Examples with EncryptedDocumentException org.apache.tika.exception.EncryptedDocumentException used on opensource projects

Example 11 with EncryptedDocumentException

use of org.apache.tika.exception.EncryptedDocumentException in project tika by apache.

the class SXSLFExtractorTest method testEncrypted.

@Test
public void testEncrypted() throws Exception {
    Map<String, String> tests = new HashMap<String, String>();
    tests.put("testPPT_protected_passtika.pptx", "This is an encrypted PowerPoint 2007 slide.");
    Parser parser = new AutoDetectParser();
    Metadata m = new Metadata();
    PasswordProvider passwordProvider = new PasswordProvider() {

        @Override
        public String getPassword(Metadata metadata) {
            return "tika";
        }
    };
    ParseContext passwordContext = new ParseContext();
    passwordContext.set(org.apache.tika.parser.PasswordProvider.class, passwordProvider);
    passwordContext.set(OfficeParserConfig.class, officeParserConfig);
    for (Map.Entry<String, String> e : tests.entrySet()) {
        try (InputStream is = getResourceAsStream("/test-documents/" + e.getKey())) {
            ContentHandler handler = new BodyContentHandler();
            parser.parse(is, handler, m, passwordContext);
            assertContains(e.getValue(), handler.toString());
        }
    }
    ParseContext context = new ParseContext();
    //now try with no password
    for (Map.Entry<String, String> e : tests.entrySet()) {
        boolean exc = false;
        try (InputStream is = getResourceAsStream("/test-documents/" + e.getKey())) {
            ContentHandler handler = new BodyContentHandler();
            parser.parse(is, handler, m, context);
        } catch (EncryptedDocumentException ex) {
            exc = true;
        }
        assertTrue(exc);
    }
}

Example 12 with EncryptedDocumentException

use of org.apache.tika.exception.EncryptedDocumentException in project tika by apache.

the class PackageParser method parse.

public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    //lazily load the MediaTypeRegistry at parse time
    //only want to call getDefaultConfig() once, and can't
    //load statically because of the ForkParser
    TikaConfig config = context.get(TikaConfig.class);
    MediaTypeRegistry mediaTypeRegistry = null;
    if (config != null) {
        mediaTypeRegistry = config.getMediaTypeRegistry();
    } else {
        if (bufferedMediaTypeRegistry == null) {
            //buffer this for next time.
            synchronized (lock) {
                //now that we're locked, check again
                if (bufferedMediaTypeRegistry == null) {
                    bufferedMediaTypeRegistry = TikaConfig.getDefaultConfig().getMediaTypeRegistry();
                }
            }
        }
        mediaTypeRegistry = bufferedMediaTypeRegistry;
    }
    // Ensure that the stream supports the mark feature
    if (!stream.markSupported()) {
        stream = new BufferedInputStream(stream);
    }
    TemporaryResources tmp = new TemporaryResources();
    ArchiveInputStream ais = null;
    try {
        ArchiveStreamFactory factory = context.get(ArchiveStreamFactory.class, new ArchiveStreamFactory());
        // At the end we want to close the archive stream to release
        // any associated resources, but the underlying document stream
        // should not be closed
        ais = factory.createArchiveInputStream(new CloseShieldInputStream(stream));
    } catch (StreamingNotSupportedException sne) {
        // Most archive formats work on streams, but a few need files
        if (sne.getFormat().equals(ArchiveStreamFactory.SEVEN_Z)) {
            // Rework as a file, and wrap
            stream.reset();
            TikaInputStream tstream = TikaInputStream.get(stream, tmp);
            // Seven Zip suports passwords, was one given?
            String password = null;
            PasswordProvider provider = context.get(PasswordProvider.class);
            if (provider != null) {
                password = provider.getPassword(metadata);
            }
            SevenZFile sevenz;
            if (password == null) {
                sevenz = new SevenZFile(tstream.getFile());
            } else {
                sevenz = new SevenZFile(tstream.getFile(), password.getBytes("UnicodeLittleUnmarked"));
            }
            // Pending a fix for COMPRESS-269 / TIKA-1525, this bit is a little nasty
            ais = new SevenZWrapper(sevenz);
        } else {
            tmp.close();
            throw new TikaException("Unknown non-streaming format " + sne.getFormat(), sne);
        }
    } catch (ArchiveException e) {
        tmp.close();
        throw new TikaException("Unable to unpack document stream", e);
    }
    updateMediaType(ais, mediaTypeRegistry, metadata);
    // Use the delegate parser to parse the contained document
    EmbeddedDocumentExtractor extractor = EmbeddedDocumentUtil.getEmbeddedDocumentExtractor(context);
    XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
    xhtml.startDocument();
    try {
        ArchiveEntry entry = ais.getNextEntry();
        while (entry != null) {
            if (!entry.isDirectory()) {
                parseEntry(ais, entry, extractor, metadata, xhtml);
            }
            entry = ais.getNextEntry();
        }
    } catch (UnsupportedZipFeatureException zfe) {
        // If it's an encrypted document of unknown password, report as such
        if (zfe.getFeature() == Feature.ENCRYPTION) {
            throw new EncryptedDocumentException(zfe);
        }
        // Otherwise throw the exception
        throw new TikaException("UnsupportedZipFeature", zfe);
    } catch (PasswordRequiredException pre) {
        throw new EncryptedDocumentException(pre);
    } finally {
        ais.close();
        tmp.close();
    }
    xhtml.endDocument();
}

Example 13 with EncryptedDocumentException

use of org.apache.tika.exception.EncryptedDocumentException in project tika by apache.

the class PackageParser method parseEntry.

private void parseEntry(ArchiveInputStream archive, ArchiveEntry entry, EmbeddedDocumentExtractor extractor, Metadata parentMetadata, XHTMLContentHandler xhtml) throws SAXException, IOException, TikaException {
    String name = entry.getName();
    if (archive.canReadEntryData(entry)) {
        // Fetch the metadata on the entry contained in the archive
        Metadata entrydata = handleEntryMetadata(name, null, entry.getLastModifiedDate(), entry.getSize(), xhtml);
        // Recurse into the entry if desired
        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 {
        name = (name == null) ? "" : name;
        if (entry instanceof ZipArchiveEntry) {
            boolean usesEncryption = ((ZipArchiveEntry) entry).getGeneralPurposeBit().usesEncryption();
            if (usesEncryption) {
                EmbeddedDocumentUtil.recordEmbeddedStreamException(new EncryptedDocumentException("stream (" + name + ") is encrypted"), parentMetadata);
            }
        } else {
            EmbeddedDocumentUtil.recordEmbeddedStreamException(new TikaException("Can't read archive stream (" + name + ")"), parentMetadata);
        }
        if (name.length() > 0) {
            xhtml.element("p", name);
        }
    }
}

Example 14 with EncryptedDocumentException

use of org.apache.tika.exception.EncryptedDocumentException in project tika by apache.

the class PDFParser method parse.

public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    PDFParserConfig localConfig = context.get(PDFParserConfig.class, defaultConfig);
    PDDocument pdfDocument = null;
    String password = "";
    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
        //TODO: make this configurable via MemoryUsageSetting
        TikaInputStream tstream = TikaInputStream.cast(stream);
        password = getPassword(metadata, context);
        if (tstream != null && tstream.hasFile()) {
            // File based -- send file directly to PDFBox
            pdfDocument = PDDocument.load(tstream.getPath().toFile(), password);
        } else {
            pdfDocument = PDDocument.load(new CloseShieldInputStream(stream), password);
        }
        metadata.set(PDF.IS_ENCRYPTED, Boolean.toString(pdfDocument.isEncrypted()));
        metadata.set(Metadata.CONTENT_TYPE, MEDIA_TYPE.toString());
        extractMetadata(pdfDocument, metadata, context);
        AccessChecker checker = localConfig.getAccessChecker();
        checker.check(metadata);
        if (handler != null) {
            if (shouldHandleXFAOnly(pdfDocument, localConfig)) {
                handleXFAOnly(pdfDocument, handler, metadata, context);
            } else if (localConfig.getOcrStrategy().equals(PDFParserConfig.OCR_STRATEGY.OCR_ONLY)) {
                metadata.add("X-Parsed-By", TesseractOCRParser.class.toString());
                OCR2XHTML.process(pdfDocument, handler, context, metadata, localConfig);
            } else {
                if (localConfig.getOcrStrategy().equals(PDFParserConfig.OCR_STRATEGY.OCR_AND_TEXT_EXTRACTION)) {
                    metadata.add("X-Parsed-By", TesseractOCRParser.class.toString());
                }
                PDF2XHTML.process(pdfDocument, handler, context, metadata, localConfig);
            }
        }
    } catch (InvalidPasswordException e) {
        metadata.set(PDF.IS_ENCRYPTED, "true");
        throw new EncryptedDocumentException(e);
    } finally {
        if (pdfDocument != null) {
            pdfDocument.close();
        }
    }
}

Example 15 with EncryptedDocumentException

use of org.apache.tika.exception.EncryptedDocumentException in project tika by apache.

the class PDFParserTest method testProtectedPDF.

/**
     * PDFs can be "protected" with the default password. This means
     * they're encrypted (potentially both text and metadata),
     * but we can decrypt them easily.
     */
@Test
public void testProtectedPDF() throws Exception {
    XMLResult r = getXML("testPDF_protected.pdf");
    Metadata metadata = r.metadata;
    assertEquals("true", metadata.get("pdf:encrypted"));
    assertEquals("application/pdf", metadata.get(Metadata.CONTENT_TYPE));
    assertEquals("The Bank of England", metadata.get(TikaCoreProperties.CREATOR));
    assertEquals("The Bank of England", metadata.get(Metadata.AUTHOR));
    assertEquals("Speeches by Andrew G Haldane", metadata.get(OfficeOpenXMLCore.SUBJECT));
    assertEquals("Speeches by Andrew G Haldane", metadata.get(Metadata.SUBJECT));
    assertEquals("Rethinking the Financial Network, Speech by Andrew G Haldane, Executive Director, Financial Stability delivered at the Financial Student Association, Amsterdam on 28 April 2009", metadata.get(TikaCoreProperties.TITLE));
    assertContains("RETHINKING THE FINANCIAL NETWORK", r.xml);
    assertContains("On 16 November 2002", r.xml);
    assertContains("In many important respects", r.xml);
    // Try again with an explicit empty password
    ParseContext context = new ParseContext();
    context.set(PasswordProvider.class, new PasswordProvider() {

        public String getPassword(Metadata metadata) {
            return "";
        }
    });
    r = getXML("testPDF_protected.pdf", context);
    metadata = r.metadata;
    assertEquals("true", metadata.get("pdf:encrypted"));
    assertEquals("application/pdf", metadata.get(Metadata.CONTENT_TYPE));
    assertEquals("The Bank of England", metadata.get(TikaCoreProperties.CREATOR));
    assertEquals("Speeches by Andrew G Haldane", metadata.get(OfficeOpenXMLCore.SUBJECT));
    assertEquals("Speeches by Andrew G Haldane", metadata.get(Metadata.SUBJECT));
    assertEquals("Rethinking the Financial Network, Speech by Andrew G Haldane, Executive Director, Financial Stability delivered at the Financial Student Association, Amsterdam on 28 April 2009", metadata.get(TikaCoreProperties.TITLE));
    assertContains("RETHINKING THE FINANCIAL NETWORK", r.xml);
    assertContains("On 16 November 2002", r.xml);
    assertContains("In many important respects", r.xml);
    //now test wrong password
    context.set(PasswordProvider.class, new PasswordProvider() {

        public String getPassword(Metadata metadata) {
            return "WRONG!!!!";
        }
    });
    boolean ex = false;
    ContentHandler handler = new BodyContentHandler();
    metadata = new Metadata();
    try (InputStream stream = PDFParserTest.class.getResourceAsStream("/test-documents/testPDF_protected.pdf")) {
        Parser parser = new AutoDetectParser();
        parser.parse(stream, handler, metadata, context);
    } catch (EncryptedDocumentException e) {
        ex = true;
    }
    assertTrue("encryption exception", ex);
    assertEquals("application/pdf", metadata.get(Metadata.CONTENT_TYPE));
    assertEquals("true", metadata.get("pdf:encrypted"));
    //pdf:encrypted, X-Parsed-By and Content-Type
    assertEquals("very little metadata should be parsed", 3, metadata.names().length);
    assertEquals(0, handler.toString().length());
}