Package org.apache.tika.parser

Examples of org.apache.tika.parser.ParseContext

249
250
251
252
253
254
255
        xhtml.endDocument();
    }

    public void parse(InputStream stream, ContentHandler handler,
            Metadata metadata) throws IOException, SAXException, TikaException {
        parse(stream, handler, metadata, new ParseContext());
    }
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
        StringWriter writer = new StringWriter();
        parser.parse(
                new ByteArrayInputStream(text.getBytes("UTF-8")),
                new WriteOutContentHandler(writer),
                metadata,
                new ParseContext());
        String content = writer.toString();

        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals("en", metadata.get(Metadata.CONTENT_LANGUAGE));
        assertEquals("en", metadata.get(Metadata.LANGUAGE));
View Full Code Here
65
66
67
68
69
70
71
72
73
74
75
        ContentHandler handler = new BodyContentHandler();
        Metadata metadata = new Metadata();
        parser.parse(
                new ByteArrayInputStream(text.getBytes("UTF-8")),
                handler, metadata, new ParseContext());
        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals("UTF-8", metadata.get(Metadata.CONTENT_ENCODING));

        assertTrue(handler.toString().contains(text));
    }
View Full Code Here
76
77
78
79
80
81
82
83
84
    public void testEmptyText() throws Exception {
        ContentHandler handler = new BodyContentHandler();
        Metadata metadata = new Metadata();
        parser.parse(
                new ByteArrayInputStream(new byte[0]), handler, metadata, new ParseContext());
        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals("\n", handler.toString());
    }
View Full Code Here
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
        final String test2 = "the name is \u00e1ndre";

        Metadata metadata = new Metadata();
        parser.parse(
                new ByteArrayInputStream(test2.getBytes("UTF-8")),
                new BodyContentHandler(),  metadata, new ParseContext());
       
        assertEquals("UTF-8", metadata.get(Metadata.CONTENT_ENCODING));

        metadata.set(Metadata.CONTENT_ENCODING, "ISO-8859-1");
        parser.parse(
                new ByteArrayInputStream(test2.getBytes("UTF-8")),
                new BodyContentHandler(),  metadata, new ParseContext());
       
        assertEquals("ISO-8859-1", metadata.get(Metadata.CONTENT_ENCODING));
    }
View Full Code Here
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
        final String test2 = "the name is \u00e1ndre";

        Metadata metadata = new Metadata();
        parser.parse(
                new ByteArrayInputStream(test2.getBytes("UTF-8")),
                new BodyContentHandler(),  metadata, new ParseContext());

        assertEquals("UTF-8", metadata.get(Metadata.CONTENT_ENCODING));

        metadata = new Metadata();
        metadata.set(Metadata.CONTENT_TYPE, "text/html; charset=ISO-8859-1");
        parser.parse(
                new ByteArrayInputStream(test2.getBytes("UTF-8")),
                new BodyContentHandler(),  metadata, new ParseContext());

        assertEquals("ISO-8859-1", metadata.get(Metadata.CONTENT_ENCODING));
    }
View Full Code Here
66
67
68
69
70
71
72
     * @deprecated This method will be removed in Apache Tika 1.0.
     */
    public void parse(
            InputStream stream, ContentHandler handler, Metadata metadata)
            throws IOException, SAXException, TikaException {
        parse(stream, handler, metadata, new ParseContext());
    }
View Full Code Here
154
155
156
157
158
159
160
161
162
            public void ignorableWhitespace(char[] ch, int off, int len) {
                // Ignore the whitespace added by XHTMLContentHandler
            }
        };
        Metadata metadata = new Metadata();
        parser.parse(new ByteArrayInputStream(input), handler, metadata, new ParseContext());
        assertEquals("text/plain", metadata.get(Metadata.CONTENT_TYPE));
        assertEquals(msg, expected, handler.toString());
    }
View Full Code Here
172
173
174
175
176
177
178
179
180
        Metadata metadata = new Metadata();
        metadata.set(Metadata.LANGUAGE, "en");

        parser.parse(
                new ByteArrayInputStream(test.getBytes("UTF-8")),
                new BodyContentHandler(),  metadata, new ParseContext());

        assertEquals("en", metadata.get(Metadata.LANGUAGE));
    }
View Full Code Here
136
137
138
139
140
141
142
     * @deprecated This method will be removed in Apache Tika 1.0.
     */
    public void parse(
            InputStream stream, ContentHandler handler, Metadata metadata)
            throws IOException, SAXException, TikaException {
        parse(stream, handler, metadata, new ParseContext());
    }
View Full Code Here
TOP

Related Classes of org.apache.tika.parser.ParseContext

  • com.cloudera.cdk.morphline.solrcell.SolrCellBuilder$SolrCell
  • com.nidhinova.tika.server.TikaService
  • com.tamingtext.tika.TikaTest
  • cx.fbn.nevernote.threads.IndexRunner
  • gov.nara.nwts.ftappImg.jpeg.JpegExtractor
  • org.apache.chemistry.opencmis.client.parser.MetadataParserTika
  • org.apache.jackrabbit.core.query.lucene.JackrabbitParser
  • org.apache.jackrabbit.core.query.lucene.LazyTextExtractorField$ParsingTask
  • org.apache.jackrabbit.oak.plugins.index.lucene.LuceneIndexEditor
  • org.apache.nutch.parse.tika.DOMContentUtilsTest
    • 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.