Package org.apache.tika.io

Examples of org.apache.tika.io.TemporaryResources

85
86
87
88
89
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
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
            InputStream stream, ContentHandler handler,
            Metadata metadata, ParseContext context)
            throws IOException, SAXException, TikaException {
      
        PDDocument pdfDocument = null;
        TemporaryResources tmp = new TemporaryResources();
        //config from context, or default if not set via context
        PDFParserConfig localConfig = context.get(PDFParserConfig.class, defaultConfig);
        try {
            // PDFBox can process entirely in memory, or can use a temp file
            //  for unpacked / processed resources
            // Decide which to do based on if we're reading from a file or not already
            TikaInputStream tstream = TikaInputStream.cast(stream);
            if (tstream != null && tstream.hasFile()) {
                // File based, take that as a cue to use a temporary file
                RandomAccess scratchFile = new RandomAccessFile(tmp.createTemporaryFile(), "rw");
                if (localConfig.getUseNonSequentialParser() == true){
                    pdfDocument = PDDocument.loadNonSeq(new CloseShieldInputStream(stream), scratchFile);
                } else {
                    pdfDocument = PDDocument.load(new CloseShieldInputStream(stream), scratchFile, true);
                }
            } else {
                // Go for the normal, stream based in-memory parsing
                if (localConfig.getUseNonSequentialParser() == true){
                    pdfDocument = PDDocument.loadNonSeq(new CloseShieldInputStream(stream), new RandomAccessBuffer());
                } else {
                    pdfDocument = PDDocument.load(new CloseShieldInputStream(stream), true);
                }
            }
           
          
            if (pdfDocument.isEncrypted()) {
                String password = null;
               
                // Did they supply a new style Password Provider?
                PasswordProvider passwordProvider = context.get(PasswordProvider.class);
                if (passwordProvider != null) {
                   password = passwordProvider.getPassword(metadata);
                }
               
                // Fall back on the old style metadata if set
                if (password == null && metadata.get(PASSWORD) != null) {
                   password = metadata.get(PASSWORD);
                }
               
                // If no password is given, use an empty string as the default
                if (password == null) {
                   password = "";
                }
              
                try {
                    pdfDocument.decrypt(password);
                } catch (Exception e) {
                    // Ignore
                }
            }
            metadata.set(Metadata.CONTENT_TYPE, "application/pdf");
            extractMetadata(pdfDocument, metadata);
            PDF2XHTML.process(pdfDocument, handler, context, metadata, localConfig);
           
        } finally {
            if (pdfDocument != null) {
               pdfDocument.close();
            }
            tmp.dispose();
        }
        handler.endDocument();
    }
View Full Code Here
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
        // Check if we have access to the document
        if (input == null) {
            return MediaType.OCTET_STREAM;
        }

        TemporaryResources tmp = new TemporaryResources();
        try {
            TikaInputStream tis = TikaInputStream.get(input, tmp);

            byte[] prefix = new byte[1024]; // enough for all known formats
            int length = tis.peek(prefix);

            MediaType type = detectArchiveFormat(prefix, length);
            if (PackageParser.isZipArchive(type)
                    && TikaInputStream.isTikaInputStream(input)) {
                return detectZipFormat(tis);
            } else if (!type.equals(MediaType.OCTET_STREAM)) {
                return type;
            } else {
                return detectCompressorFormat(prefix, length);
            }
        } finally {
            try {
                tmp.dispose();
            } catch (TikaException e) {
                // ignore
            }
        }
    }
View Full Code Here
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
                entrydata.set(Metadata.EMBEDDED_RELATIONSHIP_ID, name);
            }
            if (extractor.shouldParseEmbedded(entrydata)) {
                // For detectors to work, we need a mark/reset supporting
                // InputStream, which ArchiveInputStream isn't, so wrap
                TemporaryResources tmp = new TemporaryResources();
                try {
                    TikaInputStream tis = TikaInputStream.get(archive, tmp);
                    extractor.parseEmbedded(tis, xhtml, entrydata, true);
                } finally {
                    tmp.dispose();
                }
            }
        } else if (name != null && name.length() > 0) {
            xhtml.element("p", name);
        }
View Full Code Here
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
        // Check if we have access to the document
        if (input == null) {
            return MediaType.OCTET_STREAM;
        }

        TemporaryResources tmp = new TemporaryResources();
        try {
            TikaInputStream tis = TikaInputStream.get(input, tmp);

            byte[] prefix = new byte[1024]; // enough for all known formats
            int length = tis.peek(prefix);

            MediaType type = detectArchiveFormat(prefix, length);
            if (PackageParser.isZipArchive(type)
                    && TikaInputStream.isTikaInputStream(input)) {
                return detectZipFormat(tis);
            } else if (!type.equals(MediaType.OCTET_STREAM)) {
                return type;
            } else {
                return detectCompressorFormat(prefix, length);
            }
        } finally {
            try {
                tmp.dispose();
            } catch (TikaException e) {
                // ignore
            }
        }
    }
View Full Code Here
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
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
155
156
157
158
159
160
161
162
163
164
            InputStream stream, ContentHandler handler,
            Metadata metadata, ParseContext context)
            throws IOException, SAXException, TikaException {
      
        PDDocument pdfDocument = null;
        TemporaryResources tmp = new TemporaryResources();

        try {
            // PDFBox can process entirely in memory, or can use a temp file
            //  for unpacked / processed resources
            // Decide which to do based on if we're reading from a file or not already
            TikaInputStream tstream = TikaInputStream.cast(stream);
            if (tstream != null && tstream.hasFile()) {
               // File based, take that as a cue to use a temporary file
               RandomAccess scratchFile = new RandomAccessFile(tmp.createTemporaryFile(), "rw");
               pdfDocument = PDDocument.load(new CloseShieldInputStream(stream), scratchFile, true);
            } else {
               // Go for the normal, stream based in-memory parsing
               pdfDocument = PDDocument.load(new CloseShieldInputStream(stream), true);
            }
          
            if (pdfDocument.isEncrypted()) {
                String password = null;
               
                // Did they supply a new style Password Provider?
                PasswordProvider passwordProvider = context.get(PasswordProvider.class);
                if (passwordProvider != null) {
                   password = passwordProvider.getPassword(metadata);
                }
               
                // Fall back on the old style metadata if set
                if (password == null && metadata.get(PASSWORD) != null) {
                   password = metadata.get(PASSWORD);
                }
               
                // If no password is given, use an empty string as the default
                if (password == null) {
                   password = "";
                }
              
                try {
                    pdfDocument.decrypt(password);
                } catch (Exception e) {
                    // Ignore
                }
            }
            metadata.set(Metadata.CONTENT_TYPE, "application/pdf");
            extractMetadata(pdfDocument, metadata);
            PDF2XHTML.process(pdfDocument, handler, metadata,
                              extractAnnotationText, enableAutoSpace,
                              suppressDuplicateOverlappingText, sortByPosition);

            extractEmbeddedDocuments(context, pdfDocument, handler);
        } finally {
            if (pdfDocument != null) {
               pdfDocument.close();
            }
            tmp.dispose();
        }
    }
View Full Code Here
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
    public void parse(
            InputStream stream, ContentHandler handler,
            Metadata metadata, ParseContext context)
            throws IOException, SAXException, TikaException {
        TemporaryResources tmp = new TemporaryResources();
        try {
            TikaInputStream tis = TikaInputStream.get(stream, tmp);
            new ImageMetadataExtractor(metadata).parseTiff(tis.getFile());
            new JempboxExtractor(metadata).parse(tis);
        } finally {
            tmp.dispose();
        }

        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.endDocument();
View Full Code Here
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
    public void parse(
            InputStream stream, ContentHandler handler,
            Metadata metadata, ParseContext context)
            throws IOException, SAXException, TikaException {
        TemporaryResources tmp = new TemporaryResources();
        try {
            TikaInputStream tis = TikaInputStream.get(stream, tmp);

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

            // TIKA-216: Zip bomb prevention
            SecureContentHandler sch = new SecureContentHandler(handler, tis);
            try {
                // Parse the document
                super.parse(tis, sch, metadata, context);
            } catch (SAXException e) {
                // Convert zip bomb exceptions to TikaExceptions
                sch.throwIfCauseOf(e);
                throw e;
            }
        } finally {
            tmp.dispose();
        }
    }
View Full Code Here
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
                entrydata.set(Metadata.RESOURCE_NAME_KEY, name);
            }
            if (extractor.shouldParseEmbedded(entrydata)) {
                // For detectors to work, we need a mark/reset supporting
                // InputStream, which ArchiveInputStream isn't, so wrap
                TemporaryResources tmp = new TemporaryResources();
                try {
                    TikaInputStream tis = TikaInputStream.get(archive, tmp);
                    extractor.parseEmbedded(tis, xhtml, entrydata, true);
                } finally {
                    tmp.dispose();
                }
            }
        } else if (name != null && name.length() > 0) {
            xhtml.element("p", name);
        }
View Full Code Here
32
33
34
35
36
37
38
39
40
41
42
final class EmbeddedExtractor {

  public boolean parseEmbedded(InputStream stream, Record record, String name, Command child) {
    // Use the delegate parser to parse this entry
   
    TemporaryResources tmp = new TemporaryResources();
    try {
      final TikaInputStream newStream = TikaInputStream.get(new CloseShieldInputStream(stream), tmp);
      if (stream instanceof TikaInputStream) {
        final Object container = ((TikaInputStream) stream).getOpenContainer();
        if (container != null) {
View Full Code Here
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
        Record entrydata = new Record(); // TODO: or pass myself?
        //Record entrydata = record.copy();
       
        // For detectors to work, we need a mark/reset supporting
        // InputStream, which ArchiveInputStream isn't, so wrap
        TemporaryResources tmp = new TemporaryResources();
        try {
          TikaInputStream tis = TikaInputStream.get(archive, tmp);
          return extractor.parseEmbedded(tis, entrydata, name, getChild());
        } finally {
          try {
            tmp.dispose();
          } catch (TikaException e) {
            LOG.warn("Cannot dispose of tmp Tika resources", e);
          }
        }
      } else {
View Full Code Here
TOP

Related Classes of org.apache.tika.io.TemporaryResources

  • com.cloudera.cdk.morphline.tika.decompress.EmbeddedExtractor
  • com.cloudera.cdk.morphline.tika.decompress.UnpackBuilder$Unpack
  • org.apache.tika.extractor.ParserContainerExtractor$RecursiveParser
  • org.apache.tika.extractor.ParsingEmbeddedDocumentExtractor
  • org.apache.tika.parser.AutoDetectParser
  • org.apache.tika.parser.CompositeParser
  • org.apache.tika.parser.external.ExternalParser
  • org.apache.tika.parser.gdal.GDALParser
  • org.apache.tika.parser.image.TiffParser
  • org.apache.tika.parser.jpeg.JpegParser
    • 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.