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

Example 6 with EncryptedDocumentException

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

the class ParsingEmbeddedDocumentExtractor method parseEmbedded.

public void parseEmbedded(InputStream stream, ContentHandler handler, Metadata metadata, boolean outputHtml) throws SAXException, IOException {
    if (outputHtml) {
        AttributesImpl attributes = new AttributesImpl();
        attributes.addAttribute("", "class", "class", "CDATA", "package-entry");
        handler.startElement(XHTML, "div", "div", attributes);
    }
    String name = metadata.get(Metadata.RESOURCE_NAME_KEY);
    if (name != null && name.length() > 0 && outputHtml) {
        handler.startElement(XHTML, "h1", "h1", new AttributesImpl());
        char[] chars = name.toCharArray();
        handler.characters(chars, 0, chars.length);
        handler.endElement(XHTML, "h1", "h1");
    }
    // Use the delegate parser to parse this entry
    try (TemporaryResources tmp = new TemporaryResources()) {
        final TikaInputStream newStream = TikaInputStream.get(new CloseShieldInputStream(stream), tmp);
        if (stream instanceof TikaInputStream) {
            final Object container = ((TikaInputStream) stream).getOpenContainer();
            if (container != null) {
                newStream.setOpenContainer(container);
            }
        }
        DELEGATING_PARSER.parse(newStream, new EmbeddedContentHandler(new BodyContentHandler(handler)), metadata, context);
    } catch (EncryptedDocumentException ede) {
    // TODO: can we log a warning that we lack the password?
    // For now, just skip the content
    } catch (TikaException e) {
    // TODO: can we log a warning somehow?
    // Could not parse the entry, just skip the content
    }
    if (outputHtml) {
        handler.endElement(XHTML, "div", "div");
    }
}

Example 7 with EncryptedDocumentException

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

the class JackcessParserTest method testPassword.

@Test
public void testPassword() throws Exception {
    ParseContext c = new ParseContext();
    c.set(PasswordProvider.class, new PasswordProvider() {

        @Override
        public String getPassword(Metadata metadata) {
            return "tika";
        }
    });
    Parser p = new AutoDetectParser();
    String content = null;
    try (InputStream is = this.getResourceAsStream("/test-documents/testAccess2_encrypted.accdb")) {
        content = getText(is, p, c);
    }
    assertContains("red and brown", content);
    //now try wrong password
    c.set(PasswordProvider.class, new PasswordProvider() {

        @Override
        public String getPassword(Metadata metadata) {
            return "WRONG";
        }
    });
    boolean ex = false;
    try (InputStream is = this.getResourceAsStream("/test-documents/testAccess2_encrypted.accdb")) {
        getText(is, p, c);
    } catch (EncryptedDocumentException e) {
        ex = true;
    }
    assertTrue("failed to throw encrypted document exception for wrong password", ex);
    //now try null
    c.set(PasswordProvider.class, new PasswordProvider() {

        @Override
        public String getPassword(Metadata metadata) {
            return null;
        }
    });
    ex = false;
    try (InputStream is = this.getResourceAsStream("/test-documents/testAccess2_encrypted.accdb")) {
        getText(is, p, c);
    } catch (EncryptedDocumentException e) {
        ex = true;
    }
    assertTrue("failed to throw encrypted document exception for null password", ex);
    //now try missing password provider
    c = new ParseContext();
    ex = false;
    try (InputStream is = this.getResourceAsStream("/test-documents/testAccess2_encrypted.accdb")) {
        getText(is, p, c);
    } catch (EncryptedDocumentException e) {
        ex = true;
    }
    assertTrue("failed to throw encrypted document exception for missing password provider", ex);
    //now try password on file that doesn't need a password
    c = new ParseContext();
    c.set(PasswordProvider.class, new PasswordProvider() {

        @Override
        public String getPassword(Metadata metadata) {
            return "tika";
        }
    });
    ex = false;
    try (InputStream is = this.getResourceAsStream("/test-documents/testAccess2.accdb")) {
        content = getText(is, p, c);
    } catch (EncryptedDocumentException e) {
        ex = true;
    }
    assertFalse("shouldn't have thrown encrypted document exception for " + "opening unencrypted file that doesn't need passowrd", ex);
    assertContains("red and brown", content);
}

Example 8 with EncryptedDocumentException

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

the class OfficeParser method parse.

protected void parse(DirectoryNode root, ParseContext context, Metadata metadata, XHTMLContentHandler xhtml) throws IOException, SAXException, TikaException {
    // Parse summary entries first, to make metadata available early
    new SummaryExtractor(metadata).parseSummaries(root);
    // Parse remaining document entries
    POIFSDocumentType type = POIFSDocumentType.detectType(root);
    if (type != POIFSDocumentType.UNKNOWN) {
        setType(metadata, type.getType());
    }
    switch(type) {
        case SOLIDWORKS_PART:
        case SOLIDWORKS_ASSEMBLY:
        case SOLIDWORKS_DRAWING:
            break;
        case PUBLISHER:
            PublisherTextExtractor publisherTextExtractor = new PublisherTextExtractor(root);
            xhtml.element("p", publisherTextExtractor.getText());
            break;
        case WORDDOCUMENT:
            new WordExtractor(context, metadata).parse(root, xhtml);
            break;
        case POWERPOINT:
            new HSLFExtractor(context, metadata).parse(root, xhtml);
            break;
        case WORKBOOK:
        case XLR:
            Locale locale = context.get(Locale.class, Locale.getDefault());
            new ExcelExtractor(context, metadata).parse(root, xhtml, locale);
            break;
        case PROJECT:
            // We currently can't do anything beyond the metadata
            break;
        case VISIO:
            VisioTextExtractor visioTextExtractor = new VisioTextExtractor(root);
            for (String text : visioTextExtractor.getAllText()) {
                xhtml.element("p", text);
            }
            break;
        case OUTLOOK:
            OutlookExtractor extractor = new OutlookExtractor(root, context);
            extractor.parse(xhtml, metadata);
            break;
        case ENCRYPTED:
            EncryptionInfo info = new EncryptionInfo(root);
            Decryptor d = Decryptor.getInstance(info);
            try {
                // By default, use the default Office Password
                String password = Decryptor.DEFAULT_PASSWORD;
                // If they supplied a Password Provider, ask that for the password,
                //  and use the provider given one if available (stick with default if not)
                PasswordProvider passwordProvider = context.get(PasswordProvider.class);
                if (passwordProvider != null) {
                    String suppliedPassword = passwordProvider.getPassword(metadata);
                    if (suppliedPassword != null) {
                        password = suppliedPassword;
                    }
                }
                // Check if we've the right password or not
                if (!d.verifyPassword(password)) {
                    throw new EncryptedDocumentException();
                }
                // Decrypt the OLE2 stream, and delegate the resulting OOXML
                //  file to the regular OOXML parser for normal handling
                OOXMLParser parser = new OOXMLParser();
                parser.parse(d.getDataStream(root), new EmbeddedContentHandler(new BodyContentHandler(xhtml)), metadata, context);
            } catch (GeneralSecurityException ex) {
                throw new EncryptedDocumentException(ex);
            }
        default:
            //  is extracted, which happened above
            break;
    }
}

Example 9 with EncryptedDocumentException

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

the class WordExtractor method parse.

protected void parse(DirectoryNode root, XHTMLContentHandler xhtml) throws IOException, SAXException, TikaException {
    HWPFDocument document;
    try {
        document = new HWPFDocument(root);
    } catch (org.apache.poi.EncryptedDocumentException e) {
        throw new EncryptedDocumentException(e);
    } catch (OldWordFileFormatException e) {
        parseWord6(root, xhtml);
        return;
    }
    extractSavedByMetadata(document);
    org.apache.poi.hwpf.extractor.WordExtractor wordExtractor = new org.apache.poi.hwpf.extractor.WordExtractor(document);
    HeaderStories headerFooter = new HeaderStories(document);
    // Grab the list of pictures. As far as we can tell,
    //  the pictures should be in order, and may be directly
    //  placed or referenced from an anchor
    PicturesTable pictureTable = document.getPicturesTable();
    PicturesSource pictures = new PicturesSource(document);
    // Do any headers, if present
    Range[] headers = new Range[] { headerFooter.getFirstHeaderSubrange(), headerFooter.getEvenHeaderSubrange(), headerFooter.getOddHeaderSubrange() };
    handleHeaderFooter(headers, "header", document, pictures, pictureTable, xhtml);
    // Do the main paragraph text
    Range r = document.getRange();
    ListManager listManager = new ListManager(document);
    for (int i = 0; i < r.numParagraphs(); i++) {
        Paragraph p = r.getParagraph(i);
        i += handleParagraph(p, 0, r, document, FieldsDocumentPart.MAIN, pictures, pictureTable, listManager, xhtml);
    }
    if (officeParserConfig.getIncludeShapeBasedContent()) {
        // Do everything else
        for (String paragraph : wordExtractor.getMainTextboxText()) {
            xhtml.element("p", paragraph);
        }
    }
    for (String paragraph : wordExtractor.getFootnoteText()) {
        xhtml.element("p", paragraph);
    }
    for (String paragraph : wordExtractor.getCommentsText()) {
        xhtml.element("p", paragraph);
    }
    for (String paragraph : wordExtractor.getEndnoteText()) {
        xhtml.element("p", paragraph);
    }
    // Do any footers, if present
    Range[] footers = new Range[] { headerFooter.getFirstFooterSubrange(), headerFooter.getEvenFooterSubrange(), headerFooter.getOddFooterSubrange() };
    handleHeaderFooter(footers, "footer", document, pictures, pictureTable, xhtml);
    // Handle any pictures that we haven't output yet
    for (Picture p = pictures.nextUnclaimed(); p != null; ) {
        handlePictureCharacterRun(null, p, pictures, xhtml);
        p = pictures.nextUnclaimed();
    }
    // Handle any embeded office documents
    try {
        DirectoryEntry op = (DirectoryEntry) root.getEntry("ObjectPool");
        for (Entry entry : op) {
            if (entry.getName().startsWith("_") && entry instanceof DirectoryEntry) {
                handleEmbeddedOfficeDoc((DirectoryEntry) entry, xhtml);
            }
        }
    } catch (FileNotFoundException e) {
    }
}

Example 10 with EncryptedDocumentException

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

the class CryptoParser method parse.

public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    try {
        Cipher cipher;
        if (provider != null) {
            cipher = Cipher.getInstance(transformation, provider);
        } else {
            cipher = Cipher.getInstance(transformation);
        }
        Key key = context.get(Key.class);
        if (key == null) {
            throw new EncryptedDocumentException("No decryption key provided");
        }
        AlgorithmParameters params = context.get(AlgorithmParameters.class);
        SecureRandom random = context.get(SecureRandom.class);
        if (params != null && random != null) {
            cipher.init(Cipher.DECRYPT_MODE, key, params, random);
        } else if (params != null) {
            cipher.init(Cipher.DECRYPT_MODE, key, params);
        } else if (random != null) {
            cipher.init(Cipher.DECRYPT_MODE, key, random);
        } else {
            cipher.init(Cipher.DECRYPT_MODE, key);
        }
        super.parse(new CipherInputStream(stream, cipher), handler, metadata, context);
    } catch (GeneralSecurityException e) {
        throw new TikaException("Unable to decrypt document stream", e);
    }
}