Package org.apache.tika.mime

Examples of org.apache.tika.mime.MediaType

110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
                if (inputDoc.getContent() != null) {
                    Metadata meta = new Metadata();
                    meta.set(Metadata.RESOURCE_NAME_KEY, inputDoc.getUrl());
                    MimeType mimetype = null;
                    try {
                        MediaType mediaType = detector
                                .detect(new ByteArrayInputStream(inputDoc
                                        .getContent()), meta);
                        mimetype = mimetypes.forName(mediaType.getType() + "/"
                                + mediaType.getSubtype());
                    } catch (IOException e) {
                        LOG.error("Exception", e);
                    } catch (MimeTypeException e) {
                        LOG.error("Exception", e);
                    }
View Full Code Here
69
70
71
72
73
74
75
76
77
78
79
        if (!stream.markSupported()) {
            stream = new BufferedInputStream(stream);
        }

        // Automatically detect the MIME type of the document
        MediaType type = types.detect(stream, metadata);
        metadata.set(Metadata.CONTENT_TYPE, type.toString());

        // Parse the document
        super.parse(stream, handler, metadata);
    }
View Full Code Here
29
30
31
32
33
34
35
36
37
38
39
* Test cases for the {@link MagicDetector} class.
*/
public class MagicDetectorTest extends TestCase {

    public void testDetectNull() throws Exception {
        MediaType html = new MediaType("text", "html");
        Detector detector = new MagicDetector(html, "<html".getBytes("ASCII"));
        assertEquals(
                MediaType.OCTET_STREAM,
                detector.detect(null, new Metadata()));
    }
View Full Code Here
37
38
39
40
41
42
43
44
45
46
47
                MediaType.OCTET_STREAM,
                detector.detect(null, new Metadata()));
    }

    public void testDetectSimple() throws Exception {
        MediaType html = new MediaType("text", "html");
        Detector detector = new MagicDetector(html, "<html".getBytes("ASCII"));

        assertDetect(detector, html, "<html");
        assertDetect(detector, html, "<html><head/><body/></html>");
        assertDetect(detector, MediaType.OCTET_STREAM, "<HTML");
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
        assertDetect(detector, MediaType.OCTET_STREAM, " <html");
        assertDetect(detector, MediaType.OCTET_STREAM, "");
    }

    public void testDetectOffsetRange() throws Exception {
        MediaType html = new MediaType("text", "html");
        Detector detector = new MagicDetector(
                html, "<html".getBytes("ASCII"), null, 0, 64);

        assertDetect(detector, html, "<html");
        assertDetect(detector, html, "<html><head/><body/></html>");
View Full Code Here
73
74
75
76
77
78
79
80
81
82
83
        assertDetect(detector, MediaType.OCTET_STREAM, "");
}

    public void testDetectMask() throws Exception {
        MediaType html = new MediaType("text", "html");
        byte up = (byte) 0xdf;
        Detector detector = new MagicDetector(
                html,
                new byte[] { '<',  'H',  'T',  'M',  'L' },
                new byte[] { (byte) 0xff, up, up, up, up },
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
45
46
47
        this.detectors = detectors;
    }

    public MediaType detect(InputStream input, Metadata metadata)
            throws IOException {
        MediaType type = MediaType.OCTET_STREAM;
        for (Detector detector : detectors) {
            MediaType detected = detector.detect(input, metadata);
            if (detected.isSpecializationOf(type)) {
                type = detected;
            }
        }
        return type;
    }
View Full Code Here
43
44
45
46
47
48
49
50
51
52
53
     */
    public MediaType detect(InputStream input, Metadata metadata) {
        // Look for a type hint in the input metadata
        String hint = metadata.get(Metadata.CONTENT_TYPE);
        if (hint != null) {
            MediaType type = MediaType.parse(hint);
            if (type != null) {
                return type;
            }
        }
        return MediaType.OCTET_STREAM;
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
        checkArgument(representations != null && representations.length > 0);
        int maxIndex = 0;
        double maxQ = 0.0;
        for (int i = 0; i < representations.length; i++) {
            double q = 0.0;
            MediaType type = registry.normalize(representations[i].getType());
            for (MediaRange range : ranges) {
                q = Math.max(q, range.match(type, registry));
            }
            if (q > maxQ) {
                maxIndex = i;
View Full Code Here
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
    private final MediaType type;

    private final double q;

    public static MediaRange parse(String range, MediaTypeRegistry registry) {
        MediaType type = MediaType.parse(range);
        if (type == null) {
            return null;
        }
        type = registry.normalize(type);

        Map<String, String> parameters =
                new HashMap<String, String>(type.getParameters());
        String q = parameters.remove("q");
        if (q != null) {
            try {
                return new MediaRange(
                        new MediaType(type.getBaseType(), parameters),
                        Double.parseDouble(q));
            } catch (NumberFormatException e) {
                return null;
            }
        }
View Full Code Here
TOP

Related Classes of org.apache.tika.mime.MediaType

  • bixo.utils.HttpUtils
  • com.cloudera.cdk.morphline.solrcell.SolrCellBuilder$SolrCell
  • com.cloudera.cdk.morphline.tika.DetectMimeTypeBuilder$DetectMimeType
  • com.digitalpebble.behemoth.tika.TikaProcessor
  • controllers.CodeApp
  • dk.i2m.drupal.util.EntityUtils
  • edu.isi.karma.rdf.GenericRDFGenerator
  • io.lumify.tikaMimeType.TikaMimeTypeMapper
  • org.apache.cxf.jaxrs.ext.search.tika.TikaContentExtractor
  • org.apache.jackrabbit.oak.http.AcceptHeader
    • 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.