Examples with TemporaryResources org.apache.tika.io.TemporaryResources used on opensource projects

Example 1 with TemporaryResources

use of org.apache.tika.io.TemporaryResources in project tika by apache.

the class ExternalParser method parse.

/**
     * Executes the configured external command and passes the given document
     * stream as a simple XHTML document to the given SAX content handler.
     * Metadata is only extracted if {@link #setMetadataExtractionPatterns(Map)}
     * has been called to set patterns.
     */
public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
    TemporaryResources tmp = new TemporaryResources();
    try {
        parse(TikaInputStream.get(stream, tmp), xhtml, metadata, tmp);
    } finally {
        tmp.dispose();
    }
}

Example 2 with TemporaryResources

use of org.apache.tika.io.TemporaryResources in project tika by apache.

the class AutoDetectParser method parse.

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 = handler != null ? new SecureContentHandler(handler, tis) : null;
        //the caller hasn't specified one.
        if (context.get(EmbeddedDocumentExtractor.class) == null) {
            Parser p = context.get(Parser.class);
            if (p == null) {
                context.set(Parser.class, this);
            }
            context.set(EmbeddedDocumentExtractor.class, new ParsingEmbeddedDocumentExtractor(context));
        }
        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();
    }
}

Example 3 with TemporaryResources

use of org.apache.tika.io.TemporaryResources in project tika by apache.

the class DigestingParser method parse.

@Override
public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    TemporaryResources tmp = new TemporaryResources();
    TikaInputStream tis = TikaInputStream.get(stream, tmp);
    try {
        if (digester != null) {
            digester.digest(tis, metadata, context);
        }
        super.parse(tis, handler, metadata, context);
    } finally {
        tmp.dispose();
    }
}

Example 4 with TemporaryResources

use of org.apache.tika.io.TemporaryResources in project tika by apache.

the class ZipContainerDetector method detect.

public MediaType detect(InputStream input, Metadata metadata) throws IOException {
    // 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);
        // enough for all known formats
        byte[] prefix = new byte[1024];
        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
        }
    }
}

Example 5 with TemporaryResources

use of org.apache.tika.io.TemporaryResources in project tika by apache.

the class NetCDFParser method parse.

/*
     * (non-Javadoc)
     * 
     * @see org.apache.tika.parser.Parser#parse(java.io.InputStream,
     * org.xml.sax.ContentHandler, org.apache.tika.metadata.Metadata,
     * org.apache.tika.parser.ParseContext)
     */
public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    TemporaryResources tmp = TikaInputStream.isTikaInputStream(stream) ? null : new TemporaryResources();
    TikaInputStream tis = TikaInputStream.get(stream, tmp);
    NetcdfFile ncFile = null;
    try {
        ncFile = NetcdfFile.open(tis.getFile().getAbsolutePath());
        metadata.set("File-Type-Description", ncFile.getFileTypeDescription());
        // first parse out the set of global attributes
        for (Attribute attr : ncFile.getGlobalAttributes()) {
            Property property = resolveMetadataKey(attr.getFullName());
            if (attr.getDataType().isString()) {
                metadata.add(property, attr.getStringValue());
            } else if (attr.getDataType().isNumeric()) {
                int value = attr.getNumericValue().intValue();
                metadata.add(property, String.valueOf(value));
            }
        }
        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.newline();
        xhtml.element("h1", "dimensions");
        xhtml.startElement("ul");
        xhtml.newline();
        for (Dimension dim : ncFile.getDimensions()) {
            xhtml.element("li", dim.getFullName() + " = " + dim.getLength());
        }
        xhtml.endElement("ul");
        xhtml.element("h1", "variables");
        xhtml.startElement("ul");
        xhtml.newline();
        for (Variable var : ncFile.getVariables()) {
            xhtml.startElement("li");
            xhtml.characters(var.getDataType() + " " + var.getNameAndDimensions());
            xhtml.newline();
            List<Attribute> attributes = var.getAttributes();
            if (!attributes.isEmpty()) {
                xhtml.startElement("ul");
                for (Attribute element : attributes) {
                    xhtml.element("li", element.toString());
                }
                xhtml.endElement("ul");
            }
            xhtml.endElement("li");
        }
        xhtml.endElement("ul");
        xhtml.endDocument();
    } catch (IOException e) {
        throw new TikaException("NetCDF parse error", e);
    } finally {
        if (ncFile != null) {
            ncFile.close();
        }
        if (tmp != null) {
            tmp.dispose();
        }
    }
}