/*

* Copyright (c) 2003, 2011, Oracle and/or its affiliates. All rights reserved.

* ORACLE PROPRIETARY/CONFIDENTIAL. Use is subject to license terms.

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*/

package java.util;

import java.io.BufferedWriter;

import java.io.Closeable;

import java.io.IOException;

import java.io.File;

import java.io.FileOutputStream;

import java.io.FileNotFoundException;

import java.io.Flushable;

import java.io.OutputStream;

import java.io.OutputStreamWriter;

import java.io.PrintStream;

import java.io.UnsupportedEncodingException;

import java.math.BigDecimal;

import java.math.BigInteger;

import java.math.MathContext;

import java.math.RoundingMode;

import java.nio.charset.Charset;

import java.nio.charset.IllegalCharsetNameException;

import java.nio.charset.UnsupportedCharsetException;

import java.text.DateFormatSymbols;

import java.text.DecimalFormat;

import java.text.DecimalFormatSymbols;

import java.text.NumberFormat;

import java.util.regex.Matcher;

import java.util.regex.Pattern;

import sun.misc.FpUtils;

import sun.misc.DoubleConsts;

import sun.misc.FormattedFloatingDecimal;

/**

* An interpreter for printf-style format strings. This class provides support

* for layout justification and alignment, common formats for numeric, string,

* and date/time data, and locale-specific output. Common Java types such as

* {@code byte}, {@link java.math.BigDecimal BigDecimal}, and {@link Calendar}

* are supported. Limited formatting customization for arbitrary user types is

* provided through the {@link Formattable} interface.

*

* <p> Formatters are not necessarily safe for multithreaded access. Thread

* safety is optional and is the responsibility of users of methods in this

* class.

*

* <p> Formatted printing for the Java language is heavily inspired by C's

* {@code printf}. Although the format strings are similar to C, some

* customizations have been made to accommodate the Java language and exploit

* some of its features. Also, Java formatting is more strict than C's; for

* example, if a conversion is incompatible with a flag, an exception will be

* thrown. In C inapplicable flags are silently ignored. The format strings

* are thus intended to be recognizable to C programmers but not necessarily

* completely compatible with those in C.

*

* <p> Examples of expected usage:

*

* <blockquote><pre>

* StringBuilder sb = new StringBuilder();

* // Send all output to the Appendable object sb

* Formatter formatter = new Formatter(sb, Locale.US);

*

* // Explicit argument indices may be used to re-order output.

* formatter.format("%4$2s %3$2s %2$2s %1$2s", "a", "b", "c", "d")

* // -&gt; " d c b a"

*

* // Optional locale as the first argument can be used to get

* // locale-specific formatting of numbers. The precision and width can be

* // given to round and align the value.

* formatter.format(Locale.FRANCE, "e = %+10.4f", Math.E);

* // -&gt; "e = +2,7183"

*

* // The '(' numeric flag may be used to format negative numbers with

* // parentheses rather than a minus sign. Group separators are

* // automatically inserted.

* formatter.format("Amount gained or lost since last statement: $ %(,.2f",

* balanceDelta);

* // -&gt; "Amount gained or lost since last statement: $ (6,217.58)"

* </pre></blockquote>

*

* <p> Convenience methods for common formatting requests exist as illustrated

* by the following invocations:

*

* <blockquote><pre>

* // Writes a formatted string to System.out.

* System.out.format("Local time: %tT", Calendar.getInstance());

* // -&gt; "Local time: 13:34:18"

*

* // Writes formatted output to System.err.

* System.err.printf("Unable to open file '%1$s': %2$s",

* fileName, exception.getMessage());

* // -&gt; "Unable to open file 'food': No such file or directory"

* </pre></blockquote>

*

* <p> Like C's {@code sprintf(3)}, Strings may be formatted using the static

* method {@link String#format(String,Object...) String.format}:

*

* <blockquote><pre>

* // Format a string containing a date.

* import java.util.Calendar;

* import java.util.GregorianCalendar;

* import static java.util.Calendar.*;

*

* Calendar c = new GregorianCalendar(1995, MAY, 23);

* String s = String.format("Duke's Birthday: %1$tm %1$te,%1$tY", c);

* // -&gt; s == "Duke's Birthday: May 23, 1995"

* </pre></blockquote>

*

* <h3><a name="org">Organization</a></h3>

*

* <p> This specification is divided into two sections. The first section, <a

* href="#summary">Summary</a>, covers the basic formatting concepts. This

* section is intended for users who want to get started quickly and are

* familiar with formatted printing in other programming languages. The second

* section, <a href="#detail">Details</a>, covers the specific implementation

* details. It is intended for users who want more precise specification of

* formatting behavior.

*

* <h3><a name="summary">Summary</a></h3>

*

* <p> This section is intended to provide a brief overview of formatting

* concepts. For precise behavioral details, refer to the <a

* href="#detail">Details</a> section.

*

* <h4><a name="syntax">Format String Syntax</a></h4>

*

* <p> Every method which produces formatted output requires a <i>format

* string</i> and an <i>argument list</i>. The format string is a {@link

* String} which may contain fixed text and one or more embedded <i>format

* specifiers</i>. Consider the following example:

*

* <blockquote><pre>

* Calendar c = ...;

* String s = String.format("Duke's Birthday: %1$tm %1$te,%1$tY", c);

* </pre></blockquote>

*

* This format string is the first argument to the {@code format} method. It

* contains three format specifiers "{@code %1$tm}", "{@code %1$te}", and

* "{@code %1$tY}" which indicate how the arguments should be processed and

* where they should be inserted in the text. The remaining portions of the

* format string are fixed text including {@code "Dukes Birthday: "} and any

* other spaces or punctuation.

*

* The argument list consists of all arguments passed to the method after the

* format string. In the above example, the argument list is of size one and

* consists of the {@link java.util.Calendar Calendar} object {@code c}.

*

* <ul>

*

* <li> The format specifiers for general, character, and numeric types have

* the following syntax:

*

* <blockquote><pre>

* %[argument_index$][flags][width][.precision]conversion

* </pre></blockquote>

*

* <p> The optional <i>argument_index</i> is a decimal integer indicating the

* position of the argument in the argument list. The first argument is

* referenced by "{@code 1$}", the second by "{@code 2$}", etc.

*

* <p> The optional <i>flags</i> is a set of characters that modify the output

* format. The set of valid flags depends on the conversion.

*

* <p> The optional <i>width</i> is a non-negative decimal integer indicating

* the minimum number of characters to be written to the output.

*

* <p> The optional <i>precision</i> is a non-negative decimal integer usually

* used to restrict the number of characters. The specific behavior depends on

* the conversion.

*

* <p> The required <i>conversion</i> is a character indicating how the

* argument should be formatted. The set of valid conversions for a given

* argument depends on the argument's data type.

*

* <li> The format specifiers for types which are used to represents dates and

* times have the following syntax:

*

* <blockquote><pre>

* %[argument_index$][flags][width]conversion

* </pre></blockquote>

*

* <p> The optional <i>argument_index</i>, <i>flags</i> and <i>width</i> are

* defined as above.

*

* <p> The required <i>conversion</i> is a two character sequence. The first

* character is {@code 't'} or {@code 'T'}. The second character indicates

* the format to be used. These characters are similar to but not completely

* identical to those defined by GNU {@code date} and POSIX

* {@code strftime(3c)}.

*

* <li> The format specifiers which do not correspond to arguments have the

* following syntax:

*

* <blockquote><pre>

* %[flags][width]conversion

* </pre></blockquote>

*

* <p> The optional <i>flags</i> and <i>width</i> is defined as above.

*

* <p> The required <i>conversion</i> is a character indicating content to be

* inserted in the output.

*

* </ul>

*

* <h4> Conversions </h4>

*

* <p> Conversions are divided into the following categories:

*

* <ol>

*

* <li> <b>General</b> - may be applied to any argument

* type

*

* <li> <b>Character</b> - may be applied to basic types which represent

* Unicode characters: {@code char}, {@link Character}, {@code byte}, {@link

* Byte}, {@code short}, and {@link Short}. This conversion may also be

* applied to the types {@code int} and {@link Integer} when {@link

* Character#isValidCodePoint} returns {@code true}

*

* <li> <b>Numeric</b>

*

* <ol>

*

* <li> <b>Integral</b> - may be applied to Java integral types: {@code byte},

* {@link Byte}, {@code short}, {@link Short}, {@code int} and {@link

* Integer}, {@code long}, {@link Long}, and {@link java.math.BigInteger

* BigInteger}

*

* <li><b>Floating Point</b> - may be applied to Java floating-point types:

* {@code float}, {@link Float}, {@code double}, {@link Double}, and {@link

* java.math.BigDecimal BigDecimal}

*

* </ol>

*

* <li> <b>Date/Time</b> - may be applied to Java types which are capable of

* encoding a date or time: {@code long}, {@link Long}, {@link Calendar}, and

* {@link Date}.

*

* <li> <b>Percent</b> - produces a literal {@code '%'}

* (<tt>'&#92;u0025'</tt>)

*

* <li> <b>Line Separator</b> - produces the platform-specific line separator

*

* </ol>

*

* <p> The following table summarizes the supported conversions. Conversions

* denoted by an upper-case character (i.e. {@code 'B'}, {@code 'H'},

* {@code 'S'}, {@code 'C'}, {@code 'X'}, {@code 'E'}, {@code 'G'},

* {@code 'A'}, and {@code 'T'}) are the same as those for the corresponding

* lower-case conversion characters except that the result is converted to

* upper case according to the rules of the prevailing {@link java.util.Locale

* Locale}. The result is equivalent to the following invocation of {@link

* String#toUpperCase()}

*

* <pre>

* out.toUpperCase() </pre>

*

* <table cellpadding=5 summary="genConv">

*

* <tr><th valign="bottom"> Conversion

* <th valign="bottom"> Argument Category

* <th valign="bottom"> Description

*

* <tr><td valign="top"> {@code 'b'}, {@code 'B'}

* <td valign="top"> general

* <td> If the argument <i>arg</i> is {@code null}, then the result is

* "{@code false}". If <i>arg</i> is a {@code boolean} or {@link

* Boolean}, then the result is the string returned by {@link

* String#valueOf(boolean) String.valueOf(arg)}. Otherwise, the result is

* "true".

*

* <tr><td valign="top"> {@code 'h'}, {@code 'H'}

* <td valign="top"> general

* <td> If the argument <i>arg</i> is {@code null}, then the result is

* "{@code null}". Otherwise, the result is obtained by invoking

* {@code Integer.toHexString(arg.hashCode())}.

*

* <tr><td valign="top"> {@code 's'}, {@code 'S'}

* <td valign="top"> general

* <td> If the argument <i>arg</i> is {@code null}, then the result is

* "{@code null}". If <i>arg</i> implements {@link Formattable}, then

* {@link Formattable#formatTo arg.formatTo} is invoked. Otherwise, the

* result is obtained by invoking {@code arg.toString()}.

*

* <tr><td valign="top">{@code 'c'}, {@code 'C'}

* <td valign="top"> character

* <td> The result is a Unicode character

*

* <tr><td valign="top">{@code 'd'}

* <td valign="top"> integral

* <td> The result is formatted as a decimal integer

*

* <tr><td valign="top">{@code 'o'}

* <td valign="top"> integral

* <td> The result is formatted as an octal integer

*

* <tr><td valign="top">{@code 'x'}, {@code 'X'}

* <td valign="top"> integral

* <td> The result is formatted as a hexadecimal integer

*

* <tr><td valign="top">{@code 'e'}, {@code 'E'}

* <td valign="top"> floating point

* <td> The result is formatted as a decimal number in computerized

* scientific notation

*

* <tr><td valign="top">{@code 'f'}

* <td valign="top"> floating point

* <td> The result is formatted as a decimal number

*

* <tr><td valign="top">{@code 'g'}, {@code 'G'}

* <td valign="top"> floating point

* <td> The result is formatted using computerized scientific notation or

* decimal format, depending on the precision and the value after rounding.

*

* <tr><td valign="top">{@code 'a'}, {@code 'A'}

* <td valign="top"> floating point

* <td> The result is formatted as a hexadecimal floating-point number with

* a significand and an exponent

*

* <tr><td valign="top">{@code 't'}, {@code 'T'}

* <td valign="top"> date/time

* <td> Prefix for date and time conversion characters. See <a

* href="#dt">Date/Time Conversions</a>.

*

* <tr><td valign="top">{@code '%'}

* <td valign="top"> percent

* <td> The result is a literal {@code '%'} (<tt>'&#92;u0025'</tt>)

*

* <tr><td valign="top">{@code 'n'}

* <td valign="top"> line separator

* <td> The result is the platform-specific line separator

*

* </table>

*

* <p> Any characters not explicitly defined as conversions are illegal and are

* reserved for future extensions.

*

* <h4><a name="dt">Date/Time Conversions</a></h4>

*

* <p> The following date and time conversion suffix characters are defined for

* the {@code 't'} and {@code 'T'} conversions. The types are similar to but

* not completely identical to those defined by GNU {@code date} and POSIX

* {@code strftime(3c)}. Additional conversion types are provided to access

* Java-specific functionality (e.g. {@code 'L'} for milliseconds within the

* second).

*

* <p> The following conversion characters are used for formatting times:

*

* <table cellpadding=5 summary="time">

*

* <tr><td valign="top"> {@code 'H'}

* <td> Hour of the day for the 24-hour clock, formatted as two digits with

* a leading zero as necessary i.e. {@code 00 - 23}.

*

* <tr><td valign="top">{@code 'I'}

* <td> Hour for the 12-hour clock, formatted as two digits with a leading

* zero as necessary, i.e. {@code 01 - 12}.

*

* <tr><td valign="top">{@code 'k'}

* <td> Hour of the day for the 24-hour clock, i.e. {@code 0 - 23}.

*

* <tr><td valign="top">{@code 'l'}

* <td> Hour for the 12-hour clock, i.e. {@code 1 - 12}.

*

* <tr><td valign="top">{@code 'M'}

* <td> Minute within the hour formatted as two digits with a leading zero

* as necessary, i.e. {@code 00 - 59}.

*

* <tr><td valign="top">{@code 'S'}

* <td> Seconds within the minute, formatted as two digits with a leading

* zero as necessary, i.e. {@code 00 - 60} ("{@code 60}" is a special

* value required to support leap seconds).

*

* <tr><td valign="top">{@code 'L'}

* <td> Millisecond within the second formatted as three digits with

* leading zeros as necessary, i.e. {@code 000 - 999}.

*

* <tr><td valign="top">{@code 'N'}

* <td> Nanosecond within the second, formatted as nine digits with leading

* zeros as necessary, i.e. {@code 000000000 - 999999999}.

*

* <tr><td valign="top">{@code 'p'}

* <td> Locale-specific {@linkplain

* java.text.DateFormatSymbols#getAmPmStrings morning or afternoon} marker

* in lower case, e.g."{@code am}" or "{@code pm}". Use of the conversion

* prefix {@code 'T'} forces this output to upper case.

*

* <tr><td valign="top">{@code 'z'}

* <td> <a href="http://www.ietf.org/rfc/rfc0822.txt">RFC&nbsp;822</a>

* style numeric time zone offset from GMT, e.g. {@code -0800}. This

* value will be adjusted as necessary for Daylight Saving Time. For

* {@code long}, {@link Long}, and {@link Date} the time zone used is

* the {@linkplain TimeZone#getDefault() default time zone} for this

* instance of the Java virtual machine.

*

* <tr><td valign="top">{@code 'Z'}

* <td> A string representing the abbreviation for the time zone. This

* value will be adjusted as necessary for Daylight Saving Time. For

* {@code long}, {@link Long}, and {@link Date} the time zone used is

* the {@linkplain TimeZone#getDefault() default time zone} for this

* instance of the Java virtual machine. The Formatter's locale will

* supersede the locale of the argument (if any).

*

* <tr><td valign="top">{@code 's'}

* <td> Seconds since the beginning of the epoch starting at 1 January 1970

* {@code 00:00:00} UTC, i.e. {@code Long.MIN_VALUE/1000} to

* {@code Long.MAX_VALUE/1000}.

*

* <tr><td valign="top">{@code 'Q'}

* <td> Milliseconds since the beginning of the epoch starting at 1 January

* 1970 {@code 00:00:00} UTC, i.e. {@code Long.MIN_VALUE} to

* {@code Long.MAX_VALUE}.

*

* </table>

*

* <p> The following conversion characters are used for formatting dates:

*

* <table cellpadding=5 summary="date">

*

* <tr><td valign="top">{@code 'B'}

* <td> Locale-specific {@linkplain java.text.DateFormatSymbols#getMonths

* full month name}, e.g. {@code "January"}, {@code "February"}.

*

* <tr><td valign="top">{@code 'b'}

* <td> Locale-specific {@linkplain

* java.text.DateFormatSymbols#getShortMonths abbreviated month name},

* e.g. {@code "Jan"}, {@code "Feb"}.

*

* <tr><td valign="top">{@code 'h'}

* <td> Same as {@code 'b'}.

*

* <tr><td valign="top">{@code 'A'}

* <td> Locale-specific full name of the {@linkplain

* java.text.DateFormatSymbols#getWeekdays day of the week},

* e.g. {@code "Sunday"}, {@code "Monday"}

*

* <tr><td valign="top">{@code 'a'}

* <td> Locale-specific short name of the {@linkplain

* java.text.DateFormatSymbols#getShortWeekdays day of the week},

* e.g. {@code "Sun"}, {@code "Mon"}

*

* <tr><td valign="top">{@code 'C'}

* <td> Four-digit year divided by {@code 100}, formatted as two digits

* with leading zero as necessary, i.e. {@code 00 - 99}

*

* <tr><td valign="top">{@code 'Y'}

* <td> Year, formatted as at least four digits with leading zeros as

* necessary, e.g. {@code 0092} equals {@code 92} CE for the Gregorian

* calendar.

*

* <tr><td valign="top">{@code 'y'}

* <td> Last two digits of the year, formatted with leading zeros as

* necessary, i.e. {@code 00 - 99}.

*

* <tr><td valign="top">{@code 'j'}

* <td> Day of year, formatted as three digits with leading zeros as

* necessary, e.g. {@code 001 - 366} for the Gregorian calendar.

*

* <tr><td valign="top">{@code 'm'}

* <td> Month, formatted as two digits with leading zeros as necessary,

* i.e. {@code 01 - 13}.

*

* <tr><td valign="top">{@code 'd'}

* <td> Day of month, formatted as two digits with leading zeros as

* necessary, i.e. {@code 01 - 31}

*

* <tr><td valign="top">{@code 'e'}

* <td> Day of month, formatted as two digits, i.e. {@code 1 - 31}.

*

* </table>

*

* <p> The following conversion characters are used for formatting common

* date/time compositions.

*

* <table cellpadding=5 summary="composites">

*

* <tr><td valign="top">{@code 'R'}

* <td> Time formatted for the 24-hour clock as {@code "%tH:%tM"}

*

* <tr><td valign="top">{@code 'T'}

* <td> Time formatted for the 24-hour clock as {@code "%tH:%tM:%tS"}.

*

* <tr><td valign="top">{@code 'r'}

* <td> Time formatted for the 12-hour clock as {@code "%tI:%tM:%tS %Tp"}.

* The location of the morning or afternoon marker ({@code '%Tp'}) may be

* locale-dependent.

*

* <tr><td valign="top">{@code 'D'}

* <td> Date formatted as {@code "%tm/%td/%ty"}.

*

* <tr><td valign="top">{@code 'F'}

* <td> <a href="http://www.w3.org/TR/NOTE-datetime">ISO&nbsp;8601</a>

* complete date formatted as {@code "%tY-%tm-%td"}.

*

* <tr><td valign="top">{@code 'c'}

* <td> Date and time formatted as {@code "%ta %tb %td %tT %tZ %tY"},

* e.g. {@code "Sun Jul 20 16:17:00 EDT 1969"}.

*

* </table>

*

* <p> Any characters not explicitly defined as date/time conversion suffixes

* are illegal and are reserved for future extensions.

*

* <h4> Flags </h4>

*

* <p> The following table summarizes the supported flags. <i>y</i> means the

* flag is supported for the indicated argument types.

*

* <table cellpadding=5 summary="genConv">

*

* <tr><th valign="bottom"> Flag <th valign="bottom"> General

* <th valign="bottom"> Character <th valign="bottom"> Integral

* <th valign="bottom"> Floating Point

* <th valign="bottom"> Date/Time

* <th valign="bottom"> Description

*

* <tr><td> '-' <td align="center" valign="top"> y

* <td align="center" valign="top"> y

* <td align="center" valign="top"> y

* <td align="center" valign="top"> y

* <td align="center" valign="top"> y

* <td> The result will be left-justified.

*

* <tr><td> '#' <td align="center" valign="top"> y¹

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y³

* <td align="center" valign="top"> y

* <td align="center" valign="top"> -

* <td> The result should use a conversion-dependent alternate form

*

* <tr><td> '+' <td align="center" valign="top"> -

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y⁴

* <td align="center" valign="top"> y

* <td align="center" valign="top"> -

* <td> The result will always include a sign

*

* <tr><td> '&nbsp;&nbsp;' <td align="center" valign="top"> -

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y⁴

* <td align="center" valign="top"> y

* <td align="center" valign="top"> -

* <td> The result will include a leading space for positive values

*

* <tr><td> '0' <td align="center" valign="top"> -

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y

* <td align="center" valign="top"> y

* <td align="center" valign="top"> -

* <td> The result will be zero-padded

*

* <tr><td> ',' <td align="center" valign="top"> -

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y²

* <td align="center" valign="top"> y⁵

* <td align="center" valign="top"> -

* <td> The result will include locale-specific {@linkplain

* java.text.DecimalFormatSymbols#getGroupingSeparator grouping separators}

*

* <tr><td> '(' <td align="center" valign="top"> -

* <td align="center" valign="top"> -

* <td align="center" valign="top"> y⁴

* <td align="center" valign="top"> y⁵

* <td align="center"> -

* <td> The result will enclose negative numbers in parentheses

*

* </table>

*

* <p> ¹ Depends on the definition of {@link Formattable}.

*

* <p> ² For {@code 'd'} conversion only.

*

* <p> ³ For {@code 'o'}, {@code 'x'}, and {@code 'X'}

* conversions only.

*

* <p> ⁴ For {@code 'd'}, {@code 'o'}, {@code 'x'}, and

* {@code 'X'} conversions applied to {@link java.math.BigInteger BigInteger}

* or {@code 'd'} applied to {@code byte}, {@link Byte}, {@code short}, {@link

* Short}, {@code int} and {@link Integer}, {@code long}, and {@link Long}.

*

* <p> ⁵ For {@code 'e'}, {@code 'E'}, {@code 'f'},

* {@code 'g'}, and {@code 'G'} conversions only.

*

* <p> Any characters not explicitly defined as flags are illegal and are

* reserved for future extensions.

*

* <h4> Width </h4>

*

* <p> The width is the minimum number of characters to be written to the

* output. For the line separator conversion, width is not applicable; if it

* is provided, an exception will be thrown.

*

* <h4> Precision </h4>

*

* <p> For general argument types, the precision is the maximum number of

* characters to be written to the output.

*

* <p> For the floating-point conversions {@code 'e'}, {@code 'E'}, and

* {@code 'f'} the precision is the number of digits after the decimal

* separator. If the conversion is {@code 'g'} or {@code 'G'}, then the

* precision is the total number of digits in the resulting magnitude after

* rounding. If the conversion is {@code 'a'} or {@code 'A'}, then the

* precision must not be specified.

*

* <p> For character, integral, and date/time argument types and the percent

* and line separator conversions, the precision is not applicable; if a

* precision is provided, an exception will be thrown.

*

* <h4> Argument Index </h4>

*

* <p> The argument index is a decimal integer indicating the position of the

* argument in the argument list. The first argument is referenced by

* "{@code 1$}", the second by "{@code 2$}", etc.

*

* <p> Another way to reference arguments by position is to use the

* {@code '<'} (<tt>'&#92;u003c'</tt>) flag, which causes the argument for

* the previous format specifier to be re-used. For example, the following two

* statements would produce identical strings:

*

* <blockquote><pre>

* Calendar c = ...;

* String s1 = String.format("Duke's Birthday: %1$tm %1$te,%1$tY", c);

*

* String s2 = String.format("Duke's Birthday: %1$tm %&lt;te,%&lt;tY", c);

* </pre></blockquote>

*

* <hr>

* <h3><a name="detail">Details</a></h3>

*

* <p> This section is intended to provide behavioral details for formatting,

* including conditions and exceptions, supported data types, localization, and

* interactions between flags, conversions, and data types. For an overview of

* formatting concepts, refer to the <a href="#summary">Summary</a>

*

* <p> Any characters not explicitly defined as conversions, date/time

* conversion suffixes, or flags are illegal and are reserved for

* future extensions. Use of such a character in a format string will

* cause an {@link UnknownFormatConversionException} or {@link

* UnknownFormatFlagsException} to be thrown.

*

* <p> If the format specifier contains a width or precision with an invalid

* value or which is otherwise unsupported, then a {@link

* IllegalFormatWidthException} or {@link IllegalFormatPrecisionException}

* respectively will be thrown.

*

* <p> If a format specifier contains a conversion character that is not

* applicable to the corresponding argument, then an {@link

* IllegalFormatConversionException} will be thrown.

*

* <p> All specified exceptions may be thrown by any of the {@code format}

* methods of {@code Formatter} as well as by any {@code format} convenience

* methods such as {@link String#format(String,Object...) String.format} and

* {@link java.io.PrintStream#printf(String,Object...) PrintStream.printf}.

*

* <p> Conversions denoted by an upper-case character (i.e. {@code 'B'},

* {@code 'H'}, {@code 'S'}, {@code 'C'}, {@code 'X'}, {@code 'E'},

* {@code 'G'}, {@code 'A'}, and {@code 'T'}) are the same as those for the

* corresponding lower-case conversion characters except that the result is

* converted to upper case according to the rules of the prevailing {@link

* java.util.Locale Locale}. The result is equivalent to the following

* invocation of {@link String#toUpperCase()}

*

* <pre>

* out.toUpperCase() </pre>

*

* <h4><a name="dgen">General</a></h4>

*

* <p> The following general conversions may be applied to any argument type:

*

* <table cellpadding=5 summary="dgConv">

*

* <tr><td valign="top"> {@code 'b'}

* <td valign="top"> <tt>'&#92;u0062'</tt>

* <td> Produces either "{@code true}" or "{@code false}" as returned by

* {@link Boolean#toString(boolean)}.

*

* <p> If the argument is {@code null}, then the result is

* "{@code false}". If the argument is a {@code boolean} or {@link

* Boolean}, then the result is the string returned by {@link

* String#valueOf(boolean) String.valueOf()}. Otherwise, the result is

* "{@code true}".

*

* <p> If the {@code '#'} flag is given, then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <tr><td valign="top"> {@code 'B'}

* <td valign="top"> <tt>'&#92;u0042'</tt>

* <td> The upper-case variant of {@code 'b'}.

*

* <tr><td valign="top"> {@code 'h'}

* <td valign="top"> <tt>'&#92;u0068'</tt>

* <td> Produces a string representing the hash code value of the object.

*

* <p> If the argument, <i>arg</i> is {@code null}, then the

* result is "{@code null}". Otherwise, the result is obtained

* by invoking {@code Integer.toHexString(arg.hashCode())}.

*

* <p> If the {@code '#'} flag is given, then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <tr><td valign="top"> {@code 'H'}

* <td valign="top"> <tt>'&#92;u0048'</tt>

* <td> The upper-case variant of {@code 'h'}.

*

* <tr><td valign="top"> {@code 's'}

* <td valign="top"> <tt>'&#92;u0073'</tt>

* <td> Produces a string.

*

* <p> If the argument is {@code null}, then the result is

* "{@code null}". If the argument implements {@link Formattable}, then

* its {@link Formattable#formatTo formatTo} method is invoked.

* Otherwise, the result is obtained by invoking the argument's

* {@code toString()} method.

*

* <p> If the {@code '#'} flag is given and the argument is not a {@link

* Formattable} , then a {@link FormatFlagsConversionMismatchException}

* will be thrown.

*

* <tr><td valign="top"> {@code 'S'}

* <td valign="top"> <tt>'&#92;u0053'</tt>

* <td> The upper-case variant of {@code 's'}.

*

* </table>

*

* <p> The following <a name="dFlags">flags</a> apply to general conversions:

*

* <table cellpadding=5 summary="dFlags">

*

* <tr><td valign="top"> {@code '-'}

* <td valign="top"> <tt>'&#92;u002d'</tt>

* <td> Left justifies the output. Spaces (<tt>'&#92;u0020'</tt>) will be

* added at the end of the converted value as required to fill the minimum

* width of the field. If the width is not provided, then a {@link

* MissingFormatWidthException} will be thrown. If this flag is not given

* then the output will be right-justified.

*

* <tr><td valign="top"> {@code '#'}

* <td valign="top"> <tt>'&#92;u0023'</tt>

* <td> Requires the output use an alternate form. The definition of the

* form is specified by the conversion.

*

* </table>

*

* <p> The <a name="genWidth">width</a> is the minimum number of characters to

* be written to the

* output. If the length of the converted value is less than the width then

* the output will be padded by <tt>'&nbsp;&nbsp;'</tt> (<tt>'&#92;u0020'</tt>)

* until the total number of characters equals the width. The padding is on

* the left by default. If the {@code '-'} flag is given, then the padding

* will be on the right. If the width is not specified then there is no

* minimum.

*

* <p> The precision is the maximum number of characters to be written to the

* output. The precision is applied before the width, thus the output will be

* truncated to {@code precision} characters even if the width is greater than

* the precision. If the precision is not specified then there is no explicit

* limit on the number of characters.

*

* <h4><a name="dchar">Character</a></h4>

*

* This conversion may be applied to {@code char} and {@link Character}. It

* may also be applied to the types {@code byte}, {@link Byte},

* {@code short}, and {@link Short}, {@code int} and {@link Integer} when

* {@link Character#isValidCodePoint} returns {@code true}. If it returns

* {@code false} then an {@link IllegalFormatCodePointException} will be

* thrown.

*

* <table cellpadding=5 summary="charConv">

*

* <tr><td valign="top"> {@code 'c'}

* <td valign="top"> <tt>'&#92;u0063'</tt>

* <td> Formats the argument as a Unicode character as described in <a

* href="../lang/Character.html#unicode">Unicode Character

* Representation</a>. This may be more than one 16-bit {@code char} in

* the case where the argument represents a supplementary character.

*

* <p> If the {@code '#'} flag is given, then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <tr><td valign="top"> {@code 'C'}

* <td valign="top"> <tt>'&#92;u0043'</tt>

* <td> The upper-case variant of {@code 'c'}.

*

* </table>

*

* <p> The {@code '-'} flag defined for <a href="#dFlags">General

* conversions</a> applies. If the {@code '#'} flag is given, then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <p> The width is defined as for <a href="#genWidth">General conversions</a>.

*

* <p> The precision is not applicable. If the precision is specified then an

* {@link IllegalFormatPrecisionException} will be thrown.

*

* <h4><a name="dnum">Numeric</a></h4>

*

* <p> Numeric conversions are divided into the following categories:

*

* <ol>

*

* <li> <a href="#dnint"><b>Byte, Short, Integer, and Long</b></a>

*

* <li> <a href="#dnbint"><b>BigInteger</b></a>

*

* <li> <a href="#dndec"><b>Float and Double</b></a>

*

* <li> <a href="#dnbdec"><b>BigDecimal</b></a>

*

* </ol>

*

* <p> Numeric types will be formatted according to the following algorithm:

*

* <p><b><a name="l10n algorithm"> Number Localization Algorithm</a></b>

*

* <p> After digits are obtained for the integer part, fractional part, and

* exponent (as appropriate for the data type), the following transformation

* is applied:

*

* <ol>

*

* <li> Each digit character <i>d</i> in the string is replaced by a

* locale-specific digit computed relative to the current locale's

* {@linkplain java.text.DecimalFormatSymbols#getZeroDigit() zero digit}

* <i>z</i>; that is <i>d&nbsp;-&nbsp;</i> {@code '0'}

* <i>&nbsp;+&nbsp;z</i>.

*

* <li> If a decimal separator is present, a locale-specific {@linkplain

* java.text.DecimalFormatSymbols#getDecimalSeparator decimal separator} is

* substituted.

*

* <li> If the {@code ','} (<tt>'&#92;u002c'</tt>)

* <a name="l10n group">flag</a> is given, then the locale-specific {@linkplain

* java.text.DecimalFormatSymbols#getGroupingSeparator grouping separator} is

* inserted by scanning the integer part of the string from least significant

* to most significant digits and inserting a separator at intervals defined by

* the locale's {@linkplain java.text.DecimalFormat#getGroupingSize() grouping

* size}.

*

* <li> If the {@code '0'} flag is given, then the locale-specific {@linkplain

* java.text.DecimalFormatSymbols#getZeroDigit() zero digits} are inserted

* after the sign character, if any, and before the first non-zero digit, until

* the length of the string is equal to the requested field width.

*

* <li> If the value is negative and the {@code '('} flag is given, then a

* {@code '('} (<tt>'&#92;u0028'</tt>) is prepended and a {@code ')'}

* (<tt>'&#92;u0029'</tt>) is appended.

*

* <li> If the value is negative (or floating-point negative zero) and

* {@code '('} flag is not given, then a {@code '-'} (<tt>'&#92;u002d'</tt>)

* is prepended.

*

* <li> If the {@code '+'} flag is given and the value is positive or zero (or

* floating-point positive zero), then a {@code '+'} (<tt>'&#92;u002b'</tt>)

* will be prepended.

*

* </ol>

*

* <p> If the value is NaN or positive infinity the literal strings "NaN" or

* "Infinity" respectively, will be output. If the value is negative infinity,

* then the output will be "(Infinity)" if the {@code '('} flag is given

* otherwise the output will be "-Infinity". These values are not localized.

*

* <p><a name="dnint"><b> Byte, Short, Integer, and Long </b></a>

*

* <p> The following conversions may be applied to {@code byte}, {@link Byte},

* {@code short}, {@link Short}, {@code int} and {@link Integer},

* {@code long}, and {@link Long}.

*

* <table cellpadding=5 summary="IntConv">

*

* <tr><td valign="top"> {@code 'd'}

* <td valign="top"> <tt>'&#92;u0054'</tt>

* <td> Formats the argument as a decimal integer. The <a

* href="#l10n algorithm">localization algorithm</a> is applied.

*

* <p> If the {@code '0'} flag is given and the value is negative, then

* the zero padding will occur after the sign.

*

* <p> If the {@code '#'} flag is given then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <tr><td valign="top"> {@code 'o'}

* <td valign="top"> <tt>'&#92;u006f'</tt>

* <td> Formats the argument as an integer in base eight. No localization

* is applied.

*

* <p> If <i>x</i> is negative then the result will be an unsigned value

* generated by adding 2ⁿ to the value where {@code n} is the

* number of bits in the type as returned by the static {@code SIZE} field

* in the {@linkplain Byte#SIZE Byte}, {@linkplain Short#SIZE Short},

* {@linkplain Integer#SIZE Integer}, or {@linkplain Long#SIZE Long}

* classes as appropriate.

*

* <p> If the {@code '#'} flag is given then the output will always begin

* with the radix indicator {@code '0'}.

*

* <p> If the {@code '0'} flag is given then the output will be padded

* with leading zeros to the field width following any indication of sign.

*

* <p> If {@code '('}, {@code '+'}, '&nbsp&nbsp;', or {@code ','} flags

* are given then a {@link FormatFlagsConversionMismatchException} will be

* thrown.

*

* <tr><td valign="top"> {@code 'x'}

* <td valign="top"> <tt>'&#92;u0078'</tt>

* <td> Formats the argument as an integer in base sixteen. No

* localization is applied.

*

* <p> If <i>x</i> is negative then the result will be an unsigned value

* generated by adding 2ⁿ to the value where {@code n} is the

* number of bits in the type as returned by the static {@code SIZE} field

* in the {@linkplain Byte#SIZE Byte}, {@linkplain Short#SIZE Short},

* {@linkplain Integer#SIZE Integer}, or {@linkplain Long#SIZE Long}

* classes as appropriate.

*

* <p> If the {@code '#'} flag is given then the output will always begin

* with the radix indicator {@code "0x"}.

*

* <p> If the {@code '0'} flag is given then the output will be padded to

* the field width with leading zeros after the radix indicator or sign (if

* present).

*

* <p> If {@code '('}, <tt>'&nbsp;&nbsp;'</tt>, {@code '+'}, or

* {@code ','} flags are given then a {@link

* FormatFlagsConversionMismatchException} will be thrown.

*

* <tr><td valign="top"> {@code 'X'}

* <td valign="top"> <tt>'&#92;u0058'</tt>

* <td> The upper-case variant of {@code 'x'}. The entire string

* representing the number will be converted to {@linkplain

* String#toUpperCase upper case} including the {@code 'x'} (if any) and

* all hexadecimal digits {@code 'a'} - {@code 'f'}

* (<tt>'&#92;u0061'</tt> - <tt>'&#92;u0066'</tt>).

*

* </table>

*

* <p> If the conversion is {@code 'o'}, {@code 'x'}, or {@code 'X'} and

* both the {@code '#'} and the {@code '0'} flags are given, then result will

* contain the radix indicator ({@code '0'} for octal and {@code "0x"} or

* {@code "0X"} for hexadecimal), some number of zeros (based on the width),

* and the value.

*

* <p> If the {@code '-'} flag is not given, then the space padding will occur

* before the sign.

*

* <p> The following <a name="intFlags">flags</a> apply to numeric integral

* conversions:

*

* <table cellpadding=5 summary="intFlags">

*

* <tr><td valign="top"> {@code '+'}

* <td valign="top"> <tt>'&#92;u002b'</tt>

* <td> Requires the output to include a positive sign for all positive

* numbers. If this flag is not given then only negative values will

* include a sign.

*

* <p> If both the {@code '+'} and <tt>'&nbsp;&nbsp;'</tt> flags are given

* then an {@link IllegalFormatFlagsException} will be thrown.

*

* <tr><td valign="top"> <tt>'&nbsp;&nbsp;'</tt>

* <td valign="top"> <tt>'&#92;u0020'</tt>

* <td> Requires the output to include a single extra space

* (<tt>'&#92;u0020'</tt>) for non-negative values.

*

* <p> If both the {@code '+'} and <tt>'&nbsp;&nbsp;'</tt> flags are given

* then an {@link IllegalFormatFlagsException} will be thrown.

*

* <tr><td valign="top"> {@code '0'}

* <td valign="top"> <tt>'&#92;u0030'</tt>

* <td> Requires the output to be padded with leading {@linkplain

* java.text.DecimalFormatSymbols#getZeroDigit zeros} to the minimum field

* width following any sign or radix indicator except when converting NaN

* or infinity. If the width is not provided, then a {@link

* MissingFormatWidthException} will be thrown.

*

* <p> If both the {@code '-'} and {@code '0'} flags are given then an

* {@link IllegalFormatFlagsException} will be thrown.

*