Package org.apache.any23.extractor.html

Examples of org.apache.any23.extractor.html.HCardExtractor

321
322
323
324
325
326
327
328
329
330
331
        fw.setAnnotated(annotate);
        outputMediaType = factory.getMimeType();
        List<TripleHandler> tripleHandlers = new ArrayList<TripleHandler>();
        tripleHandlers.add(new IgnoreAccidentalRDFa(fw));
        tripleHandlers.add(new CountingTripleHandler());
        rdfWriter = new CompositeTripleHandler(tripleHandlers);
        reporter = new ReportingTripleHandler(rdfWriter);
        rdfWriter = new IgnoreAccidentalRDFa(
            new IgnoreTitlesOfEmptyDocuments(reporter),
            true    // suppress stylesheet triples.
        );
View Full Code Here
232
233
234
235
236
237
238
239
240
241
242
243
244
    private SingleDocumentExtraction getInstance(String file) throws FileNotFoundException, IOException {
        baos = new ByteArrayOutputStream();
        rdfxmlWriter = new RDFXMLWriter(baos);
        repositoryWriter = new RepositoryWriter(conn);

        final CompositeTripleHandler cth = new CompositeTripleHandler();
        cth.addChild(rdfxmlWriter);
        cth.addChild(repositoryWriter);

        final ModifiableConfiguration configuration = DefaultConfiguration.copy();
        configuration.setProperty("any23.extraction.metadata.domain.per.entity", "on");
        SingleDocumentExtraction instance =  new SingleDocumentExtraction(
                configuration,
View Full Code Here
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
        final ExtractionContext extractionContext = new ExtractionContext(
                extractor.getDescription().getExtractorName(),
                RDFUtils.uri("file://" + resource)
        );
        final InputStream is = this.getClass().getResourceAsStream(resource);
        final CompositeTripleHandler compositeTripleHandler = new CompositeTripleHandler();
        final TripleHandler verifierTripleHandler = Mockito.mock(TripleHandler.class);
        compositeTripleHandler.addChild(verifierTripleHandler);
        final CountingTripleHandler countingTripleHandler = new CountingTripleHandler();
        compositeTripleHandler.addChild(countingTripleHandler);
        final ByteArrayOutputStream out = new ByteArrayOutputStream();
        compositeTripleHandler.addChild( new NTriplesWriter(out) );
        final ExtractionResult extractionResult = new ExtractionResultImpl(
                extractionContext, extractor, compositeTripleHandler
        );
        extractor.run(extractionParameters, extractionContext, is, extractionResult);
        compositeTripleHandler.close();
        logger.info(out.toString());

        verifyPredicateOccurrence(verifierTripleHandler, Excel.getInstance().containsSheet, 2 );
        verifyPredicateOccurrence(verifierTripleHandler, Excel.getInstance().containsRow  , 6 );
        verifyPredicateOccurrence(verifierTripleHandler, Excel.getInstance().containsCell , 18);
View Full Code Here
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
        ByteArrayOutputStream baos = new ByteArrayOutputStream();

        CountingTripleHandler cth1 = new CountingTripleHandler();
        NTriplesWriter ctw1 = new NTriplesWriter(baos);
        CompositeTripleHandler compositeTH1 = new CompositeTripleHandler();
        compositeTH1.addChild(cth1);
        compositeTH1.addChild(ctw1);
        try {
            runner.extract(
                    new ExtractionParameters(
                            DefaultConfiguration.singleton(),
                            ValidationMode.None
                    ),
                    source,
                    compositeTH1
            );
        } finally {
            compositeTH1.close();
        }
        logger.info(baos.toString());
        Assert.assertEquals("Unexpected number of triples.", EXPECTED_TRIPLES, cth1.getCount() );

        baos.reset();
        CountingTripleHandler cth2 = new CountingTripleHandler();
        NTriplesWriter ctw2 = new NTriplesWriter(baos);
        CompositeTripleHandler compositeTH2 = new CompositeTripleHandler();
        compositeTH2.addChild(cth2);
        compositeTH2.addChild(ctw2);
        runner.extract(
                new ExtractionParameters(
                        DefaultConfiguration.singleton(),
                        ValidationMode.ValidateAndFix
                ),
View Full Code Here
320
321
322
323
324
325
326
327
328
329
330
        FormatWriter fw = factory.getRdfWriter(byteOutStream);
        fw.setAnnotated(annotate);
        outputMediaType = factory.getMimeType();
        List<TripleHandler> tripleHandlers = new ArrayList<TripleHandler>();
        tripleHandlers.add(new IgnoreAccidentalRDFa(fw));
        tripleHandlers.add(new CountingTripleHandler());
        rdfWriter = new CompositeTripleHandler(tripleHandlers);
        reporter = new ReportingTripleHandler(rdfWriter);
        rdfWriter = new IgnoreAccidentalRDFa(
            new IgnoreTitlesOfEmptyDocuments(reporter),
            true    // suppress stylesheet triples.
View Full Code Here
85
86
87
88
89
90
91
92
93
94
95
        );
        final InputStream is = this.getClass().getResourceAsStream(resource);
        final CompositeTripleHandler compositeTripleHandler = new CompositeTripleHandler();
        final TripleHandler verifierTripleHandler = Mockito.mock(TripleHandler.class);
        compositeTripleHandler.addChild(verifierTripleHandler);
        final CountingTripleHandler countingTripleHandler = new CountingTripleHandler();
        compositeTripleHandler.addChild(countingTripleHandler);
        final ByteArrayOutputStream out = new ByteArrayOutputStream();
        compositeTripleHandler.addChild( new NTriplesWriter(out) );
        final ExtractionResult extractionResult = new ExtractionResultImpl(
                extractionContext, extractor, compositeTripleHandler
View Full Code Here
315
316
317
318
319
320
321
322
323
324
325
326
                    null,
                    report
            );
            return false;
        }
        FormatWriter fw = factory.getRdfWriter(byteOutStream);
        fw.setAnnotated(annotate);
        outputMediaType = factory.getMimeType();
        List<TripleHandler> tripleHandlers = new ArrayList<TripleHandler>();
        tripleHandlers.add(new IgnoreAccidentalRDFa(fw));
        tripleHandlers.add(new CountingTripleHandler());
        rdfWriter = new CompositeTripleHandler(tripleHandlers);
View Full Code Here
88
89
90
91
92
93
94
95
96
97
98
        final TripleHandler verifierTripleHandler = Mockito.mock(TripleHandler.class);
        compositeTripleHandler.addChild(verifierTripleHandler);
        final CountingTripleHandler countingTripleHandler = new CountingTripleHandler();
        compositeTripleHandler.addChild(countingTripleHandler);
        final ByteArrayOutputStream out = new ByteArrayOutputStream();
        compositeTripleHandler.addChild( new NTriplesWriter(out) );
        final ExtractionResult extractionResult = new ExtractionResultImpl(
                extractionContext, extractor, compositeTripleHandler
        );
        extractor.run(extractionParameters, extractionContext, is, extractionResult);
        compositeTripleHandler.close();
View Full Code Here
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    @Test
    public void testTypedLiteralIncompatibleValueSupport()
    throws IOException, ExtractionException, TripleHandlerException {
        final URI uri = RDFUtils.uri("http://host.com/test-malformed-literal.turtle");
        ByteArrayOutputStream baos = new ByteArrayOutputStream();
        final TripleHandler th = new RDFXMLWriter(baos);
        final ExtractionContext extractionContext = new ExtractionContext("turtle-extractor", uri);
        final ExtractionResult result = new ExtractionResultImpl(extractionContext, extractor, th);
        extractor.setStopAtFirstError(false);
        try {
            extractor.run(
                    ExtractionParameters.newDefault(),
                    extractionContext,
                    this.getClass().getResourceAsStream("/org/apache/any23/extractor/rdf/testMalformedLiteral"),
                    result
            );
        } finally {
            logger.debug(baos.toString());
            th.close();
            result.close();
        }
    }
View Full Code Here
229
230
231
232
233
234
235
236
237
238
239
        assertTripleCount(vSINDICE.getProperty(SINDICE.NESTING_ORIGINAL)  , vREVIEW.hasReview, 1);
    }

    private SingleDocumentExtraction getInstance(String file) throws FileNotFoundException, IOException {
        baos = new ByteArrayOutputStream();
        rdfxmlWriter = new RDFXMLWriter(baos);
        repositoryWriter = new RepositoryWriter(conn);

        final CompositeTripleHandler cth = new CompositeTripleHandler();
        cth.addChild(rdfxmlWriter);
        cth.addChild(repositoryWriter);
View Full Code Here
TOP

Related Classes of org.apache.any23.extractor.html.HCardExtractor

  • org.apache.any23.AbstractAny23TestBase
  • org.apache.any23.Any23
  • org.apache.any23.Any23Test
  • org.apache.any23.cli.ExtractorDocumentation
  • org.apache.any23.cli.MicrodataParser
  • org.apache.any23.cli.MicrodataParser$MicrodataParserDocumentSourceConverter
  • org.apache.any23.cli.MimeDetector
  • org.apache.any23.cli.MimeDetector$MimeDetectorDocumentSourceConverter
  • org.apache.any23.cli.Rover
  • org.apache.any23.cli.Tool
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.