Package org.apache.tika.exception

Examples of org.apache.tika.exception.TikaException

109
110
111
112
113
114
115
116
            Parser parser = config.getParser(mimeType);
            ContentHandler handler = new BodyContentHandler();
            parser.parse(stream, handler, new Metadata());
            return handler.toString();
        } catch (SAXException e) {
            throw new TikaException("Unexpected SAX error", e);
        }
    }
View Full Code Here
90
91
92
93
94
95
96
97
98
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
125
126
127
128
129
            }
        }

        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");
        }

        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
56
57
58
59
60
61
62
63
64
            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
56
57
58
59
60
61
62
63
            SAXParser parser = factory.newSAXParser();
            parser.parse(
                    new CloseShieldInputStream(stream),
                    new OfflineContentHandler(xhtml));
        } catch (ParserConfigurationException e) {
            throw new TikaException("XML parser configuration error", e);
        }
    }
View Full Code Here
61
62
63
64
65
66
67
68
69
70
71
72
73
                   
                    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);
        }

        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.endDocument();
View Full Code Here
43
44
45
46
47
48
49
50
51
            reader.accept(visitor, ClassReader.SKIP_FRAMES | ClassReader.SKIP_CODE);
        } catch (RuntimeException e) {
            if (e.getCause() instanceof SAXException) {
                throw (SAXException) e.getCause();
            } else {
                throw new TikaException("Failed to parse a Java class", e);
            }
        }
    }
View Full Code Here
154
155
156
157
158
159
160
161
162
163
164
165
                parse(new DocumentSummaryInformation(properties), metadata);
            }
        } catch (FileNotFoundException e) {
            // entry does not exist, just skip it
        } catch (NoPropertySetStreamException e) {
            throw new TikaException("Not a HPSF document", e);
        } catch (UnexpectedPropertySetTypeException e) {
            throw new TikaException("Unexpected HPSF document", e);
        } catch (MarkUnsupportedException e) {
            throw new TikaException("Invalid DocumentInputStream", e);
        }
    }
View Full Code Here
321
322
323
324
325
326
327
328
            parser.parse(
                    new CloseShieldInputStream(stream),
                    new OfflineContentHandler(
                            new NSNormalizerContentHandler(dh)));
        } catch (ParserConfigurationException e) {
            throw new TikaException("XML parser configuration error", e);
        }
    }
View Full Code Here
48
49
50
51
52
53
54
55
                new XHTMLContentHandler(handler, metadata);
            xhtml.startDocument();
            xhtml.element("p", sd.getText(0, sd.getLength()));
            xhtml.endDocument();
        } catch (BadLocationException e) {
            throw new TikaException("Error parsing an RTF document", e);
        }
    }
View Full Code Here
109
110
111
112
113
114
115
116
117
118
119
            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
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.