Package org.apache.pdfbox.pdmodel.common.filespecification

Examples of org.apache.pdfbox.pdmodel.common.filespecification.PDEmbeddedFile

391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
        if (embeddedFileNames == null){
            return;
        }
        for (Map.Entry<String,COSObjectable> ent : embeddedFileNames.entrySet()) {
            PDComplexFileSpecification spec = (PDComplexFileSpecification) ent.getValue();
            PDEmbeddedFile file = spec.getEmbeddedFile();

            Metadata metadata = new Metadata();
            // TODO: other metadata?
            metadata.set(Metadata.RESOURCE_NAME_KEY, ent.getKey());
            metadata.set(Metadata.CONTENT_TYPE, file.getSubtype());
            metadata.set(Metadata.CONTENT_LENGTH, Long.toString(file.getSize()));

            if (embeddedExtractor.shouldParseEmbedded(metadata)) {
                TikaInputStream stream = TikaInputStream.get(file.createInputStream());
                try {
                    embeddedExtractor.parseEmbedded(
                            stream,
                            new EmbeddedContentHandler(handler),
                            metadata, false);
View Full Code Here
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
                                if (debug)
                                {
                                    System.err.println("Processing embedded file " + ent.getKey() + ":");
                                }
                                PDComplexFileSpecification spec = (PDComplexFileSpecification) ent.getValue();
                                PDEmbeddedFile file = spec.getEmbeddedFile();
                                if (file != null && file.getSubtype().equals("application/pdf"))
                                {
                                    if (debug)
                                    {
                                        System.err.println("  is PDF (size=" + file.getSize() + ")");
                                    }
                                    InputStream fis = file.createInputStream();
                                    PDDocument subDoc = null;
                                    try
                                    {
                                        subDoc = PDDocument.load(fis);
                                    }
View Full Code Here
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
        Map<String, COSObjectable> embeddedFileNames = embeddedFiles.getNames();
        Assert.assertEquals(1, embeddedFileNames.size());
        Map.Entry<String, COSObjectable> entry = embeddedFileNames.entrySet().iterator().next();
        LOG.info("Processing embedded file " + entry.getKey() + ":");
        PDComplexFileSpecification complexFileSpec = (PDComplexFileSpecification) entry.getValue();
        PDEmbeddedFile embeddedFile = complexFileSpec.getEmbeddedFile();

        File resultFile = new File(testResultsDir, name);
        FileOutputStream fos = new FileOutputStream(resultFile);
        InputStream is = embeddedFile.createInputStream();
        IOUtils.copy(is, fos);
        fos.close();
        is.close();

        LOG.info("  size: " + embeddedFile.getSize());
        assertEquals(embeddedFile.getSize(), resultFile.length());

        return resultFile;
    }
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
            fs.setFile( "Test.txt" );
            //create a dummy file stream, this would probably normally be a FileInputStream
            byte[] data = "This is the contents of the embedded file".getBytes("ISO-8859-1");
            ByteArrayInputStream fakeFile =
                new ByteArrayInputStream( data );
            PDEmbeddedFile ef = new PDEmbeddedFile(doc, fakeFile );
            //now lets some of the optional parameters
            ef.setSubtype( "test/plain" );
            ef.setSize( data.length );
            ef.setCreationDate( new GregorianCalendar() );
            fs.setEmbeddedFile( ef );

            // create a new tree node and add the embedded file
            PDEmbeddedFilesNameTreeNode treeNode = new PDEmbeddedFilesNameTreeNode();
            treeNode.setNames( Collections.singletonMap( "My first attachment",  fs ) );
View Full Code Here
34
35
36
37
38
39
40
41
42
43
44
public class TestEmbeddedFiles extends TestCase
{
    @Test
    public void testNullEmbeddedFile() throws IOException
    {
        PDEmbeddedFile embeddedFile = null;
        boolean ok = false;
        try
        {
            PDDocument doc = PDDocument.load(TestEmbeddedFiles.class.getResourceAsStream(
                "null_PDComplexFileSpecification.pdf"));
View Full Code Here
72
73
74
75
76
77
78
79
80
81
82
83
84
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
    }

    @Test
    public void testOSSpecificAttachments() throws IOException
    {
        PDEmbeddedFile nonOSFile = null;
        PDEmbeddedFile macFile = null;
        PDEmbeddedFile dosFile = null;
        PDEmbeddedFile unixFile = null;

        PDDocument doc = PDDocument.load(TestEmbeddedFiles.class
                .getResourceAsStream("testPDF_multiFormatEmbFiles.pdf"));

        PDDocumentCatalog catalog = doc.getDocumentCatalog();
        PDDocumentNameDictionary names = catalog.getNames();
        PDEmbeddedFilesNameTreeNode treeNode = names.getEmbeddedFiles();
        List<PDNameTreeNode> kids = treeNode.getKids();
        for (PDNameTreeNode kid : kids)
        {
            Map<String, COSObjectable> tmpNames = kid.getNames();
            COSObjectable obj = tmpNames.get("My first attachment");
            if (obj instanceof PDComplexFileSpecification)
            {
                PDComplexFileSpecification spec = (PDComplexFileSpecification) obj;
                nonOSFile = spec.getEmbeddedFile();
                macFile = spec.getEmbeddedFileMac();
                dosFile = spec.getEmbeddedFileDos();
                unixFile = spec.getEmbeddedFileUnix();
            }
        }

        assertTrue("non os specific",
                byteArrayContainsLC("non os specific", nonOSFile.getByteArray(), "ISO-8859-1"));

        assertTrue("mac", byteArrayContainsLC("mac embedded", macFile.getByteArray(), "ISO-8859-1"));

        assertTrue("dos", byteArrayContainsLC("dos embedded", dosFile.getByteArray(), "ISO-8859-1"));

        assertTrue("unix",
                byteArrayContainsLC("unix embedded", unixFile.getByteArray(), "ISO-8859-1"));

    }
View Full Code Here
111
112
113
114
115
116
117
118
119
120
121
                    {
                        if (annotation instanceof PDAnnotationFileAttachment)
                        {
                            PDAnnotationFileAttachment annotationFileAttachment = (PDAnnotationFileAttachment) annotation;
                            PDComplexFileSpecification fileSpec = (PDComplexFileSpecification) annotationFileAttachment.getFile();
                            PDEmbeddedFile embeddedFile = getEmbeddedFile(fileSpec);
                            extractFile(filePath, fileSpec.getFilename(), embeddedFile);
                        }
                    }
                }
               
View Full Code Here
134
135
136
137
138
139
140
141
142
            throws IOException
    {
        for (String filename : names.keySet())
        {
            PDComplexFileSpecification fileSpec = (PDComplexFileSpecification)names.get(filename);
            PDEmbeddedFile embeddedFile = getEmbeddedFile(fileSpec);
            extractFile(filePath, filename, embeddedFile);
        }
    }
View Full Code Here
153
154
155
156
157
158
159
160
161
162
163
    }
   
    private static PDEmbeddedFile getEmbeddedFile(PDComplexFileSpecification fileSpec )
    {
        // search for the first available alternative of the embedded file
        PDEmbeddedFile embeddedFile = null;
        if (fileSpec != null)
        {
            embeddedFile = fileSpec.getEmbeddedFileUnicode();
            if (embeddedFile == null)
            {
View Full Code Here
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
                                if (debug)
                                {
                                    System.err.println("Processing embedded file " + ent.getKey() + ":");
                                }
                                PDComplexFileSpecification spec = (PDComplexFileSpecification) ent.getValue();
                                PDEmbeddedFile file = spec.getEmbeddedFile();
                                if (file.getSubtype().equals("application/pdf"))
                                {
                                    if (debug)
                                    {
                                        System.err.println("  is PDF (size=" + file.getSize() + ")");
                                    }
                                    InputStream fis = file.createInputStream();
                                    PDDocument subDoc = null;
                                    try
                                    {
                                        subDoc = PDDocument.load(fis);
                                    }
View Full Code Here
TOP

Related Classes of org.apache.pdfbox.pdmodel.common.filespecification.PDEmbeddedFile

  • de.pdf_scrutinizer.document.DocumentAdapter
  • org.apache.pdfbox.cos.COSDictionary
  • org.apache.pdfbox.encryption.TestSymmetricKeyEncryption
  • org.apache.pdfbox.examples.pdmodel.EmbeddedFiles
  • org.apache.pdfbox.examples.pdmodel.ExtractEmbeddedFiles
  • org.apache.pdfbox.ExtractText
  • org.apache.pdfbox.pdmodel.common.TestEmbeddedFiles
  • org.apache.pdfbox.tools.ExtractText
  • org.apache.tika.parser.pdf.PDF2XHTML
  • org.apache.tika.parser.pdf.PDFParser
    • 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.