XmlUtilities.java

/**
 * Extract the XML encoding from XML content, to help parse content correctly.
 */
package org.androiddaisyreader.model;

import java.io.DataInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.StringReader;
import java.util.regex.Pattern;

import org.xml.sax.EntityResolver;
import org.xml.sax.InputSource;

/**
 * Simple Utility class to extract the XML encoding from a text file.
 * 
 * @author Julian Harty
 */
public final class XmlUtilities {
    private static final int ENOUGH = 200;
    protected static final String XML_TRAILER = "\"?>";
    protected static final String EXTRACT_ENCODING_REGEX = ".*encoding=\"";
    protected static final String XML_FIRST_LINE_REGEX = "<\\?xml version=\"1\\.0\" encoding=\"(.*)\"?>";

    // Hide the constructor for this utility class.
    private XmlUtilities() {
    };

    /**
     * Helper method to extract the XML file encoding
     * 
     * @param line the first line of an XML file
     * @return The value of the encoding in lower-case.
     */
    protected static String extractEncoding(String line) {
        Pattern p = Pattern.compile(EXTRACT_ENCODING_REGEX);
        String matches[] = p.split(line);
        // We want the value after encoding="
        String value = matches[1];
        // We don't need anything after the first " after the value
        String cleanup[] = value.split("\"");
        String encoding = cleanup[0];
        return encoding.toLowerCase();
    }

    /**
     * Helper method to map an unsupported XML encoding to a similar encoding.
     * 
     * Currently limited to processing windows-1252 encoding.
     * 
     * @param encoding The encoding string e.g. "windows-1252"
     * @return a similar, hopefully supported encoding, where we have a suitable
     *         match, else the original encoding.
     */
    public static String mapUnsupportedEncoding(String encoding) {
        if (encoding.equalsIgnoreCase("windows-1252")) {
            return "iso-8859-1";
        }
        return encoding;
    }

    /**
     * Helper method to obtain the content encoding from an input stream.
     * 
     * @param bis file to parse
     * @return the encoding if we are able to extract and parse it, else the
     *         default value expected by the expat parser, i.e. "UTF-8"
     * @throws {@link IOException} if there is a problem reading from the file.
     * @throws {@link IllegalArgumentException} if the InputStream does not
     *         support mark and reset.
     */
    public static String obtainEncodingStringFromInputStream(InputStream bis) throws IOException {
        String encoding = "UTF-8";
        if (!bis.markSupported()) {
            throw new IllegalArgumentException(
                    "Error in the program, InputStream needs to support markSupported()");
        }

        // read the first line after setting the mark, then reset
        // before calling the parser.
        bis.mark(ENOUGH);
        DataInputStream dis = new DataInputStream(bis);

        String line = dis.readLine();
        line = line.replace("'", "\"");
        if (line.matches(XML_FIRST_LINE_REGEX)) {
            encoding = extractEncoding(line);
        }
        bis.reset();
        return encoding;
    }

    /**
     * Provide a dummy XML resolver.
     * 
     * @return the dummy resolver.
     */
    public static EntityResolver dummyEntityResolver() {
        // Thanks to http://www.junlu.com/msg/202604.html
        EntityResolver er = new EntityResolver() {
            public InputSource resolveEntity(String publicId, String systemId) {
                return new InputSource(new StringReader(" "));
            }
        };
        return er;
    }
}