Package org.apache.any23.configuration

Examples of org.apache.any23.configuration.ModifiableConfiguration

387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
        RDFXMLWriter ctw1 = new RDFXMLWriter(baos);
        CompositeTripleHandler compositeTH1 = new CompositeTripleHandler();
        compositeTH1.addChild(cth1);
        compositeTH1.addChild(ctw1);
        runner.extract(
                new ExtractionParameters(
                        DefaultConfiguration.singleton(),
                        ValidationMode.None, true
                ),
                source,
                compositeTH1
        );
        compositeTH1.close();
        logger.debug("Out1: " + baos.toString());
        Assert.assertEquals("Unexpected number of triples.", EXPECTED_TRIPLES + 3, 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, false),
                source,
                compositeTH2
        );
View Full Code Here
260
261
262
263
264
265
266
267
                String.format("Invalid value '%s' for '%s' parameter.", validationMode, PARAMETER)
        );
    }
   
    private ExtractionParameters getExtractionParameters(HttpServletRequest request) {
        final ValidationMode mode = getValidationMode(request);
        return new ExtractionParameters(DefaultConfiguration.singleton(), mode);
    }
View Full Code Here
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    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
65
66
67
68
69
70
71
72
73
74
75
    }

    @Test
    public void testRun() throws IOException, ExtractionException {
        final InputStream is = this.getClass().getResourceAsStream("html-scraper-extractor-test.html");
        final ExtractionResult extractionResult = mock(ExtractionResult.class);
        final URI pageURI = ValueFactoryImpl.getInstance().createURI("http://fake/test/page/testrun");
        final ExtractionContext extractionContext = new ExtractionContext(
                extractor.getDescription().getExtractorName(),
                pageURI
        );
View Full Code Here
89
90
91
92
93
94
95
96
97
98
99
        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());
View Full Code Here
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
    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
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
        );
        hCardExtractor.run(
                ExtractionParameters.newDefault(),
                hcExtractionContext,
                document,
                new ExtractionResultImpl(
                        hcExtractionContext,
                        hCardExtractor,
                        new RepositoryWriter(getConnection())
                )
        );
        XFNExtractor xfnExtractor = new XFNExtractorFactory().createExtractor();
        ExtractionContext xfnExtractionContext = new ExtractionContext(
                xfnExtractor.getDescription().getExtractorName(),
                baseURI
        );
        xfnExtractor.run(
                        ExtractionParameters.newDefault(),
                        xfnExtractionContext,
                        document,
                        new ExtractionResultImpl(
                                xfnExtractionContext,
                                hCardExtractor,
                                new RepositoryWriter(getConnection())
                        )
                );
View Full Code Here
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
        );
        hCardExtractor.run(
                ExtractionParameters.newDefault(),
                hCardExtractionContext,
                document,
                new ExtractionResultImpl(
                        hCardExtractionContext,
                        hCardExtractor, new RepositoryWriter(getConnection())
                )
        );

        GeoExtractor geoExtractor = new GeoExtractorFactory().createExtractor();
        ExtractionContext geoExtractionContext = new ExtractionContext(
                geoExtractor.getDescription().getExtractorName(), baseURI
        );
        geoExtractor.run(
                ExtractionParameters.newDefault(),
                geoExtractionContext,
                document,
                new ExtractionResultImpl(
                        geoExtractionContext,
                        geoExtractor,
                        new RepositoryWriter(getConnection())
                )
        );

        AdrExtractor adrExtractor = new AdrExtractorFactory().createExtractor();
        ExtractionContext adrExtractionContext = new ExtractionContext(
                adrExtractor.getDescription().getExtractorName(), baseURI
        );
        adrExtractor.run(
                ExtractionParameters.newDefault(),
                adrExtractionContext,
                document,
                new ExtractionResultImpl(
                        adrExtractionContext,
                        adrExtractor,
                        new RepositoryWriter(getConnection())
                )
        );
View Full Code Here
529
530
531
532
533
534
535
536
537
538
539
        );
        hReviewExtractor.run(
                ExtractionParameters.newDefault(),
                hreviewExtractionContext,
                document,
                new ExtractionResultImpl(
                        hreviewExtractionContext,
                        hReviewExtractor,
                        new RepositoryWriter(getConnection())
                )
        );
View Full Code Here
89
90
91
92
93
94
95
96
97
98
99
        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());
View Full Code Here
TOP

Related Classes of org.apache.any23.configuration.ModifiableConfiguration

  • 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.configuration.ModifiableConfiguration
    • 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.