am dbca824d: Merge "Add DateFormat.getBestDateTimePattern." into jb-mr2-dev

    method public static java.lang.CharSequence format(java.lang.CharSequence, long);

    method public static java.lang.CharSequence format(java.lang.CharSequence, java.util.Date);

    method public static java.lang.CharSequence format(java.lang.CharSequence, java.util.Calendar);

    method public static java.lang.String getBestDateTimePattern(java.util.Locale, java.lang.String);

    method public static java.text.DateFormat getDateFormat(android.content.Context);

    method public static char[] getDateFormatOrder(android.content.Context);

    method public static java.text.DateFormat getLongDateFormat(android.content.Context);

import java.util.TimeZone;

import java.text.SimpleDateFormat;

import libcore.icu.ICU;

import libcore.icu.LocaleData;

/**

 * for both formatting and parsing dates. For the canonical documentation

 * of format strings, see {@link java.text.SimpleDateFormat}.

 *

 * <p>In cases where the system does not provide a suitable pattern,

 * this class offers the {@link #getBestDateTimePattern} method.

 *

 * <p>The {@code format} methods in this class implement a subset of Unicode

 * <a href="http://www.unicode.org/reports/tr35/#Date_Format_Patterns">UTS #35</a> patterns.

 * The subset currently supported by this class includes the following format characters:

        return value.equals("24");

    }

    /**

     * Returns the best possible localized form of the given skeleton for the given

     * locale. A skeleton is similar to, and uses the same format characters as, a Unicode

     * <a href="http://www.unicode.org/reports/tr35/#Date_Format_Patterns">UTS #35</a>

     * pattern.

     *

     * <p>One difference is that order is irrelevant. For example, "MMMMd" will return

     * "MMMM d" in the {@code en_US} locale, but "d. MMMM" in the {@code de_CH} locale.

     *

     * <p>Note also in that second example that the necessary punctuation for German was

     * added. For the same input in {@code es_ES}, we'd have even more extra text:

     * "d 'de' MMMM".

     *

     * <p>This method will automatically correct for grammatical necessity. Given the

     * same "MMMMd" input, this method will return "d LLLL" in the {@code fa_IR} locale,

     * where stand-alone months are necessary. Lengths are preserved where meaningful,

     * so "Md" would give a different result to "MMMd", say, except in a locale such as

     * {@code ja_JP} where there is only one length of month.

     *

     * <p>This method will only return patterns that are in CLDR, and is useful whenever

     * you know what elements you want in your format string but don't want to make your

     * code specific to any one locale.

     *

     * @param locale the locale into which the skeleton should be localized

     * @param skeleton a skeleton as described above

     * @return a string pattern suitable for use with {@link java.text.SimpleDateFormat}.

     */

    public static String getBestDateTimePattern(Locale locale, String skeleton) {

        return ICU.getBestDateTimePattern(skeleton, locale.toString());

    }

    /**

     * Returns a {@link java.text.DateFormat} object that can format the time according

     * to the current locale and the user's 12-/24-hour clock preference.