of the new graph.

The example:

{:prefixes {:census "<http://census.marley/>"
 :create   :census/data}}

becomes:

PREFIX census: <http://census.marley/>
CREATE census:data

:drop/:drop-silent

Reference: 3.2.2 DROP

The :drop update deletes one or more graphs in an RDF store. The :drop clause consists of either an explicit IRI or prefixed IRI, :default, :named, or :all.

The example:

{:drop :default}

becomes:

DROP DEFAULT

:copy/:copy-silent and :to

Reference: 3.2.3 COPY

The :copy update transfers data from a source to a target graph, overwriting the latter in the process. Both the :copy and :to clauses consist of either an IRI or prefixed IRI or the :default keyword.

The example:

{:copy "<http://census.marley/data>"
 :to   :default}

becomes:

COPY <http://census.marley/data>
TO DEFAULT

:move/:move-silent and :to

Reference: 3.2.4 MOVE

The :move update moves data from a source to a target graph, deleting the former and overwriting the latter. Both the :move and :to clauses consist of either an IRI or prefixed IRI or the :default keyword.

The example:

{:move "<http://census.marley/data>"
 :to   :default}

becomes:

MOVE <http://census.marley/data>
TO DEFAULT

:add/:add-silent and :to

Reference: 3.2.5 ADD

The :add update appends data from a source to a target graph. Both the :add and :to clauses consist of either an IRI or prefixed IRI or the :default keyword.

The example:

{:add "<http://census.marley/data>"
 :to  :default}

becomes:

ADD <http://census.marley/data>
TO DEFAULT

Update request sequences

Unlike SPARQL queries, SPARQL update requests can be chained together into sequences. In Flint, this is supported by the format-updates function, which accepts a collection of update maps instead of a single one.

The example:

[{:prefixes    {:foaf "<http://xmlns.com/foaf/0.1/>"}
  :delete-data [[:graph "<http://census.marley/data>"
                        [["<http://census.marley/entry#221325>" :foaf/familyName "Brown"]]]]}
 {:insert-data [[:graph "<http://census.marley/data>"
                        [["<http://census.marley/entry#221325>" :foaf/familyName "Braun"]]]]}]

becomes:

PREFIX foaf: <http://xmlns.com/foaf/0.1/>
DELETE DATA {
    GRAPH <http://census.marley/data> {
        <http://census.marley/entry#212325> foaf:familyName "Brown" .
    }
};
INSERT DATA {
    GRAPH <http://census.marley/data> {
        <http://census.marley/entry#212325> foaf:familyName "Braun" .
    }
}

Note that only the first update map has a :prefixes clause. In Flint, prefixes in subsequent update maps are merged and overwrite duplicate prefixes.