如何修改Jena Java程序，将SPARQL单变量查询改为双变量并返回Java数组

Got it, let's tweak your existing Jena query method to run the requested SPARQL query and send back the results as a 2D Java array (each row holds a ?s and ?class pair). Here's the adjusted code and breakdown:

// Updated return type from void to Object[][] to return query results
private Object[][] runQuery(String query, Model model) {
    String defaultNameSpace = "your-actual-namespace-uri"; // Replace with your real ESCO namespace
    List<Object[]> resultsList = new ArrayList<>();

    // Build full query with required prefixes
    StringBuffer queryStr = new StringBuffer();
    queryStr.append("PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> ");
    queryStr.append("PREFIX esco: <" + defaultNameSpace + "> ");
    queryStr.append(query); // Attach the incoming query string

    try {
        // Create and execute the SPARQL query
        Query sparqlQuery = QueryFactory.create(queryStr.toString());
        // Use try-with-resources to auto-close QueryExecution and avoid leaks
        try (QueryExecution qexec = QueryExecutionFactory.create(sparqlQuery, model)) {
            ResultSet rs = qexec.execSelect();

            // Loop through results and collect each ?s + ?class pair
            while (rs.hasNext()) {
                QuerySolution soln = rs.nextSolution();
                // Extract URIs as strings (you can keep Resource objects if needed instead)
                String subjectUri = soln.getResource("s").getURI();
                String classUri = soln.getResource("class").getURI();
                resultsList.add(new Object[]{subjectUri, classUri});
            }
        }
    } catch (QueryParseException e) {
        // Handle query parsing errors gracefully
        e.printStackTrace();
        return new Object[0][0]; // Return empty array on failure
    }

    // Convert the list to a 2D Java array and return
    return resultsList.toArray(new Object[resultsList.size()][]);
}

Key Adjustments Explained: #

• Return Type Shift: Swapped void for Object[][] so we can pass back the query results. If you want a strictly typed array like String[][], just adjust the array declarations accordingly.
• Critical Prefix Addition: Added the rdf: prefix since your query uses rdf:type—omitting this would cause parsing errors.
• Result Collection: Used an ArrayList to gather result pairs first (flexible for dynamic result counts), then converted it to a 2D array to meet your requirement.
• Resource Safety: Used a try-with-resources block for QueryExecution to ensure it's properly closed after use, preventing resource leaks.
• Error Handling: Added basic catch logic for query parsing issues, returning an empty array if something goes wrong.

How to Use the Updated Method: #

Call it with your target query string like this:

String targetQuery = "select distinct ?s ?class where {?s rdf:type ?class }";
Object[][] queryResults = runQuery(targetQuery, yourJenaModel);

// Iterate over the results to use them
for (Object[] row : queryResults) {
    String subject = (String) row[0];
    String clazz = (String) row[1];
    System.out.printf("Subject: %s, Class: %s%n", subject, clazz);
}

内容的提问来源于stack exchange，提问作者marco