Fix #780: Validate all IRIs in AbstractRDFParser by default

Signed-off-by: James Leigh <james.leigh@ontotext.com>
eclipse-rdf4j · May 31, 2017 · 828be54 · 828be54
1 parent c38e52f
commit 828be54
Show file tree

Hide file tree

Showing 10 changed files with 68 additions and 19 deletions.
diff --git a/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/rdfjson/RDFJSONParserCustomTest.java b/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/rdfjson/RDFJSONParserCustomTest.java
@@ -24,7 +24,7 @@ public class RDFJSONParserCustomTest {
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(17, Rio.createParser(RDFFormat.RDFJSON).getSupportedSettings().size());
+		assertEquals(18, Rio.createParser(RDFFormat.RDFJSON).getSupportedSettings().size());
 	}
 
 }
diff --git a/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/rdfxml/RDFXMLParserCustomTest.java b/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/rdfxml/RDFXMLParserCustomTest.java
@@ -223,6 +223,6 @@ public void testParseCommentAtStart()
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(21, Rio.createParser(RDFFormat.RDFXML).getSupportedSettings().size());
+		assertEquals(22, Rio.createParser(RDFFormat.RDFXML).getSupportedSettings().size());
 	}
 }
diff --git a/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/trig/TriGParserCustomTest.java b/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/trig/TriGParserCustomTest.java
@@ -226,7 +226,7 @@ public void testBadPname02()
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(12, Rio.createParser(RDFFormat.TRIG).getSupportedSettings().size());
+		assertEquals(13, Rio.createParser(RDFFormat.TRIG).getSupportedSettings().size());
 	}
 
 }
diff --git a/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/turtle/CustomTurtleParserTest.java b/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/turtle/CustomTurtleParserTest.java
@@ -7,19 +7,25 @@
  *******************************************************************************/
 package org.eclipse.rdf4j.rio.turtle;
 
-import static org.junit.Assert.*;
+import static org.junit.Assert.assertEquals;
+import static org.junit.Assert.assertFalse;
+import static org.junit.Assert.assertTrue;
+import static org.junit.Assert.fail;
 
 import java.io.StringReader;
 import java.io.StringWriter;
 import java.util.Collections;
+import java.util.function.Consumer;
 
 import org.eclipse.rdf4j.model.IRI;
 import org.eclipse.rdf4j.model.Literal;
 import org.eclipse.rdf4j.model.Model;
 import org.eclipse.rdf4j.model.Namespace;
+import org.eclipse.rdf4j.model.Value;
 import org.eclipse.rdf4j.model.ValueFactory;
 import org.eclipse.rdf4j.model.impl.LinkedHashModel;
 import org.eclipse.rdf4j.model.impl.NamespaceImpl;
+import org.eclipse.rdf4j.model.impl.SimpleValueFactory;
 import org.eclipse.rdf4j.model.impl.ValueFactoryImpl;
 import org.eclipse.rdf4j.model.util.Models;
 import org.eclipse.rdf4j.model.vocabulary.RDF;
@@ -206,7 +212,7 @@ public void testLiteralWithNewlines()
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(12, parser.getSupportedSettings().size());
+		assertEquals(13, parser.getSupportedSettings().size());
 	}
 
 	@Test
@@ -507,4 +513,24 @@ public void testParsingNamespacesWithOverride()
 		assertTrue(model.contains(vf.createIRI("urn:a"), vf.createIRI("urn:not_skos:broader"),
 				vf.createIRI("urn:b")));
 	}
+
+	@Test
+	public void test780IRISpace()
+		throws Exception
+	{
+		String ttl = "_:b25978837	a <http://purl.bioontology.org/ontology/UATC/\\u0020SERINE\\u0020\\u0020> .";
+		try {
+			Rio.parse(new StringReader(ttl), "", RDFFormat.TURTLE);
+			fail();
+		}
+		catch (RDFParseException e) {
+			// Invalid IRI
+		}
+		Model model = Rio.parse(new StringReader(ttl), "", RDFFormat.TURTLE,
+				new ParserConfig().set(BasicParserSettings.VERIFY_URI_SYNTAX, false),
+				SimpleValueFactory.getInstance(), new ParseErrorLogger());
+		assertEquals(1, model.size());
+		model.filter(null, RDF.TYPE, null).objects().forEach(obj -> assertEquals(
+				"http://purl.bioontology.org/ontology/UATC/%20SERINE%20%20", obj.stringValue()));
+	}
 }
diff --git a/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/turtle/TurtleParserTest.java b/compliance/rio/src/test/java/org/eclipse/rdf4j/rio/turtle/TurtleParserTest.java
@@ -7,13 +7,11 @@
  *******************************************************************************/
 package org.eclipse.rdf4j.rio.turtle;
 
-import junit.framework.Test;
-
 import org.eclipse.rdf4j.rio.RDFParser;
-import org.eclipse.rdf4j.rio.helpers.TurtleParserSettings;
+import org.eclipse.rdf4j.rio.helpers.BasicParserSettings;
 import org.eclipse.rdf4j.rio.ntriples.NTriplesParser;
-import org.eclipse.rdf4j.rio.turtle.TurtleParser;
-import org.eclipse.rdf4j.rio.turtle.TurtleParserTestCase;
+
+import junit.framework.Test;
 
 /**
  * JUnit test for the Turtle parser that uses the tests that are available
@@ -30,11 +28,16 @@ public static Test suite()
 	@Override
 	protected RDFParser createTurtleParser() {
 		RDFParser result = new TurtleParser();
+		// localName_with_assigned_nfc_PN_CHARS_BASE_character_boundaries contains illegal trailing character
+		result.set(BasicParserSettings.VERIFY_URI_SYNTAX, false);
 		return result;
 	}
 
 	@Override
 	protected RDFParser createNTriplesParser() {
-		return new NTriplesParser();
+		NTriplesParser result = new NTriplesParser();
+		// localName_with_assigned_nfc_PN_CHARS_BASE_character_boundaries contains illegal trailing character
+		result.set(BasicParserSettings.VERIFY_URI_SYNTAX, false);
+		return result;
 	}
 }
diff --git a/core/rio/api/src/main/java/org/eclipse/rdf4j/rio/helpers/AbstractRDFParser.java b/core/rio/api/src/main/java/org/eclipse/rdf4j/rio/helpers/AbstractRDFParser.java
@@ -198,6 +198,9 @@ public Collection<RioSetting<?>> getSupportedSettings() {
 		// Supported in RDFParserBase.resolveURI
 		result.add(BasicParserSettings.VERIFY_RELATIVE_URIS);
 
+		// Supported in createURI
+		result.add(BasicParserSettings.VERIFY_URI_SYNTAX);
+
 		// Supported in RDFParserBase.createBNode(String)
 		result.add(BasicParserSettings.PRESERVE_BNODE_IDS);
 
@@ -376,7 +379,15 @@ protected IRI resolveURI(String uriSpec)
 		throws RDFParseException
 	{
 		// Resolve relative URIs against base URI
-		ParsedIRI uri = ParsedIRI.create(uriSpec);
+		ParsedIRI uri;
+		try {
+			uri = new ParsedIRI(uriSpec);
+		}
+		catch (URISyntaxException e) {
+			reportError("Invalid IRI '" + uriSpec,
+					BasicParserSettings.VERIFY_URI_SYNTAX);
+			uri = ParsedIRI.create(uriSpec);
+		}
 
 		if (!uri.isAbsolute()) {
 			if (baseURI == null) {
@@ -403,6 +414,15 @@ protected IRI resolveURI(String uriSpec)
 	protected IRI createURI(String uri)
 		throws RDFParseException
 	{
+		if (getParserConfig().get(BasicParserSettings.VERIFY_URI_SYNTAX)) {
+			try {
+				new ParsedIRI(uri);
+			}
+			catch (URISyntaxException e) {
+				reportError(e.getMessage(),
+						BasicParserSettings.VERIFY_URI_SYNTAX);
+			}
+		}
 		try {
 			return valueFactory.createIRI(uri);
 		}

diff --git a/testsuites/rio/src/main/java/org/eclipse/rdf4j/rio/nquads/AbstractNQuadsParserUnitTest.java b/testsuites/rio/src/main/java/org/eclipse/rdf4j/rio/nquads/AbstractNQuadsParserUnitTest.java
@@ -709,7 +709,7 @@ public void assertHandler(int expected) {
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(12, parser.getSupportedSettings().size());
+		assertEquals(13, parser.getSupportedSettings().size());
 	}
 
 	protected abstract RDFParser createRDFParser();

diff --git a/...ites/rio/src/main/java/org/eclipse/rdf4j/rio/ntriples/AbstractNTriplesParserUnitTest.java b/...ites/rio/src/main/java/org/eclipse/rdf4j/rio/ntriples/AbstractNTriplesParserUnitTest.java
@@ -380,7 +380,7 @@ public void testBlankNodeIdentifiersRDF11()
 	public void testSupportedSettings()
 		throws Exception
 	{
-		assertEquals(12, createRDFParser().getSupportedSettings().size());
+		assertEquals(13, createRDFParser().getSupportedSettings().size());
 	}
 
 	@Test

diff --git a/testsuites/rio/src/main/resources/testcases/rdfjson/rdfjson-eval-file-11.nt b/testsuites/rio/src/main/resources/testcases/rdfjson/rdfjson-eval-file-11.nt
@@ -1,2 +1,2 @@
-<:_about> <http://purl.org/dc/terms/creator> <http://example.org/anna> .
-<:_anna> <http://xmlns.com/foaf/0.1/name> "Anna" .
+_:about <http://purl.org/dc/terms/creator> <http://example.org/anna> .
+_:anna <http://xmlns.com/foaf/0.1/name> "Anna" .
diff --git a/testsuites/rio/src/main/resources/testcases/rdfjson/rdfjson-syntax-file-11.rj b/testsuites/rio/src/main/resources/testcases/rdfjson/rdfjson-syntax-file-11.rj
@@ -1,10 +1,10 @@
 {
-  ":_about" : {
+  "_:about" : {
       "http://purl.org/dc/terms/creator" : [ { "value" : "http://example.org/anna", 
                                                "type" : "uri" } ] 
   },
-  ":_anna" : {
+  "_:anna" : {
       "http://xmlns.com/foaf/0.1/name" : [ { "value" : "Anna", 
                                              "type" : "literal" } ] 
   }
-}                                               
+}