(jena) branch main updated: GH-2250: Determine likely syntax using source URI

[andy]

This is an automated email from the ASF dual-hosted git repository.

andy pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/jena.git


The following commit(s) were added to refs/heads/main by this push:
     new c14b9d9db3 GH-2250: Determine likely syntax using source URI
c14b9d9db3 is described below

commit c14b9d9db396662d442654b51854363162958445
Author: Andy Seaborne <a...@apache.org>
AuthorDate: Tue Feb 6 16:36:58 2024 +0000

    GH-2250: Determine likely syntax using source URI
---
 .../main/java/org/apache/jena/riot/RDFParser.java  | 27 ++++++++++++++++------
 .../org/apache/jena/sparql/util/DatasetUtils.java  |  2 +-
 2 files changed, 21 insertions(+), 8 deletions(-)

diff --git a/jena-arq/src/main/java/org/apache/jena/riot/RDFParser.java 
b/jena-arq/src/main/java/org/apache/jena/riot/RDFParser.java
index 7ac679a379..8bde96cdb6 100644
--- a/jena-arq/src/main/java/org/apache/jena/riot/RDFParser.java
+++ b/jena-arq/src/main/java/org/apache/jena/riot/RDFParser.java
@@ -379,25 +379,38 @@ public class RDFParser {
     private void parseURI(StreamRDF destination) {
         // Source by uri or path.
         try (TypedInputStream input = openTypedInputStream(uri, path)) {
-            ReaderRIOT reader;
+            ReaderRIOT readerRiot;
             ContentType ct;
             if ( forceLang != null ) {
                 ReaderRIOTFactory r = RDFParserRegistry.getFactory(forceLang);
                 if ( r == null )
                     throw new RiotException("No parser registered for 
language: " + forceLang);
                 ct = forceLang.getContentType();
-                reader = createReader(r, forceLang);
+                readerRiot = createReader(r, forceLang);
             } else {
                 // No forced language.
-                // Conneg and hint, ignoring text/plain.
-                ct = WebContent.determineCT(input.getContentType(), hintLang, 
baseURI);
+                // Determine the syntax based on
+                //   Content-type, ignoring text/plain
+                //   hintLanguage
+                //   Any pathname extension from file or URI.
+                //
+                // Prefer the uri being read for more information, or oath, 
and the base if all else fails.
+
+                String target;
+                if ( uri != null )
+                    target = uri;
+                else if ( path != null )
+                    target = path.toString();
+                else
+                    target = baseURI;
+                ct = WebContent.determineCT(input.getContentType(), hintLang, 
target);
                 if ( ct == null )
                     throw new RiotException("Failed to determine the content 
type: (URI=" + baseURI + " : stream=" + input.getContentType()+")");
-                reader = createReader(ct);
-                if ( reader == null )
+                readerRiot = createReader(ct);
+                if ( readerRiot == null )
                     throw new RiotException("No parser registered for content 
type: " + ct.getContentTypeStr());
             }
-            read(reader, input, null, baseURI, context, ct, destination);
+            read(readerRiot, input, null, baseURI, context, ct, destination);
         }
     }
 
diff --git 
a/jena-arq/src/main/java/org/apache/jena/sparql/util/DatasetUtils.java 
b/jena-arq/src/main/java/org/apache/jena/sparql/util/DatasetUtils.java
index a0476ea448..e61d25e1d5 100644
--- a/jena-arq/src/main/java/org/apache/jena/sparql/util/DatasetUtils.java
+++ b/jena-arq/src/main/java/org/apache/jena/sparql/util/DatasetUtils.java
@@ -206,7 +206,7 @@ public class DatasetUtils
             for ( Iterator<String> iter = namedSourceList.iterator() ; 
iter.hasNext() ; ) {
                 String sourceURI = iter.next();
                 String absURI = baseURI(sourceURI, absBaseURI);
-                // Read to a tmp graph in case of syntax errors.
+                // Read into a tmp graph in case of syntax errors.
                 Graph gTmp = GraphFactory.createJenaDefaultGraph();
                 RDFDataMgr.read(gTmp, sourceURI, absBaseURI, null);
                 Node gn = NodeFactory.createURI(sourceURI);