Examples with EndDocumentShieldingContentHandler org.apache.tika.sax.EndDocumentShieldingContentHandler used on opensource projects

Example 1 with EndDocumentShieldingContentHandler

use of org.apache.tika.sax.EndDocumentShieldingContentHandler in project tika by apache.

the class OpenDocumentParser method parse.

public void parse(InputStream stream, ContentHandler baseHandler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    // Open the Zip stream
    // Use a File if we can, and an already open zip is even better
    ZipFile zipFile = null;
    ZipInputStream zipStream = null;
    if (stream instanceof TikaInputStream) {
        TikaInputStream tis = (TikaInputStream) stream;
        Object container = ((TikaInputStream) stream).getOpenContainer();
        if (container instanceof ZipFile) {
            zipFile = (ZipFile) container;
        } else if (tis.hasFile()) {
            zipFile = new ZipFile(tis.getFile());
        } else {
            zipStream = new ZipInputStream(stream);
        }
    } else {
        zipStream = new ZipInputStream(stream);
    }
    // Prepare to handle the content
    XHTMLContentHandler xhtml = new XHTMLContentHandler(baseHandler, metadata);
    // As we don't know which of the metadata or the content
    //  we'll hit first, catch the endDocument call initially
    EndDocumentShieldingContentHandler handler = new EndDocumentShieldingContentHandler(xhtml);
    if (zipFile != null) {
        try {
            handleZipFile(zipFile, metadata, context, handler);
        } finally {
            //Do we want to close silently == catch an exception here?
            zipFile.close();
        }
    } else {
        try {
            handleZipStream(zipStream, metadata, context, handler);
        } finally {
            //Do we want to close silently == catch an exception here?
            zipStream.close();
        }
    }
    // Only now call the end document
    if (handler.getEndDocumentWasCalled()) {
        handler.reallyEndDocument();
    }
}