SPARQL query in R

Ekaterina Chuprikova

unread,

Sep 23, 2021, 11:31:24 AM9/23/21

to ontop4obda

Hi all,

I would like to ask for help regarding SPARQL endpoint and R. The endpoint is running, I can make queries. But when I try to access the data from R it gives me such an error:

SPARQL(endpoint, query)

Space required after the Public Identifier

SystemLiteral " or ' expected

SYSTEM or PUBLIC, the URI is missing

Error: 1: Space required after the Public Identifier

2: SystemLiteral " or ' expected

3: SYSTEM or PUBLIC, the URI is missing

When I do the same in Python it works partially. I can access the data, but I cannot create a data frame with such an error:

RuntimeWarning: unknown response content type 'text/sparql-results+csv;charset=UTF-8' returning raw response...
warnings.warn("unknown response content type '%s' returning raw response..." %(ct), RuntimeWarning)

My R code below.

Thank your in advance!

Best,

Ekaterina

library(DBI)

library(RPostgreSQL)

library(ggplot2)

library(XML)

library(RCurl)

library(SPARQL) # SPARQL querying package

endpoint <- 'http://localhost:8080/sparql'

prefix <- c("","http://ob-visly.com/apples/")

sparql_prefix <- "PREFIX : <http://ob-visly.com/apples/>

PREFIX base: <http://ob-visly.com/apples/>

PREFIX owl: <http://www.w3.org/2002/07/owl#>

PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>

PREFIX xml: <http://www.w3.org/XML/1998/namespace>

PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>

PREFIX obda: <https://w3id.org/obda/vocabulary#>

PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>

"

q <- paste(sparql_prefix,

'SELECT DISTINCT ?apple {

?apple a :sorts .

}'

)

SPARQL(endpoint,q,ns=prefix,extra=options)$results

roman.ko...@gmail.com

unread,

Sep 26, 2021, 12:16:52 PM9/26/21

to ontop4obda

Hi Ekaterina

I am no R guru, but a quick search online suggests that such an error (Space required after the Public Identifier ..) is usually a result of the endpoint misconfiguration - the server (endpoint) returns an HTML that begins with "<!DOCTYPE HTML PUBLIC" instead of the expected SPARQL result. When you say "the endpoint is running", what exactly do you mean? You also mention Python code - could you provide it please?

Best

Roman

Ekaterina Chuprikova

unread,

Sep 29, 2021, 12:24:15 PM9/29/21

to ontop4obda

Thank you, Roman! I might call the thing in a wrong way. What I meant by running endpoint, is that I can access the data through the http://localhost:8080/sparql and query it.

What do you recommend to check in the endpoint misconfiguration? Could it be a problem of encoding? In the ontology, I have everything in English and I check the files they are saved as UTF-8.

Thank you again!

My python code and its output:

from SPARQLWrapper import SPARQLWrapper, JSON

import sparql_dataframe
from pandas import DataFrame
from rdflib.plugins.sparql.processor import SPARQLResult

from SPARQLWrapper import SPARQLWrapper, SPARQLWrapper2, JSON, JSONLD, CSV, TSV, N3, RDF, RDFXML, TURTLE
import pandas as pds
import itertools

endpoint = "http://localhost:8080/sparql"
q = """

PREFIX : <http://ob-visly.com/apples/>
PREFIX base: <http://ob-visly.com/apples/>
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX xml: <http://www.w3.org/XML/1998/namespace>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
PREFIX obda: <https://w3id.org/obda/vocabulary#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>

SELECT DISTINCT ?apple {
?apple a :sorts .
}
"""

sparql = SPARQLWrapper("http://localhost:8080/sparql")
sparql.setQuery(q)
sparql.setReturnFormat(JSON)
results = sparql.query().convert()
print(results)

df = sparql_dataframe.get(endpoint, q)

Output:

"C:\Program Files\Python39\python.exe" "C:/PATH/sparql/tests.py"

{'head': {'vars': ['apple']}, 'results': {'bindings': [{'apple': {'type': 'uri', 'value': 'http://ob-visly.com/apples/sorts/ID_sort=7'}}, {'apple': {'type': 'uri', 'value': 'http://ob-visly.com/apples/sorts/ID_sort=21'}},.........OTHER ID_SORT ....., {'apple': {'type': 'uri', 'value': 'http://ob-visly.com/apples/sorts/ID_sort=1076483'}}]}}

C:\ PATH \AppData\Roaming\Python\Python39\site-packages\SPARQLWrapper\Wrapper.py:1346: RuntimeWarning: unknown response content type 'text/sparql-results+csv;charset=UTF-8' returning raw response...

warnings.warn("unknown response content type '%s' returning raw response..." %(ct), RuntimeWarning)

Process finished with exit code 0

Ekaterina Chuprikova

unread,

Oct 14, 2021, 11:46:35 AM10/14/21

to ontop4obda

Thank you all for answering my question:

@Peter Hopfgartner has suggested a solution that worked for me. I publish it below:

I was able to solve the issue with the out-of-date libcurl version.

For that you do have to compile RCurl on your own using Rtools (https://cran.r-project.org/bin/windows/Rtools/). With these, you have to install the current versions of libcurl (see https://github.com/r-windows/docs/blob/master/rtools40.md#readme), by opening the Rtools bash and running the commands:

pacman -Sy

pacman -S mingw-w64-{i686,x86_64}-curl

After that, the RCurl package can be resinstalled from source:

install.packages("RCurl", type="source")

All the best,

Ekaterina

Reply all

Reply to author

Forward