Skip to content

Latest commit

 

History

History
122 lines (97 loc) · 4.44 KB

README.md

File metadata and controls

122 lines (97 loc) · 4.44 KB
Raw

SPARQL-cheatsheet

All those SPARQL commands & tricks that you keep on forgetting ヽ༼ຈل͜ຈ༽ﾉ

Contribuitons are welcomed. Send you pull requests

Query Forms

SELECT

...

CONSTRUCT

Returns a graph specified by a graph template.

within the CONSTRUCT triples are created based on the variables established within the WHERE query

PREFIX wd: <http://www.wikidata.org/entity/>
PREFIX wdt: <http://www.wikidata.org/prop/direct/>
PREFIX mo: <http://semantics.id/ns/example/movie#>
prefix ex: <http://semantics.id/ns/example#> 
CONSTRUCT
    { 
     ?iri a mo:Movie ;
        mo:hasWikidataLink ?s ;
        mo:hasIMDBResource ?imdbRes ;
        mo:title ?title ;
    }
WHERE {
    SERVICE <https://query.wikidata.org/sparql>
    {
        ?s wdt:P31 wd:Q24869 ; #instance_of feature film  
        wdt:P345 ?imdbId ; 
        wdt:P1476 ?title .
        BIND(IRI(CONCAT("https://www.imdb.com/title/", ?imdbId )) AS ?imdbRes) .
        BIND( IRI(CONCAT("http://semantics.id/ns/example#", ?imdbId )) AS ?iri) .
   }
}
LIMIT 2

Resulting in new triples

@prefix ex:  <http://semantics.id/ns/example#> .
@prefix mo:  <http://semantics.id/ns/example/movie#> .
@prefix wd:  <http://www.wikidata.org/entity/> .
@prefix wdt: <http://www.wikidata.org/prop/direct/> .

ex:tt0461770  a             mo:Movie ;
        mo:hasIMDBResource  <https://www.imdb.com/title/tt0461770> ;
        mo:hasWikidataLink  wd:Q490464 ;
        mo:title            "Enchanted"@en .

ex:tt0070215  a             mo:Movie ;
        mo:hasIMDBResource  <https://www.imdb.com/title/tt0070215> ;
        mo:hasWikidataLink  wd:Q964009 ;
        mo:title            "Il mio nome è Nessuno"@it .

ASK

DESCRIBE

String Operations

  • UCASE() convert to upper-case
  • LCASE() convert to lower-case
  • STRLEN() lenght of a string
  • SUBSTR() substring of first arg, second arg: start position; third arg: character lenght;
  • STRSTARTS()string in first arg starts with string in second arg
  • STRENDS() string in first arg ends with string in second arg
  • CONTAINS() string in second arg is within first arg. Example: CONTAINS("SPARQL", "QL") is True. FILTER(CONTAINS(str(?term), "iao") || CONTAINS(str(?term), "bfo") || CONTAINS(str(?term), "obo"))

Logic operators

  • sameTerm & !sameTerm (more efficient and =)
  • =, !=, >, >=, <, <=
  • AND: && OR: ||
  • IN and NOT IN: ?foo IN ("bar", "anotherbar")
  • isIRI isIRI
  • isBlank: only bank nodes FILTER (isBlank(?term)) filter out FILTER (!isBlank(?term))
  • isLiteral

Functions

  • lang Returns the language tag.
    • FILTER(lang(?label) = 'es')
    • any @en variant (will match all of these: @en, @en-US, @en-GB) FILTER(langMatches(lang(?label), 'en'))
  • datatype returns the datatype of a literal
    • BIND( datatype(?shoeSize) AS ?datatype)

Aggregation Functions

  • COUNT
  • SUM
  • AVG
  • MIN, MAX
  • SAMPLE
  • GROUP_CONCAT

Tips to speedup SPARQL Queries:

  • reduce search space
  • avoid OPTIONAL clauses (when possible), or have them follown non-optional clauses
  • avoid FILTER or try having them earlier - so search space is reduced:
    • FILTER(sameTerm(?term1, ?term2)) is faster that FILTER(?terms = ?term)
    • idem for !sameTerm being faster that !=
    • regexs are expensive
  • avoid DISTINCT
  • avoid SORT BY in large amount of results
  • reduce printout named variables - allows for more optimization flexibility from the SPARQL processor

SPARQL Resources