view src/share/jaxws_classes/com/sun/xml/internal/bind/DatatypeConverterImpl.java @ 809:cfc3c8286c3b

8025030: Enhance stream handling Summary: Avoiding caching data initialized via TCCL in static context; fix also reviewed by Iaroslav Savytskyi, Alexander Fomin Reviewed-by: ahgross, mgrebac, skoivu
author mkos
date Thu, 13 Feb 2014 18:20:10 +0100
parents 913b9e4d8c51
children 2cc5b85692ba
line wrap: on
line source

/*
 * Copyright (c) 1997, 2014, Oracle and/or its affiliates. All rights reserved.
 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
 *
 * This code is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 only, as
 * published by the Free Software Foundation.  Oracle designates this
 * particular file as subject to the "Classpath" exception as provided
 * by Oracle in the LICENSE file that accompanied this code.
 *
 * This code is distributed in the hope that it will be useful, but WITHOUT
 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
 * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
 * version 2 for more details (a copy is included in the LICENSE file that
 * accompanied this code).
 *
 * You should have received a copy of the GNU General Public License version
 * 2 along with this work; if not, write to the Free Software Foundation,
 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
 *
 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA
 * or visit www.oracle.com if you need additional information or have any
 * questions.
 */

package com.sun.xml.internal.bind;

import java.math.BigDecimal;
import java.math.BigInteger;
import java.security.AccessController;
import java.security.PrivilegedAction;
import java.util.Calendar;
import java.util.Collections;
import java.util.GregorianCalendar;
import java.util.Map;
import java.util.TimeZone;
import java.util.WeakHashMap;

import javax.xml.bind.DatatypeConverter;
import javax.xml.bind.DatatypeConverterInterface;
import javax.xml.datatype.DatatypeConfigurationException;
import javax.xml.datatype.DatatypeFactory;
import javax.xml.namespace.NamespaceContext;
import javax.xml.namespace.QName;
import javax.xml.stream.XMLStreamException;
import javax.xml.stream.XMLStreamWriter;

/**
 * This class is the JAXB RI's default implementation of the
 * {@link DatatypeConverterInterface}.
 *
 * <p>
 * When client applications specify the use of the static print/parse
 * methods in {@link DatatypeConverter}, it will delegate
 * to this class.
 *
 * <p>
 * This class is responsible for whitespace normalization.
 *
 * @author <ul><li>Ryan Shoemaker, Martin Grebac</li></ul>
 * @since JAXB1.0
 * @deprecated in JAXB 2.2.4 - use javax.xml.bind.DatatypeConverterImpl instead
 * or let us know why you can't
 */
public final class DatatypeConverterImpl {

    protected DatatypeConverterImpl() {
        // shall not be used
    }

    public static BigInteger _parseInteger(CharSequence s) {
        return new BigInteger(removeOptionalPlus(WhiteSpaceProcessor.trim(s)).toString());
    }

    public static String _printInteger(BigInteger val) {
        return val.toString();
    }

    /**
     * Faster but less robust String->int conversion.
     *
     * Note that:
     * <ol>
     *  <li>XML Schema allows '+', but {@link Integer#valueOf(String)} is not.
     *  <li>XML Schema allows leading and trailing (but not in-between) whitespaces.
     *      {@link Integer#valueOf(String)} doesn't allow any.
     * </ol>
     */
    public static int _parseInt(CharSequence s) {
        int len = s.length();
        int sign = 1;

        int r = 0;

        for (int i = 0; i < len; i++) {
            char ch = s.charAt(i);
            if (WhiteSpaceProcessor.isWhiteSpace(ch)) {
                // skip whitespace
            } else if ('0' <= ch && ch <= '9') {
                r = r * 10 + (ch - '0');
            } else if (ch == '-') {
                sign = -1;
            } else if (ch == '+') {
                // noop
            } else {
                throw new NumberFormatException("Not a number: " + s);
            }
        }

        return r * sign;
    }

    public static long _parseLong(CharSequence s) {
        return Long.valueOf(removeOptionalPlus(WhiteSpaceProcessor.trim(s)).toString());
    }

    public static short _parseShort(CharSequence s) {
        return (short) _parseInt(s);
    }

    public static String _printShort(short val) {
        return String.valueOf(val);
    }

    public static BigDecimal _parseDecimal(CharSequence content) {
        content = WhiteSpaceProcessor.trim(content);

        if (content.length() <= 0) {
            return null;
        }

        return new BigDecimal(content.toString());

        // from purely XML Schema perspective,
        // this implementation has a problem, since
        // in xs:decimal "1.0" and "1" is equal whereas the above
        // code will return different values for those two forms.
        //
        // the code was originally using com.sun.msv.datatype.xsd.NumberType.load,
        // but a profiling showed that the process of normalizing "1.0" into "1"
        // could take non-trivial time.
        //
        // also, from the user's point of view, one might be surprised if
        // 1 (not 1.0) is returned from "1.000"
    }

    public static float _parseFloat(CharSequence _val) {
        String s = WhiteSpaceProcessor.trim(_val).toString();
        /* Incompatibilities of XML Schema's float "xfloat" and Java's float "jfloat"

         * jfloat.valueOf ignores leading and trailing whitespaces,
        whereas this is not allowed in xfloat.
         * jfloat.valueOf allows "float type suffix" (f, F) to be
        appended after float literal (e.g., 1.52e-2f), whereare
        this is not the case of xfloat.

        gray zone
        ---------
         * jfloat allows ".523". And there is no clear statement that mentions
        this case in xfloat. Although probably this is allowed.
         *
         */

        if (s.equals("NaN")) {
            return Float.NaN;
        }
        if (s.equals("INF")) {
            return Float.POSITIVE_INFINITY;
        }
        if (s.equals("-INF")) {
            return Float.NEGATIVE_INFINITY;
        }

        if (s.length() == 0
                || !isDigitOrPeriodOrSign(s.charAt(0))
                || !isDigitOrPeriodOrSign(s.charAt(s.length() - 1))) {
            throw new NumberFormatException();
        }

        // these screening process is necessary due to the wobble of Float.valueOf method
        return Float.parseFloat(s);
    }

    public static String _printFloat(float v) {
        if (Float.isNaN(v)) {
            return "NaN";
        }
        if (v == Float.POSITIVE_INFINITY) {
            return "INF";
        }
        if (v == Float.NEGATIVE_INFINITY) {
            return "-INF";
        }
        return String.valueOf(v);
    }

    public static double _parseDouble(CharSequence _val) {
        String val = WhiteSpaceProcessor.trim(_val).toString();

        if (val.equals("NaN")) {
            return Double.NaN;
        }
        if (val.equals("INF")) {
            return Double.POSITIVE_INFINITY;
        }
        if (val.equals("-INF")) {
            return Double.NEGATIVE_INFINITY;
        }

        if (val.length() == 0
                || !isDigitOrPeriodOrSign(val.charAt(0))
                || !isDigitOrPeriodOrSign(val.charAt(val.length() - 1))) {
            throw new NumberFormatException(val);
        }


        // these screening process is necessary due to the wobble of Float.valueOf method
        return Double.parseDouble(val);
    }

    public static Boolean _parseBoolean(CharSequence literal) {
        if (literal == null) {
            return null;
        }

        int i = 0;
        int len = literal.length();
        char ch;
        boolean value = false;

        if (literal.length() <= 0) {
            return null;
        }

        do {
            ch = literal.charAt(i++);
        } while (WhiteSpaceProcessor.isWhiteSpace(ch) && i < len);

        int strIndex = 0;

        switch (ch) {
            case '1':
                value = true;
                break;
            case '0':
                value = false;
                break;
            case 't':
                String strTrue = "rue";
                do {
                    ch = literal.charAt(i++);
                } while ((strTrue.charAt(strIndex++) == ch) && i < len && strIndex < 3);

                if (strIndex == 3) {
                    value = true;
                } else {
                    return false;
                }
//                    throw new IllegalArgumentException("String \"" + literal + "\" is not valid boolean value.");

                break;
            case 'f':
                String strFalse = "alse";
                do {
                    ch = literal.charAt(i++);
                } while ((strFalse.charAt(strIndex++) == ch) && i < len && strIndex < 4);


                if (strIndex == 4) {
                    value = false;
                } else {
                    return false;
                }
//                    throw new IllegalArgumentException("String \"" + literal + "\" is not valid boolean value.");

                break;
        }

        if (i < len) {
            do {
                ch = literal.charAt(i++);
            } while (WhiteSpaceProcessor.isWhiteSpace(ch) && i < len);
        }

        if (i == len) {
            return value;
        } else {
            return null;
        }
//            throw new IllegalArgumentException("String \"" + literal + "\" is not valid boolean value.");
    }

    public static String _printBoolean(boolean val) {
        return val ? "true" : "false";
    }

    public static byte _parseByte(CharSequence literal) {
        return (byte) _parseInt(literal);
    }

    public static String _printByte(byte val) {
        return String.valueOf(val);
    }

    /**
     * @return null if fails to convert.
     */
    public static QName _parseQName(CharSequence text, NamespaceContext nsc) {
        int length = text.length();

        // trim whitespace
        int start = 0;
        while (start < length && WhiteSpaceProcessor.isWhiteSpace(text.charAt(start))) {
            start++;
        }

        int end = length;
        while (end > start && WhiteSpaceProcessor.isWhiteSpace(text.charAt(end - 1))) {
            end--;
        }

        if (end == start) {
            throw new IllegalArgumentException("input is empty");
        }


        String uri;
        String localPart;
        String prefix;

        // search ':'
        int idx = start + 1;    // no point in searching the first char. that's not valid.
        while (idx < end && text.charAt(idx) != ':') {
            idx++;
        }

        if (idx == end) {
            uri = nsc.getNamespaceURI("");
            localPart = text.subSequence(start, end).toString();
            prefix = "";
        } else {
            // Prefix exists, check everything
            prefix = text.subSequence(start, idx).toString();
            localPart = text.subSequence(idx + 1, end).toString();
            uri = nsc.getNamespaceURI(prefix);
            // uri can never be null according to javadoc,
            // but some users reported that there are implementations that return null.
            if (uri == null || uri.length() == 0) // crap. the NamespaceContext interface is broken.
            // error: unbound prefix
            {
                throw new IllegalArgumentException("prefix " + prefix + " is not bound to a namespace");
            }
        }

        return new QName(uri, localPart, prefix);
    }

    public static GregorianCalendar _parseDateTime(CharSequence s) {
        String val = WhiteSpaceProcessor.trim(s).toString();
        return getDatatypeFactory().newXMLGregorianCalendar(val).toGregorianCalendar();
    }

    public static String _printDateTime(Calendar val) {
        return CalendarFormatter.doFormat("%Y-%M-%DT%h:%m:%s%z", val);
    }

    public static String _printDate(Calendar val) {
        return CalendarFormatter.doFormat((new StringBuilder("%Y-%M-%D").append("%z")).toString(),val);
    }

    public static String _printInt(int val) {
        return String.valueOf(val);
    }

    public static String _printLong(long val) {
        return String.valueOf(val);
    }

    public static String _printDecimal(BigDecimal val) {
        return val.toPlainString();
    }

    public static String _printDouble(double v) {
        if (Double.isNaN(v)) {
            return "NaN";
        }
        if (v == Double.POSITIVE_INFINITY) {
            return "INF";
        }
        if (v == Double.NEGATIVE_INFINITY) {
            return "-INF";
        }
        return String.valueOf(v);
    }

    public static String _printQName(QName val, NamespaceContext nsc) {
        // Double-check
        String qname;
        String prefix = nsc.getPrefix(val.getNamespaceURI());
        String localPart = val.getLocalPart();

        if (prefix == null || prefix.length() == 0) { // be defensive
            qname = localPart;
        } else {
            qname = prefix + ':' + localPart;
        }

        return qname;
    }

// base64 decoder
    private static final byte[] decodeMap = initDecodeMap();
    private static final byte PADDING = 127;

    private static byte[] initDecodeMap() {
        byte[] map = new byte[128];
        int i;
        for (i = 0; i < 128; i++) {
            map[i] = -1;
        }

        for (i = 'A'; i <= 'Z'; i++) {
            map[i] = (byte) (i - 'A');
        }
        for (i = 'a'; i <= 'z'; i++) {
            map[i] = (byte) (i - 'a' + 26);
        }
        for (i = '0'; i <= '9'; i++) {
            map[i] = (byte) (i - '0' + 52);
        }
        map['+'] = 62;
        map['/'] = 63;
        map['='] = PADDING;

        return map;
    }

    /**
     * computes the length of binary data speculatively.
     *
     * <p>
     * Our requirement is to create byte[] of the exact length to store the binary data.
     * If we do this in a straight-forward way, it takes two passes over the data.
     * Experiments show that this is a non-trivial overhead (35% or so is spent on
     * the first pass in calculating the length.)
     *
     * <p>
     * So the approach here is that we compute the length speculatively, without looking
     * at the whole contents. The obtained speculative value is never less than the
     * actual length of the binary data, but it may be bigger. So if the speculation
     * goes wrong, we'll pay the cost of reallocation and buffer copying.
     *
     * <p>
     * If the base64 text is tightly packed with no indentation nor illegal char
     * (like what most web services produce), then the speculation of this method
     * will be correct, so we get the performance benefit.
     */
    private static int guessLength(String text) {
        final int len = text.length();

        // compute the tail '=' chars
        int j = len - 1;
        for (; j >= 0; j--) {
            byte code = decodeMap[text.charAt(j)];
            if (code == PADDING) {
                continue;
            }
            if (code == -1) // most likely this base64 text is indented. go with the upper bound
            {
                return text.length() / 4 * 3;
            }
            break;
        }

        j++;    // text.charAt(j) is now at some base64 char, so +1 to make it the size
        int padSize = len - j;
        if (padSize > 2) // something is wrong with base64. be safe and go with the upper bound
        {
            return text.length() / 4 * 3;
        }

        // so far this base64 looks like it's unindented tightly packed base64.
        // take a chance and create an array with the expected size
        return text.length() / 4 * 3 - padSize;
    }

    /**
     * @param text
     *      base64Binary data is likely to be long, and decoding requires
     *      each character to be accessed twice (once for counting length, another
     *      for decoding.)
     *
     *      A benchmark showed that taking {@link String} is faster, presumably
     *      because JIT can inline a lot of string access (with data of 1K chars, it was twice as fast)
     */
    public static byte[] _parseBase64Binary(String text) {
        final int buflen = guessLength(text);
        final byte[] out = new byte[buflen];
        int o = 0;

        final int len = text.length();
        int i;

        final byte[] quadruplet = new byte[4];
        int q = 0;

        // convert each quadruplet to three bytes.
        for (i = 0; i < len; i++) {
            char ch = text.charAt(i);
            byte v = decodeMap[ch];

            if (v != -1) {
                quadruplet[q++] = v;
            }

            if (q == 4) {
                // quadruplet is now filled.
                out[o++] = (byte) ((quadruplet[0] << 2) | (quadruplet[1] >> 4));
                if (quadruplet[2] != PADDING) {
                    out[o++] = (byte) ((quadruplet[1] << 4) | (quadruplet[2] >> 2));
                }
                if (quadruplet[3] != PADDING) {
                    out[o++] = (byte) ((quadruplet[2] << 6) | (quadruplet[3]));
                }
                q = 0;
            }
        }

        if (buflen == o) // speculation worked out to be OK
        {
            return out;
        }

        // we overestimated, so need to create a new buffer
        byte[] nb = new byte[o];
        System.arraycopy(out, 0, nb, 0, o);
        return nb;
    }
    private static final char[] encodeMap = initEncodeMap();

    private static char[] initEncodeMap() {
        char[] map = new char[64];
        int i;
        for (i = 0; i < 26; i++) {
            map[i] = (char) ('A' + i);
        }
        for (i = 26; i < 52; i++) {
            map[i] = (char) ('a' + (i - 26));
        }
        for (i = 52; i < 62; i++) {
            map[i] = (char) ('0' + (i - 52));
        }
        map[62] = '+';
        map[63] = '/';

        return map;
    }

    public static char encode(int i) {
        return encodeMap[i & 0x3F];
    }

    public static byte encodeByte(int i) {
        return (byte) encodeMap[i & 0x3F];
    }

    public static String _printBase64Binary(byte[] input) {
        return _printBase64Binary(input, 0, input.length);
    }

    public static String _printBase64Binary(byte[] input, int offset, int len) {
        char[] buf = new char[((len + 2) / 3) * 4];
        int ptr = _printBase64Binary(input, offset, len, buf, 0);
        assert ptr == buf.length;
        return new String(buf);
    }

    /**
     * Encodes a byte array into a char array by doing base64 encoding.
     *
     * The caller must supply a big enough buffer.
     *
     * @return
     *      the value of {@code ptr+((len+2)/3)*4}, which is the new offset
     *      in the output buffer where the further bytes should be placed.
     */
    public static int _printBase64Binary(byte[] input, int offset, int len, char[] buf, int ptr) {
        // encode elements until only 1 or 2 elements are left to encode
        int remaining = len;
        int i;
        for (i = offset;remaining >= 3; remaining -= 3, i += 3) {
            buf[ptr++] = encode(input[i] >> 2);
            buf[ptr++] = encode(
                    ((input[i] & 0x3) << 4)
                    | ((input[i + 1] >> 4) & 0xF));
            buf[ptr++] = encode(
                    ((input[i + 1] & 0xF) << 2)
                    | ((input[i + 2] >> 6) & 0x3));
            buf[ptr++] = encode(input[i + 2] & 0x3F);
        }
        // encode when exactly 1 element (left) to encode
        if (remaining == 1) {
            buf[ptr++] = encode(input[i] >> 2);
            buf[ptr++] = encode(((input[i]) & 0x3) << 4);
            buf[ptr++] = '=';
            buf[ptr++] = '=';
        }
        // encode when exactly 2 elements (left) to encode
        if (remaining == 2) {
            buf[ptr++] = encode(input[i] >> 2);
            buf[ptr++] = encode(((input[i] & 0x3) << 4)
                    | ((input[i + 1] >> 4) & 0xF));
            buf[ptr++] = encode((input[i + 1] & 0xF) << 2);
            buf[ptr++] = '=';
        }
        return ptr;
    }

    public static void _printBase64Binary(byte[] input, int offset, int len, XMLStreamWriter output) throws XMLStreamException {
        int remaining = len;
        int i;
        char[] buf = new char[4];

        for (i = offset; remaining >= 3; remaining -= 3, i += 3) {
            buf[0] = encode(input[i] >> 2);
            buf[1] = encode(
                    ((input[i] & 0x3) << 4)
                    | ((input[i + 1] >> 4) & 0xF));
            buf[2] = encode(
                    ((input[i + 1] & 0xF) << 2)
                    | ((input[i + 2] >> 6) & 0x3));
            buf[3] = encode(input[i + 2] & 0x3F);
            output.writeCharacters(buf, 0, 4);
        }
        // encode when exactly 1 element (left) to encode
        if (remaining == 1) {
            buf[0] = encode(input[i] >> 2);
            buf[1] = encode(((input[i]) & 0x3) << 4);
            buf[2] = '=';
            buf[3] = '=';
            output.writeCharacters(buf, 0, 4);
        }
        // encode when exactly 2 elements (left) to encode
        if (remaining == 2) {
            buf[0] = encode(input[i] >> 2);
            buf[1] = encode(((input[i] & 0x3) << 4)
                    | ((input[i + 1] >> 4) & 0xF));
            buf[2] = encode((input[i + 1] & 0xF) << 2);
            buf[3] = '=';
            output.writeCharacters(buf, 0, 4);
        }
    }

    /**
     * Encodes a byte array into another byte array by first doing base64 encoding
     * then encoding the result in ASCII.
     *
     * The caller must supply a big enough buffer.
     *
     * @return
     *      the value of {@code ptr+((len+2)/3)*4}, which is the new offset
     *      in the output buffer where the further bytes should be placed.
     */
    public static int _printBase64Binary(byte[] input, int offset, int len, byte[] out, int ptr) {
        byte[] buf = out;
        int remaining = len;
        int i;
        for (i=offset; remaining >= 3; remaining -= 3, i += 3 ) {
            buf[ptr++] = encodeByte(input[i]>>2);
            buf[ptr++] = encodeByte(
                        ((input[i]&0x3)<<4) |
                        ((input[i+1]>>4)&0xF));
            buf[ptr++] = encodeByte(
                        ((input[i+1]&0xF)<<2)|
                        ((input[i+2]>>6)&0x3));
            buf[ptr++] = encodeByte(input[i+2]&0x3F);
        }
        // encode when exactly 1 element (left) to encode
        if (remaining == 1) {
            buf[ptr++] = encodeByte(input[i]>>2);
            buf[ptr++] = encodeByte(((input[i])&0x3)<<4);
            buf[ptr++] = '=';
            buf[ptr++] = '=';
        }
        // encode when exactly 2 elements (left) to encode
        if (remaining == 2) {
            buf[ptr++] = encodeByte(input[i]>>2);
            buf[ptr++] = encodeByte(
                        ((input[i]&0x3)<<4) |
                        ((input[i+1]>>4)&0xF));
            buf[ptr++] = encodeByte((input[i+1]&0xF)<<2);
            buf[ptr++] = '=';
        }

        return ptr;
    }

    private static CharSequence removeOptionalPlus(CharSequence s) {
        int len = s.length();

        if (len <= 1 || s.charAt(0) != '+') {
            return s;
        }

        s = s.subSequence(1, len);
        char ch = s.charAt(0);
        if ('0' <= ch && ch <= '9') {
            return s;
        }
        if ('.' == ch) {
            return s;
        }

        throw new NumberFormatException();
    }

    private static boolean isDigitOrPeriodOrSign(char ch) {
        if ('0' <= ch && ch <= '9') {
            return true;
        }
        if (ch == '+' || ch == '-' || ch == '.') {
            return true;
        }
        return false;
    }

    private static final Map<ClassLoader, DatatypeFactory> DF_CACHE = Collections.synchronizedMap(new WeakHashMap<ClassLoader, DatatypeFactory>());

    public static DatatypeFactory getDatatypeFactory() {
        ClassLoader tccl = AccessController.doPrivileged(new PrivilegedAction<ClassLoader>() {
            public ClassLoader run() {
                return Thread.currentThread().getContextClassLoader();
            }
        });
        DatatypeFactory df = DF_CACHE.get(tccl);
        if (df == null) {
            synchronized (DatatypeConverterImpl.class) {
                df = DF_CACHE.get(tccl);
                if (df == null) { // to prevent multiple initialization
                    try {
                        df = DatatypeFactory.newInstance();
                    } catch (DatatypeConfigurationException e) {
                        throw new Error(Messages.FAILED_TO_INITIALE_DATATYPE_FACTORY.format(),e);
                    }
                    DF_CACHE.put(tccl, df);
                }
            }
        }
        return df;
    }

    private static final class CalendarFormatter {

        public static String doFormat(String format, Calendar cal) throws IllegalArgumentException {
            int fidx = 0;
            int flen = format.length();
            StringBuilder buf = new StringBuilder();

            while (fidx < flen) {
                char fch = format.charAt(fidx++);

                if (fch != '%') {  // not a meta character
                    buf.append(fch);
                    continue;
                }

                // seen meta character. we don't do error check against the format
                switch (format.charAt(fidx++)) {
                    case 'Y': // year
                        formatYear(cal, buf);
                        break;

                    case 'M': // month
                        formatMonth(cal, buf);
                        break;

                    case 'D': // days
                        formatDays(cal, buf);
                        break;

                    case 'h': // hours
                        formatHours(cal, buf);
                        break;

                    case 'm': // minutes
                        formatMinutes(cal, buf);
                        break;

                    case 's': // parse seconds.
                        formatSeconds(cal, buf);
                        break;

                    case 'z': // time zone
                        formatTimeZone(cal, buf);
                        break;

                    default:
                        // illegal meta character. impossible.
                        throw new InternalError();
                }
            }

            return buf.toString();
        }

        private static void formatYear(Calendar cal, StringBuilder buf) {
            int year = cal.get(Calendar.YEAR);

            String s;
            if (year <= 0) // negative value
            {
                s = Integer.toString(1 - year);
            } else // positive value
            {
                s = Integer.toString(year);
            }

            while (s.length() < 4) {
                s = '0' + s;
            }
            if (year <= 0) {
                s = '-' + s;
            }

            buf.append(s);
        }

        private static void formatMonth(Calendar cal, StringBuilder buf) {
            formatTwoDigits(cal.get(Calendar.MONTH) + 1, buf);
        }

        private static void formatDays(Calendar cal, StringBuilder buf) {
            formatTwoDigits(cal.get(Calendar.DAY_OF_MONTH), buf);
        }

        private static void formatHours(Calendar cal, StringBuilder buf) {
            formatTwoDigits(cal.get(Calendar.HOUR_OF_DAY), buf);
        }

        private static void formatMinutes(Calendar cal, StringBuilder buf) {
            formatTwoDigits(cal.get(Calendar.MINUTE), buf);
        }

        private static void formatSeconds(Calendar cal, StringBuilder buf) {
            formatTwoDigits(cal.get(Calendar.SECOND), buf);
            if (cal.isSet(Calendar.MILLISECOND)) { // milliseconds
                int n = cal.get(Calendar.MILLISECOND);
                if (n != 0) {
                    String ms = Integer.toString(n);
                    while (ms.length() < 3) {
                        ms = '0' + ms; // left 0 paddings.
                    }
                    buf.append('.');
                    buf.append(ms);
                }
            }
        }

        /** formats time zone specifier. */
        private static void formatTimeZone(Calendar cal, StringBuilder buf) {
            TimeZone tz = cal.getTimeZone();

            if (tz == null) {
                return;
            }

            // otherwise print out normally.
            int offset = tz.getOffset(cal.getTime().getTime());

            if (offset == 0) {
                buf.append('Z');
                return;
            }

            if (offset >= 0) {
                buf.append('+');
            } else {
                buf.append('-');
                offset *= -1;
            }

            offset /= 60 * 1000; // offset is in milli-seconds

            formatTwoDigits(offset / 60, buf);
            buf.append(':');
            formatTwoDigits(offset % 60, buf);
        }

        /** formats Integer into two-character-wide string. */
        private static void formatTwoDigits(int n, StringBuilder buf) {
            // n is always non-negative.
            if (n < 10) {
                buf.append('0');
            }
            buf.append(n);
        }
    }
}