Package org.apache.tika.metadata

Examples of org.apache.tika.metadata.Metadata

303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
        return shorter;
    }

    public String getType(String typeName, String url, byte[] data) {
        try {
            Metadata metadata = new Metadata();
            if (url != null) {
                metadata.set(Metadata.RESOURCE_NAME_KEY, url);
            }
            if (typeName != null) {
                metadata.set(Metadata.CONTENT_TYPE, typeName);
            }
            return detect(new ByteArrayInputStream(data), metadata).toString();
        } catch (IOException e) {
            throw new IllegalStateException(
                    "ByteArrayInputStream throws an IOException!", e);
View Full Code Here
330
331
332
333
334
335
336
337
338
339
340
341
     * @throws IOException if the document can not be accessed
     */
    public String getType(URL url) throws IOException {
        InputStream stream = url.openStream();
        try {
            Metadata metadata = new Metadata();
            metadata.set(Metadata.RESOURCE_NAME_KEY, url.toString());
            return detect(stream, metadata).toString();
        } finally {
            stream.close();
        }
    }
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
public class OutlookParserTest extends TestCase {

    public void testOutlookParsing() throws Exception {
        Parser parser = new AutoDetectParser(); // Should auto-detect!
        ContentHandler handler = new BodyContentHandler();
        Metadata metadata = new Metadata();

        InputStream stream = OutlookParserTest.class.getResourceAsStream(
                "/test-documents/test-outlook.msg");
        try {
            parser.parse(stream, handler, metadata);
        } finally {
            stream.close();
        }

        assertEquals(
                "application/vnd.ms-outlook",
                metadata.get(Metadata.CONTENT_TYPE));
        assertEquals(
                "Microsoft Outlook Express 6",
                metadata.get(Metadata.TITLE));
        // TODO: There's apparently some encoding issue in POI
        //assertEquals(
        //        "L'\u00C9quipe Microsoft Outlook Express",
        //        metadata.get(Metadata.AUTHOR));
View Full Code Here
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
     * @see <a href="https://issues.apache.org/jira/browse/TIKA-197">TIKA-197</a>
     */
    public void testMultipleCopies() throws Exception {
        Parser parser = new AutoDetectParser();
        ContentHandler handler = new BodyContentHandler();
        Metadata metadata = new Metadata();

        InputStream stream = OutlookParserTest.class.getResourceAsStream(
                "/test-documents/testMSG.msg");
        try {
            parser.parse(stream, handler, metadata);
        } finally {
            stream.close();
        }

        assertEquals(
                "application/vnd.ms-outlook",
                metadata.get(Metadata.CONTENT_TYPE));

        String content = handler.toString();
        Pattern pattern = Pattern.compile("From");
        Matcher matcher = pattern.matcher(content);
        assertTrue(matcher.find());
View Full Code Here
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
     * @see <a href="https://issues.apache.org/jira/browse/TIKA-395">TIKA-395</a>
     */
    public void testOutlookNew() throws Exception {
        Parser parser = new AutoDetectParser();
        ContentHandler handler = new BodyContentHandler();
        Metadata metadata = new Metadata();

        InputStream stream = OutlookParserTest.class.getResourceAsStream(
                "/test-documents/test-outlook2003.msg");
        try {
            parser.parse(stream, handler, metadata);
        } finally {
            stream.close();
        }

        assertEquals(
                "application/vnd.ms-outlook",
                metadata.get(Metadata.CONTENT_TYPE));
        assertEquals(
                "Welcome to Microsoft Office Outlook 2003",
                metadata.get(Metadata.TITLE));

        String content = handler.toString();
        assertTrue(content.contains("Outlook 2003"));
        assertTrue(content.contains("Streamlined Mail Experience"));
        assertTrue(content.contains("Navigation Pane"));
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
    public void testExcel() throws Exception {
        InputStream input = OOXMLParserTest.class
                .getResourceAsStream("/test-documents/testEXCEL.xlsx");

        Parser parser = new AutoDetectParser();
        Metadata metadata = new Metadata();
        // TODO: should auto-detect without the resource name
        metadata.set(Metadata.RESOURCE_NAME_KEY, "testEXCEL.xlsx");
        ContentHandler handler = new BodyContentHandler();

        try {
            parser.parse(input, handler, metadata);

            assertEquals(
                    "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
                    metadata.get(Metadata.CONTENT_TYPE));
            assertEquals("Simple Excel document", metadata.get(Metadata.TITLE));
            assertEquals("Keith Bennett", metadata.get(Metadata.AUTHOR));
            String content = handler.toString();
            assertTrue(content.contains("Sample Excel Worksheet"));
            assertTrue(content.contains("Numbers and their Squares"));
            assertTrue(content.contains("9"));
            assertFalse(content.contains("9.0"));
            assertTrue(content.contains("196"));
            assertFalse(content.contains("196.0"));
            assertEquals("false", metadata.get(TikaMetadataKeys.PROTECTED));
        } finally {
            input.close();
        }
    }
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
    public void testExcelFormats() throws Exception {
        InputStream input = OOXMLParserTest.class
                .getResourceAsStream("/test-documents/testEXCEL-formats.xlsx");

        Parser parser = new AutoDetectParser();
        Metadata metadata = new Metadata();
        // TODO: should auto-detect without the resource name
        metadata.set(Metadata.RESOURCE_NAME_KEY, "testEXCEL-formats.xlsx");
        ContentHandler handler = new BodyContentHandler();

        try {
            parser.parse(input, handler, metadata);

            assertEquals(
                    "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
                    metadata.get(Metadata.CONTENT_TYPE));

            String content = handler.toString();

            // Number #,##0.00
            assertTrue(content.contains("1,599.99"));
View Full Code Here
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
    public void testPowerPoint() throws Exception {
        InputStream input = OOXMLParserTest.class
                .getResourceAsStream("/test-documents/testPPT.pptx");

        Parser parser = new AutoDetectParser();
        Metadata metadata = new Metadata();
        // TODO: should auto-detect without the resource name
        metadata.set(Metadata.RESOURCE_NAME_KEY, "testPPT.pptx");
        ContentHandler handler = new BodyContentHandler();

        try {
            parser.parse(input, handler, metadata);

            assertEquals(
                    "application/vnd.openxmlformats-officedocument.presentationml.presentation",
                    metadata.get(Metadata.CONTENT_TYPE));
            assertEquals("Sample Powerpoint Slide", metadata.get(Metadata.TITLE));
            assertEquals("Keith Bennett", metadata.get(Metadata.AUTHOR));
            String content = handler.toString();
            assertTrue(content.contains("Sample Powerpoint Slide"));
            assertTrue(content.contains("Powerpoint X for Mac"));
        } finally {
            input.close();
View Full Code Here
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
    public void testExcelParser() throws Exception {
        InputStream input = ExcelParserTest.class.getResourceAsStream(
                "/test-documents/testEXCEL.xls");
        try {
            Metadata metadata = new Metadata();
            ContentHandler handler = new BodyContentHandler();
            new OfficeParser().parse(input, handler, metadata);

            assertEquals(
                    "application/vnd.ms-excel",
                    metadata.get(Metadata.CONTENT_TYPE));
            assertEquals("Simple Excel document", metadata.get(Metadata.TITLE));
            assertEquals("Keith Bennett", metadata.get(Metadata.AUTHOR));
            String content = handler.toString();
            assertTrue(content.contains("Sample Excel Worksheet"));
            assertTrue(content.contains("Numbers and their Squares"));
            assertTrue(content.contains("\t\tNumber\tSquare"));
            assertTrue(content.contains("9"));
View Full Code Here
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
    public void testWord() throws Exception {
        InputStream input = OOXMLParserTest.class
                .getResourceAsStream("/test-documents/testWORD.docx");

        Parser parser = new AutoDetectParser();
        Metadata metadata = new Metadata();
        // TODO: should auto-detect without the resource name
        metadata.set(Metadata.RESOURCE_NAME_KEY, "testWORD.docx");
        ContentHandler handler = new BodyContentHandler();

        try {
            parser.parse(input, handler, metadata);

            assertEquals(
                    "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
                    metadata.get(Metadata.CONTENT_TYPE));
            assertEquals("Sample Word Document", metadata.get(Metadata.TITLE));
            assertEquals("Keith Bennett", metadata.get(Metadata.AUTHOR));
            assertTrue(handler.toString().contains("Sample Word Document"));
        } finally {
            input.close();
        }
    }
View Full Code Here
TOP

Related Classes of org.apache.tika.metadata.Metadata

  • bixo.parser.TikaCallableTest
  • com.tamingtext.tika.TikaTest
  • cx.fbn.nevernote.threads.IndexRunner
  • org.apache.any23.mime.TikaMIMETypeDetector
  • org.apache.chemistry.opencmis.client.parser.MetadataParserTika
  • org.apache.cxf.jaxrs.ext.search.tika.TikaContentExtractor
  • org.apache.jackrabbit.core.query.lucene.NodeIndexer
  • org.apache.jackrabbit.oak.plugins.index.lucene.LuceneIndexEditor
  • org.apache.jackrabbit.server.io.DefaultHandler
  • org.apache.jackrabbit.server.io.ImportContextImpl
    • 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.