apacheGH-2039: Case insensitive langtags (initial)

AKSW · Dec 27, 2023 · 74a134d · 74a134d
1 parent 7bf97bf
commit 74a134d
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 3 deletions.
diff --git a/jena-arq/src/main/java/org/apache/jena/sparql/lang/QueryParserBase.java b/jena-arq/src/main/java/org/apache/jena/sparql/lang/QueryParserBase.java
@@ -215,7 +215,7 @@ protected Node createLiteral(String lexicalForm, String langTag, String datatype
             RDFDatatype dType = TypeMapper.getInstance().getSafeTypeByName(datatypeURI);
             n = NodeFactory.createLiteral(lexicalForm, dType);
         } else if ( langTag != null && !langTag.isEmpty() )
-            n = NodeFactory.createLiteral(lexicalForm, langTag);
+            n = NodeFactory.createLiteralLang(lexicalForm, langTag);
         else
             n = NodeFactory.createLiteral(lexicalForm);
         return n;

diff --git a/jena-core/src/main/java/org/apache/jena/graph/impl/LiteralLabel.java b/jena-core/src/main/java/org/apache/jena/graph/impl/LiteralLabel.java
@@ -106,7 +106,7 @@ private enum ValueMode { EAGER , LAZY }
     /*package*/ LiteralLabel(String lex, String lang, RDFDatatype dtype) {
         this.lexicalForm = lex;
         this.dtype = Objects.requireNonNull(dtype);
-        this.lang = (lang == null ? "" : lang);
+        this.lang = formLangTag(lang);
         hash = calcHashCode();
         if ( valueMode == ValueMode.EAGER ) {
             this.wellformed = setValue(lex, dtype);
@@ -116,6 +116,24 @@ private enum ValueMode { EAGER , LAZY }
             value = null;
     }
 
+    private static final boolean legacyLangTag = true;
+
+    /** Prepare the language tag - apply formatting normalization */
+    private static String formLangTag(String input) {
+        if ( legacyLangTag )
+            return (input == null ? "" : input);
+        // Format.
+        return (input == null ? "" : input.toLowerCase(Locale.ROOT));
+    }
+
+    /** Calculate the indexing form for a language tag */
+    private static String indexingLang(String lang) {
+        if ( legacyLangTag )
+            return lang.toLowerCase(Locale.ROOT);
+        return lang;
+    }
+
+
     /**
      * Build a typed literal label from its value form using
      * whatever datatype is currently registered as the default
@@ -304,7 +322,8 @@ public Object getIndexingValue() {
         if ( indexingValueIsSelf() )
             return this;
         if ( !lang.equals("") )
-            return getLexicalForm() + "@" + lang.toLowerCase(Locale.ROOT);
+            // Assumed formatted/case-insensitive language tags.
+            return getLexicalForm() + "@" + indexingLang(lang);
         if ( wellformed ) {
             Object value = getValue();
             // JENA-1936