Package org.apache.tika.exception

Examples of org.apache.tika.exception.TikaException

149
150
151
152
153
154
155
156
157
    public static List<Parser> getParsersFromZip(URL zip, TikaConfig config)
            throws TikaException, IOException {
        String zipMimeType = config.getMimeRepository().getMimeType(zip)
        .getName();
        if (!zipMimeType.equalsIgnoreCase("application/zip")) {
            throw new TikaException("The file you are using is note a zip file");
        }
        return getParsersFromZip(zip.openStream(), config);
    }
View Full Code Here
172
173
174
175
176
177
178
179
            StringWriter writer = new StringWriter();
            parser.parse(
                    stream, new WriteOutContentHandler(writer), new Metadata());
            return writer.toString();
        } catch (SAXException e) {
            throw new TikaException("Unexpected SAX error", e);
        }
    }
View Full Code Here
79
80
81
82
83
84
85
86
87
88
89
        Parser parser = config.getParser(type.getName());
        if (parser == null) {
            parser = config.getParser(MimeTypes.DEFAULT);
        }
        if (parser == null) {
            throw new TikaException("No parsers available: " + type.getName());
        }

        // Parse the document
        parser.parse(stream, handler, metadata);
    }
View Full Code Here
53
54
55
56
57
58
59
60
61
            new PDF2XHTML(handler, metadata).getText(document);
        } catch (IOException e) {
            if (e.getCause() instanceof SAXException) {
                throw (SAXException) e.getCause();
            } else {
                throw new TikaException("Unable to extract PDF content", e);
            }
        }
    }
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
70
71
        din.read(header);
        din.close();

        int info = LittleEndian.getShort(header, 0xa);
        if ((info & 0x4) != 0) {
            throw new TikaException("Fast-saved files are unsupported");
        }
        if ((info & 0x100) != 0) {
            throw new TikaException("This document is password protected");
        }

        // determine the version of Word this document came from.
        int nFib = LittleEndian.getShort(header, 0x2);
        switch (nFib) {
View Full Code Here
53
54
55
56
57
58
59
60
61
62
                    Tag tag = (Tag)tags.next();
                    metadata.set(tag.getTagName(), tag.getDescription());
                }
            }
        } catch (JpegProcessingException e) {
            throw new TikaException("Can't read JPEG metadata", e);
        } catch (MetadataException e) {
            throw new TikaException("Can't read JPEG metadata", e);
        }
    }
View Full Code Here
71
72
73
74
75
76
77
78
79
80
81
                    metadata.set("height", Integer.toString(reader.getHeight(0)));
                    metadata.set("width", Integer.toString(reader.getWidth(0)));
                    reader.dispose();
                }
            } catch (IIOException e) {
                throw new TikaException(type + " parse error", e);
            }
        }

        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
View Full Code Here
142
143
144
145
146
147
148
149
        // Return the remaining substring
        try {
            return new String(buffer, start, end - start, "ISO-8859-1");
        } catch (UnsupportedEncodingException e) {
            throw new TikaException("ISO-8859-1 encoding is not available", e);
        }
    }
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
            return parser;
        } else {
            try {
                return getSAXParserFactory(context).newSAXParser();
            } catch (ParserConfigurationException e) {
                throw new TikaException("Unable to configure a SAX parser", e);
            } catch (SAXException e) {
                throw new TikaException("Unable to create a SAX parser", e);
            }
        }
    }
View Full Code Here
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
            }
        }

        String encoding = metadata.get(Metadata.CONTENT_ENCODING);
        if (encoding == null) {
            throw new TikaException(
                    "Text encoding could not be detected and no encoding"
                    + " hint is available in document metadata");
        }

        // TIKA-341: Only stomp on content-type after we're done trying to
        // use it to guess at the charset.
        metadata.set(Metadata.CONTENT_TYPE, "text/plain");

        try {
            Reader reader =
                new BufferedReader(new InputStreamReader(stream, encoding));

            // TIKA-240: Drop the BOM when extracting plain text
            reader.mark(1);
            int bom = reader.read();
            if (bom != '\ufeff') { // zero-width no-break space
                reader.reset();
            }

            XHTMLContentHandler xhtml =
                new XHTMLContentHandler(handler, metadata);
            xhtml.startDocument();

            xhtml.startElement("p");
            char[] buffer = new char[4096];
            int n = reader.read(buffer);
            while (n != -1) {
                xhtml.characters(buffer, 0, n);
                n = reader.read(buffer);
            }
            xhtml.endElement("p");

            xhtml.endDocument();
        } catch (UnsupportedEncodingException e) {
            throw new TikaException(
                    "Unsupported text encoding: " + encoding, e);
        }
    }
View Full Code Here
TOP

Related Classes of org.apache.tika.exception.TikaException

  • de.innosystec.unrar.tika.RARParser
  • org.apache.jackrabbit.core.query.lucene.JackrabbitParser
  • org.apache.jackrabbit.core.query.pdf.PDF2XHTML
  • org.apache.nutch.parse.tika.TikaConfig
  • org.apache.tika.config.TikaConfig
  • org.apache.tika.detect.AutoDetectReader
  • org.apache.tika.embedder.ExternalEmbedder
  • org.apache.tika.extractor.ParserContainerExtractor
  • org.apache.tika.fork.ForkClient
  • org.apache.tika.fork.ForkParser
    • 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.