IRI with space

[skodapetr]

File input.ttl contains IRI with encoded space:

_:b25978837	a <http://purl.bioontology.org/ontology/UATC/\u0020SERINE\u0020\u0020> .

The Turtle parser is able to read this IRI (with encoded spaces) as a valid IRI.
However Turtle writer does not encode the space on output, resulting in:

_:genid-ca45343d5be841429acf1ca16a6fe928-b25978837 a <http://purl.bioontology.org/ontology/UATC/ SERINE  > .

When we try to load this statement the parser fail with:

Exception in thread "main" org.eclipse.rdf4j.rio.RDFParseException: IRI included an unencoded space: '32' [line 2]
	at org.eclipse.rdf4j.rio.helpers.RDFParserHelper.reportError(RDFParserHelper.java:322)
	at org.eclipse.rdf4j.rio.helpers.AbstractRDFParser.reportError(AbstractRDFParser.java:601)
...

A simple program can be utilized to reproduce this error:

    public static void main(String[] args) throws Exception {
        try (InputStream input = new FileInputStream(new File("input.ttl"));
                OutputStream output = new FileOutputStream(new File("output.ttl"))) {
            RDFParser parser = Rio.createParser(RDFFormat.TURTLE);
            parser.setRDFHandler(Rio.createWriter(RDFFormat.TURTLE, output));
            parser.parse(input, "http://localhost/base/");
        }
        try (InputStream input = new FileInputStream(new File("output.ttl"))) {
            RDFParser parser = Rio.createParser(RDFFormat.TURTLE);
            parser.parse(input, "http://localhost/base/");
        }
    }

This problem also propagates to the querying remote SPARQL endpoints.

The same issue can be experienced when using N-Triples, however for JSON-LD and RDF/XML no error is given.
The comment in TurtleParser suggest that the \n and similar (spaces?) should be handled as errors.

It's not clear what is an intended behaviour.

Tested with rdf4j 2.2 .

[barthanssens]

I assume this is related to #69 (and that there is no strict checking on generating/writing IRIs because it impacts performance)

[abrokenjester]

Also related to #712

[KnowledgeGarden]

Importing a ttl file with the TURTLE format, line 1 of which is this:
http://dbpedia.org/property/name http://www.w3.org/1999/02/22-rdf-syntax-ns#type http://www.w3.org/1999/02/22-rdf-syntax-ns#Property .
landed this log statement:
org.eclipse.rdf4j.rio.RDFParseException: IRI included an unencoded space: '32' [line 1]
at org.eclipse.rdf4j.rio.helpers.RDFParserHelper.reportError(RDFParserHelper.java:322)

[KnowledgeGarden]

I now believe that my errors reported here were based on inappropriate use of the APIs. By switching to Model-based parsing before adding to the repo, all errors disappeared.