286N/Apackage com.sun.org.apache.xml.internal.serializer;

286N/A

286N/Aimport java.io.InputStream;

286N/Aimport java.io.OutputStream;

286N/Aimport java.io.OutputStreamWriter;

286N/Aimport java.io.UnsupportedEncodingException;

286N/Aimport java.io.Writer;

286N/Aimport java.io.BufferedWriter;

286N/Aimport java.net.URL;

286N/Aimport java.util.Enumeration;

286N/Aimport java.util.HashMap;

286N/Aimport java.util.Properties;

286N/Aimport java.util.StringTokenizer;

566N/Aimport java.io.IOException;

566N/Aimport java.net.MalformedURLException;

566N/Aimport java.nio.charset.Charset;

566N/Aimport java.nio.charset.IllegalCharsetNameException;

566N/Aimport java.nio.charset.UnsupportedCharsetException;

566N/Aimport java.util.Collections;

566N/Aimport java.util.Map;

566N/Aimport java.util.Map.Entry;

286N/A

286N/Aimport com.sun.org.apache.xalan.internal.utils.SecuritySupport;

286N/A

286N/A

286N/Apublic final class Encodings extends Object

286N/A{

286N/A

286N/A /**

286N/A private static final int m_defaultLastPrintable = 0x7F;

286N/A

286N/A /**

286N/A private static final String ENCODINGS_FILE = "com/sun/org/apache/xml/internal/serializer/Encodings.properties";

286N/A

286N/A /**

286N/A private static final String ENCODINGS_PROP = "com.sun.org.apache.xalan.internal.serialize.encodings";

286N/A

286N/A

286N/A /**

286N/A static Writer getWriter(OutputStream output, String encoding)

286N/A throws UnsupportedEncodingException

286N/A {

286N/A

566N/A final EncodingInfo ei = _encodingInfos.findEncoding(toUpperCaseFast(encoding));

566N/A if (ei != null) {

566N/A try {

566N/A return new BufferedWriter(new OutputStreamWriter(

566N/A output, ei.javaName));

566N/A } catch (UnsupportedEncodingException usee) {

566N/A // keep trying

286N/A }

286N/A }

286N/A

566N/A return new BufferedWriter(new OutputStreamWriter(output, encoding));

286N/A }

286N/A

286N/A

286N/A /**

286N/A public static int getLastPrintable()

286N/A {

286N/A return m_defaultLastPrintable;

286N/A }

286N/A

286N/A

286N/A

286N/A /**

286N/A static EncodingInfo getEncodingInfo(String encoding)

286N/A {

286N/A EncodingInfo ei;

286N/A

286N/A String normalizedEncoding = toUpperCaseFast(encoding);

566N/A ei = _encodingInfos.findEncoding(normalizedEncoding);

286N/A if (ei == null) {

286N/A // We shouldn't have to do this, but just in case.

566N/A try {

566N/A // This may happen if the caller tries to use

566N/A // an encoding that wasn't registered in the

566N/A // (java name)->(preferred mime name) mapping file.

566N/A // In that case we attempt to load the charset for the

566N/A // given encoding, and if that succeeds - we create a new

566N/A // EncodingInfo instance - assuming the canonical name

566N/A // of the charset can be used as the mime name.

566N/A final Charset c = Charset.forName(encoding);

566N/A final String name = c.name();

566N/A ei = new EncodingInfo(name, name);

566N/A _encodingInfos.putEncoding(normalizedEncoding, ei);

566N/A } catch (IllegalCharsetNameException | UnsupportedCharsetException x) {

566N/A ei = new EncodingInfo(null,null);

566N/A }

286N/A }

286N/A

286N/A return ei;

286N/A }

286N/A

286N/A /**

286N/A static private String toUpperCaseFast(final String s) {

286N/A

286N/A boolean different = false;

286N/A final int mx = s.length();

286N/A char[] chars = new char[mx];

286N/A for (int i=0; i < mx; i++) {

286N/A char ch = s.charAt(i);

286N/A // is the character a lower case ASCII one?

286N/A if ('a' <= ch && ch <= 'z') {

286N/A // a cheap and fast way to uppercase that is good enough

286N/A ch = (char) (ch + ('A' - 'a'));

286N/A different = true; // the uppercased String is different

286N/A }

286N/A chars[i] = ch;

286N/A }

286N/A

286N/A // A little optimization, don't call String.valueOf() if

286N/A // the uppercased string is the same as the input string.

286N/A final String upper;

286N/A if (different)

286N/A upper = String.valueOf(chars);

286N/A else

286N/A upper = s;

286N/A

286N/A return upper;

286N/A }

286N/A

286N/A /** The default encoding, ISO style, ISO style. */

286N/A static final String DEFAULT_MIME_ENCODING = "UTF-8";

286N/A

286N/A /**

286N/A static String getMimeEncoding(String encoding)

286N/A {

286N/A

286N/A if (null == encoding)

286N/A {

286N/A try

286N/A {

286N/A

286N/A // Get the default system character encoding. This may be

286N/A // incorrect if they passed in a writer, but right now there

286N/A // seems to be no way to get the encoding from a writer.

524N/A encoding = SecuritySupport.getSystemProperty("file.encoding", "UTF8");

286N/A

286N/A if (null != encoding)

286N/A {

286N/A

286N/A /*

286N/A String jencoding =

286N/A (encoding.equalsIgnoreCase("Cp1252")

286N/A || encoding.equalsIgnoreCase("ISO8859_1")

286N/A || encoding.equalsIgnoreCase("8859_1")

286N/A || encoding.equalsIgnoreCase("UTF8"))

286N/A ? DEFAULT_MIME_ENCODING

286N/A : convertJava2MimeEncoding(encoding);

286N/A

286N/A encoding =

286N/A (null != jencoding) ? jencoding : DEFAULT_MIME_ENCODING;

286N/A }

286N/A else

286N/A {

286N/A encoding = DEFAULT_MIME_ENCODING;

286N/A }

286N/A }

286N/A catch (SecurityException se)

286N/A {

286N/A encoding = DEFAULT_MIME_ENCODING;

286N/A }

286N/A }

286N/A else

286N/A {

286N/A encoding = convertJava2MimeEncoding(encoding);

286N/A }

286N/A

286N/A return encoding;

286N/A }

286N/A

286N/A /**

286N/A private static String convertJava2MimeEncoding(String encoding)

286N/A {

566N/A final EncodingInfo enc =

566N/A _encodingInfos.getEncodingFromJavaKey(toUpperCaseFast(encoding));

286N/A if (null != enc)

286N/A return enc.name;

286N/A return encoding;

286N/A }

286N/A

286N/A /**

286N/A public static String convertMime2JavaEncoding(String encoding)

286N/A {

566N/A final EncodingInfo info = _encodingInfos.findEncoding(toUpperCaseFast(encoding));

566N/A return info != null ? info.javaName : encoding;

286N/A }

286N/A

566N/A // Using an inner static class here prevent initialization races

566N/A // where the hash maps could be used before they were populated.

566N/A //

566N/A private final static class EncodingInfos {

566N/A // These maps are final and not modified after initialization.

566N/A private final Map<String, EncodingInfo> _encodingTableKeyJava = new HashMap<>();

566N/A private final Map<String, EncodingInfo> _encodingTableKeyMime = new HashMap<>();

566N/A // This map will be added to after initialization: make sure it's

566N/A // thread-safe. This map should not be used frequently - only in cases

566N/A // where the mapping requested was not declared in the Encodings.properties

566N/A // file.

566N/A private final Map<String, EncodingInfo> _encodingDynamicTable =

566N/A Collections.synchronizedMap(new HashMap<String, EncodingInfo>());

566N/A

566N/A private EncodingInfos() {

566N/A loadEncodingInfo();

566N/A }

566N/A

566N/A // Opens the file/resource containing java charset name -> preferred mime

566N/A // name mapping and returns it as an InputStream.

566N/A private InputStream openEncodingsFileStream() throws MalformedURLException, IOException {

286N/A String urlString = null;

286N/A InputStream is = null;

286N/A

566N/A try {

524N/A urlString = SecuritySupport.getSystemProperty(ENCODINGS_PROP, "");

566N/A } catch (SecurityException e) {

286N/A }

286N/A

286N/A if (urlString != null && urlString.length() > 0) {

286N/A URL url = new URL(urlString);

286N/A is = url.openStream();

286N/A }

286N/A

286N/A if (is == null) {

286N/A is = SecuritySupport.getResourceAsStream(ENCODINGS_FILE);

286N/A }

566N/A return is;

566N/A }

286N/A

566N/A // Loads the Properties resource containing the mapping:

566N/A // java charset name -> preferred mime name

566N/A // and returns it.

566N/A private Properties loadProperties() throws MalformedURLException, IOException {

286N/A Properties props = new Properties();

566N/A final InputStream is = openEncodingsFileStream();

566N/A try {

566N/A if (is != null) {

566N/A props.load(is);

566N/A } else {

566N/A // Seems to be no real need to force failure here, let the

566N/A // system do its best... The issue is not really very critical,

566N/A // and the output will be in any case _correct_ though maybe not

566N/A // always human-friendly... :)

566N/A // But maybe report/log the resource problem?

566N/A // Any standard ways to report/log errors (in static context)?

566N/A }

566N/A } finally {

566N/A if (is != null) {

566N/A is.close();

566N/A }

286N/A }

566N/A return props;

566N/A }

286N/A

566N/A // Parses the mime list associated to a java charset name.

566N/A // The first mime name in the list is supposed to be the preferred

566N/A // mime name.

566N/A private String[] parseMimeTypes(String val) {

566N/A int pos = val.indexOf(' ');

566N/A //int lastPrintable;

566N/A if (pos < 0) {

566N/A // Maybe report/log this problem?

566N/A // "Last printable character not defined for encoding " +

566N/A // mimeName + " (" + val + ")" ...

566N/A return new String[] { val };

566N/A //lastPrintable = 0x00FF;

566N/A }

566N/A //lastPrintable =

566N/A // Integer.decode(val.substring(pos).trim()).intValue();

566N/A StringTokenizer st =

566N/A new StringTokenizer(val.substring(0, pos), ",");

566N/A String[] values = new String[st.countTokens()];

566N/A for (int i=0; st.hasMoreTokens(); i++) {

566N/A values[i] = st.nextToken();

566N/A }

566N/A return values;

566N/A }

566N/A

566N/A // This method here attempts to find the canonical charset name for the

566N/A // the given name - which is supposed to be either a java name or a mime

566N/A // name.

566N/A // For that, it attempts to load the charset using the given name, and

566N/A // then returns the charset's canonical name.

566N/A // If the charset could not be loaded from the given name,

566N/A // the method returns null.

566N/A private String findCharsetNameFor(String name) {

566N/A try {

566N/A return Charset.forName(name).name();

566N/A } catch (Exception x) {

566N/A return null;

286N/A }

566N/A }

566N/A

566N/A // This method here attempts to find the canonical charset name for the

566N/A // the set javaName+mimeNames - which are supposed to all refer to the

566N/A // same charset.

566N/A // For that it attempts to load the charset using the javaName, and if

566N/A // not found, attempts again using each of the mime names in turn.

566N/A // If the charset could be loaded from the javaName, then the javaName

566N/A // itself is returned as charset name. Otherwise, each of the mime names

566N/A // is tried in turn, until a charset can be loaded from one of the names,

566N/A // and the loaded charset's canonical name is returned.

566N/A // If no charset can be loaded from either the javaName or one of the

566N/A // mime names, then null is returned.

566N/A //

566N/A // Note that the returned name is the 'java' name that will be used in

566N/A // instances of EncodingInfo.

566N/A // This is important because EncodingInfo uses that 'java name' later on

566N/A // in calls to String.getBytes(javaName).

566N/A // As it happens, sometimes only one element of the set mime names/javaName

566N/A // is known by Charset: sometimes only one of the mime names is known,

566N/A // sometime only the javaName is known, sometimes all are known.

566N/A //

566N/A // By using this method here, we fix the problem where one of the mime

566N/A // names is known but the javaName is unknown, by associating the charset

566N/A // loaded from one of the mime names with the unrecognized javaName.

566N/A //

566N/A // When none of the mime names or javaName are known - there's not much we can

566N/A // do... It can mean that this encoding is not supported for this

566N/A // OS. If such a charset is ever use it will result in having all characters

566N/A // escaped.

566N/A //

566N/A private String findCharsetNameFor(String javaName, String[] mimes) {

566N/A String cs = findCharsetNameFor(javaName);

566N/A if (cs != null) return javaName;

566N/A for (String m : mimes) {

566N/A cs = findCharsetNameFor(m);

566N/A if (cs != null) break;

566N/A }

566N/A return cs;

566N/A }

566N/A

566N/A /**

566N/A private void loadEncodingInfo() {

566N/A try {

566N/A // load (java name)->(preferred mime name) mapping.

566N/A final Properties props = loadProperties();

566N/A

566N/A // create instances of EncodingInfo from the loaded mapping

566N/A Enumeration keys = props.keys();

566N/A Map<String, EncodingInfo> canonicals = new HashMap<>();

566N/A while (keys.hasMoreElements()) {

566N/A final String javaName = (String) keys.nextElement();

566N/A final String[] mimes = parseMimeTypes(props.getProperty(javaName));

566N/A

566N/A final String charsetName = findCharsetNameFor(javaName, mimes);

566N/A if (charsetName != null) {

566N/A final String kj = toUpperCaseFast(javaName);

566N/A final String kc = toUpperCaseFast(charsetName);

566N/A for (int i = 0; i < mimes.length; ++i) {

566N/A final String mimeName = mimes[i];

566N/A final String km = toUpperCaseFast(mimeName);

566N/A EncodingInfo info = new EncodingInfo(mimeName, charsetName);

566N/A _encodingTableKeyMime.put(km, info);

566N/A if (!canonicals.containsKey(kc)) {

566N/A // canonicals will map the charset name to

566N/A // the info containing the prefered mime name

566N/A // (the preferred mime name is the first mime

566N/A // name in the list).

566N/A canonicals.put(kc, info);

566N/A _encodingTableKeyJava.put(kc, info);

566N/A }

566N/A _encodingTableKeyJava.put(kj, info);

566N/A }

566N/A } else {

566N/A // None of the java or mime names on the line were

566N/A // recognized => this charset is not supported?

286N/A }

286N/A }

566N/A

566N/A // Fix up the _encodingTableKeyJava so that the info mapped to

566N/A // the java name contains the preferred mime name.

566N/A // (a given java name can correspond to several mime name,

566N/A // but we want the _encodingTableKeyJava to point to the

566N/A // preferred mime name).

566N/A for (Entry<String, EncodingInfo> e : _encodingTableKeyJava.entrySet()) {

566N/A e.setValue(canonicals.get(toUpperCaseFast(e.getValue().javaName)));

566N/A }

566N/A

566N/A } catch (java.net.MalformedURLException mue) {

566N/A throw new com.sun.org.apache.xml.internal.serializer.utils.WrappedRuntimeException(mue);

566N/A } catch (java.io.IOException ioe) {

566N/A throw new com.sun.org.apache.xml.internal.serializer.utils.WrappedRuntimeException(ioe);

286N/A }

286N/A }

566N/A

566N/A EncodingInfo findEncoding(String normalizedEncoding) {

566N/A EncodingInfo info = _encodingTableKeyJava.get(normalizedEncoding);

566N/A if (info == null) {

566N/A info = _encodingTableKeyMime.get(normalizedEncoding);

566N/A }

566N/A if (info == null) {

566N/A info = _encodingDynamicTable.get(normalizedEncoding);

566N/A }

566N/A return info;

286N/A }

566N/A

566N/A EncodingInfo getEncodingFromMimeKey(String normalizedMimeName) {

566N/A return _encodingTableKeyMime.get(normalizedMimeName);

566N/A }

566N/A

566N/A EncodingInfo getEncodingFromJavaKey(String normalizedJavaName) {

566N/A return _encodingTableKeyJava.get(normalizedJavaName);

566N/A }

566N/A

566N/A void putEncoding(String key, EncodingInfo info) {

566N/A _encodingDynamicTable.put(key, info);

286N/A }

286N/A }

286N/A

286N/A /**

286N/A static boolean isHighUTF16Surrogate(char ch) {

286N/A return ('\uD800' <= ch && ch <= '\uDBFF');

286N/A }

286N/A /**

286N/A static boolean isLowUTF16Surrogate(char ch) {

286N/A return ('\uDC00' <= ch && ch <= '\uDFFF');

286N/A }

286N/A /**

286N/A static int toCodePoint(char highSurrogate, char lowSurrogate) {

286N/A int codePoint =

286N/A ((highSurrogate - 0xd800) << 10)

286N/A + (lowSurrogate - 0xdc00)

286N/A + 0x10000;

286N/A return codePoint;

286N/A }

286N/A /**

286N/A static int toCodePoint(char ch) {

286N/A int codePoint = ch;

286N/A return codePoint;

286N/A }

286N/A

566N/A private final static EncodingInfos _encodingInfos = new EncodingInfos();

566N/A

286N/A}