icuSources/i18n/unicode/msgfmt.h

   1 /*
   2 * Copyright (C) 2007-2013, International Business Machines Corporation and
   3 * others. All Rights Reserved.
   4 ********************************************************************************
   5 *
   6 * File MSGFMT.H
   7 *
   8 * Modification History:
   9 *
  10 *   Date        Name        Description
  11 *   02/19/97    aliu        Converted from java.
  12 *   03/20/97    helena      Finished first cut of implementation.
  13 *   07/22/98    stephen     Removed operator!= (defined in Format)
  14 *   08/19/2002  srl         Removing Javaisms
  15 *******************************************************************************/
  16
  17 #ifndef MSGFMT_H
  18 #define MSGFMT_H
  19
  20 #include "unicode/utypes.h"
  21
  22 /**
  23  * \file
  24  * \brief C++ API: Formats messages in a language-neutral way.
  25  */
  26
  27 #if !UCONFIG_NO_FORMATTING
  28
  29 #include "unicode/format.h"
  30 #include "unicode/locid.h"
  31 #include "unicode/messagepattern.h"
  32 #include "unicode/parseerr.h"
  33 #include "unicode/plurfmt.h"
  34 #include "unicode/plurrule.h"
  35
  36 U_CDECL_BEGIN
  37 // Forward declaration.
  38 struct UHashtable;
  39 typedef struct UHashtable UHashtable; /**< @internal */
  40 U_CDECL_END
  41
  42 U_NAMESPACE_BEGIN
  43
  44 class AppendableWrapper;
  45 class DateFormat;
  46 class NumberFormat;
  47
  48 /**
  49  * <p>MessageFormat prepares strings for display to users,
  50  * with optional arguments (variables/placeholders).
  51  * The arguments can occur in any order, which is necessary for translation
  52  * into languages with different grammars.
  53  *
  54  * <p>A MessageFormat is constructed from a <em>pattern</em> string
  55  * with arguments in {curly braces} which will be replaced by formatted values.
  56  *
  57  * <p><code>MessageFormat</code> differs from the other <code>Format</code>
  58  * classes in that you create a <code>MessageFormat</code> object with one
  59  * of its constructors (not with a <code>createInstance</code> style factory
  60  * method). Factory methods aren't necessary because <code>MessageFormat</code>
  61  * itself doesn't implement locale-specific behavior. Any locale-specific
  62  * behavior is defined by the pattern that you provide and the
  63  * subformats used for inserted arguments.
  64  *
  65  * <p>Arguments can be named (using identifiers) or numbered (using small ASCII-digit integers).
  66  * Some of the API methods work only with argument numbers and throw an exception
  67  * if the pattern has named arguments (see {@link #usesNamedArguments()}).
  68  *
  69  * <p>An argument might not specify any format type. In this case,
  70  * a Number value is formatted with a default (for the locale) NumberFormat,
  71  * a Date value is formatted with a default (for the locale) DateFormat,
  72  * and for any other value its toString() value is used.
  73  *
  74  * <p>An argument might specify a "simple" type for which the specified
  75  * Format object is created, cached and used.
  76  *
  77  * <p>An argument might have a "complex" type with nested MessageFormat sub-patterns.
  78  * During formatting, one of these sub-messages is selected according to the argument value
  79  * and recursively formatted.
  80  *
  81  * <p>After construction, a custom Format object can be set for
  82  * a top-level argument, overriding the default formatting and parsing behavior
  83  * for that argument.
  84  * However, custom formatting can be achieved more simply by writing
  85  * a typeless argument in the pattern string
  86  * and supplying it with a preformatted string value.
  87  *
  88  * <p>When formatting, MessageFormat takes a collection of argument values
  89  * and writes an output string.
  90  * The argument values may be passed as an array
  91  * (when the pattern contains only numbered arguments)
  92  * or as an array of names and and an array of arguments (which works for both named
  93  * and numbered arguments).
  94  *
  95  * <p>Each argument is matched with one of the input values by array index or argument name
  96  * and formatted according to its pattern specification
  97  * (or using a custom Format object if one was set).
  98  * A numbered pattern argument is matched with an argument name that contains that number
  99  * as an ASCII-decimal-digit string (without leading zero).
 100  *
 101  * <h4><a name="patterns">Patterns and Their Interpretation</a></h4>
 102  *
 103  * <code>MessageFormat</code> uses patterns of the following form:
 104  * <pre>
 105  * message = messageText (argument messageText)*
 106  * argument = noneArg | simpleArg | complexArg
 107  * complexArg = choiceArg | pluralArg | selectArg | selectordinalArg
 108  *
 109  * noneArg = '{' argNameOrNumber '}'
 110  * simpleArg = '{' argNameOrNumber ',' argType [',' argStyle] '}'
 111  * choiceArg = '{' argNameOrNumber ',' "choice" ',' choiceStyle '}'
 112  * pluralArg = '{' argNameOrNumber ',' "plural" ',' pluralStyle '}'
 113  * selectArg = '{' argNameOrNumber ',' "select" ',' selectStyle '}'
 114  * selectordinalArg = '{' argNameOrNumber ',' "selectordinal" ',' pluralStyle '}'
 115  *
 116  * choiceStyle: see {@link ChoiceFormat}
 117  * pluralStyle: see {@link PluralFormat}
 118  * selectStyle: see {@link SelectFormat}
 119  *
 120  * argNameOrNumber = argName | argNumber
 121  * argName = [^[[:Pattern_Syntax:][:Pattern_White_Space:]]]+
 122  * argNumber = '0' | ('1'..'9' ('0'..'9')*)
 123  *
 124  * argType = "number" | "date" | "time" | "spellout" | "ordinal" | "duration"
 125  * argStyle = "short" | "medium" | "long" | "full" | "integer" | "currency" | "percent" | argStyleText
 126  * </pre>
 127  *
 128  * <ul>
 129  *   <li>messageText can contain quoted literal strings including syntax characters.
 130  *       A quoted literal string begins with an ASCII apostrophe and a syntax character
 131  *       (usually a {curly brace}) and continues until the next single apostrophe.
 132  *       A double ASCII apostrohpe inside or outside of a quoted string represents
 133  *       one literal apostrophe.
 134  *   <li>Quotable syntax characters are the {curly braces} in all messageText parts,
 135  *       plus the '#' sign in a messageText immediately inside a pluralStyle,
 136  *       and the '|' symbol in a messageText immediately inside a choiceStyle.
 137  *   <li>See also {@link #UMessagePatternApostropheMode}
 138  *   <li>In argStyleText, every single ASCII apostrophe begins and ends quoted literal text,
 139  *       and unquoted {curly braces} must occur in matched pairs.
 140  * </ul>
 141  *
 142  * <p>Recommendation: Use the real apostrophe (single quote) character
 143  * \htmlonly&#x2019;\endhtmlonly (U+2019) for
 144  * human-readable text, and use the ASCII apostrophe ' (U+0027)
 145  * only in program syntax, like quoting in MessageFormat.
 146  * See the annotations for U+0027 Apostrophe in The Unicode Standard.
 147  *
 148  * <p>The <code>choice</code> argument type is deprecated.
 149  * Use <code>plural</code> arguments for proper plural selection,
 150  * and <code>select</code> arguments for simple selection among a fixed set of choices.
 151  *
 152  * <p>The <code>argType</code> and <code>argStyle</code> values are used to create
 153  * a <code>Format</code> instance for the format element. The following
 154  * table shows how the values map to Format instances. Combinations not
 155  * shown in the table are illegal. Any <code>argStyleText</code> must
 156  * be a valid pattern string for the Format subclass used.
 157  *
 158  * <p><table border=1>
 159  *    <tr>
 160  *       <th>argType
 161  *       <th>argStyle
 162  *       <th>resulting Format object
 163  *    <tr>
 164  *       <td colspan=2><i>(none)</i>
 165  *       <td><code>null</code>
 166  *    <tr>
 167  *       <td rowspan=5><code>number</code>
 168  *       <td><i>(none)</i>
 169  *       <td><code>NumberFormat.createInstance(getLocale(), status)</code>
 170  *    <tr>
 171  *       <td><code>integer</code>
 172  *       <td><code>NumberFormat.createInstance(getLocale(), kNumberStyle, status)</code>
 173  *    <tr>
 174  *       <td><code>currency</code>
 175  *       <td><code>NumberFormat.createCurrencyInstance(getLocale(), status)</code>
 176  *    <tr>
 177  *       <td><code>percent</code>
 178  *       <td><code>NumberFormat.createPercentInstance(getLocale(), status)</code>
 179  *    <tr>
 180  *       <td><i>argStyleText</i>
 181  *       <td><code>new DecimalFormat(argStyleText, new DecimalFormatSymbols(getLocale(), status), status)</code>
 182  *    <tr>
 183  *       <td rowspan=6><code>date</code>
 184  *       <td><i>(none)</i>
 185  *       <td><code>DateFormat.createDateInstance(kDefault, getLocale(), status)</code>
 186  *    <tr>
 187  *       <td><code>short</code>
 188  *       <td><code>DateFormat.createDateInstance(kShort, getLocale(), status)</code>
 189  *    <tr>
 190  *       <td><code>medium</code>
 191  *       <td><code>DateFormat.createDateInstance(kDefault, getLocale(), status)</code>
 192  *    <tr>
 193  *       <td><code>long</code>
 194  *       <td><code>DateFormat.createDateInstance(kLong, getLocale(), status)</code>
 195  *    <tr>
 196  *       <td><code>full</code>
 197  *       <td><code>DateFormat.createDateInstance(kFull, getLocale(), status)</code>
 198  *    <tr>
 199  *       <td><i>argStyleText</i>
 200  *       <td><code>new SimpleDateFormat(argStyleText, getLocale(), status)
 201  *    <tr>
 202  *       <td rowspan=6><code>time</code>
 203  *       <td><i>(none)</i>
 204  *       <td><code>DateFormat.createTimeInstance(kDefault, getLocale(), status)</code>
 205  *    <tr>
 206  *       <td><code>short</code>
 207  *       <td><code>DateFormat.createTimeInstance(kShort, getLocale(), status)</code>
 208  *    <tr>
 209  *       <td><code>medium</code>
 210  *       <td><code>DateFormat.createTimeInstance(kDefault, getLocale(), status)</code>
 211  *    <tr>
 212  *       <td><code>long</code>
 213  *       <td><code>DateFormat.createTimeInstance(kLong, getLocale(), status)</code>
 214  *    <tr>
 215  *       <td><code>full</code>
 216  *       <td><code>DateFormat.createTimeInstance(kFull, getLocale(), status)</code>
 217  *    <tr>
 218  *       <td><i>argStyleText</i>
 219  *       <td><code>new SimpleDateFormat(argStyleText, getLocale(), status)
 220  *    <tr>
 221  *       <td><code>spellout</code>
 222  *       <td><i>argStyleText (optional)</i>
 223  *       <td><code>new RuleBasedNumberFormat(URBNF_SPELLOUT, getLocale(), status)
 224  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 225  *    <tr>
 226  *       <td><code>ordinal</code>
 227  *       <td><i>argStyleText (optional)</i>
 228  *       <td><code>new RuleBasedNumberFormat(URBNF_ORDINAL, getLocale(), status)
 229  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 230  *    <tr>
 231  *       <td><code>duration</code>
 232  *       <td><i>argStyleText (optional)</i>
 233  *       <td><code>new RuleBasedNumberFormat(URBNF_DURATION, getLocale(), status)
 234  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 235  * </table>
 236  * <p>
 237  *
 238  * <h4>Usage Information</h4>
 239  *
 240  * <p>Here are some examples of usage:
 241  * Example 1:
 242  *
 243  * <pre>
 244  * \code
 245  *     UErrorCode success = U_ZERO_ERROR;
 246  *     GregorianCalendar cal(success);
 247  *     Formattable arguments[] = {
 248  *         7L,
 249  *         Formattable( (Date) cal.getTime(success), Formattable::kIsDate),
 250  *         "a disturbance in the Force"
 251  *     };
 252  *
 253  *     UnicodeString result;
 254  *     MessageFormat::format(
 255  *          "At {1,time} on {1,date}, there was {2} on planet {0,number}.",
 256  *          arguments, 3, result, success );
 257  *
 258  *     cout << "result: " << result << endl;
 259  *     //<output>: At 4:34:20 PM on 23-Mar-98, there was a disturbance
 260  *     //             in the Force on planet 7.
 261  * \endcode
 262  * </pre>
 263  *
 264  * Typically, the message format will come from resources, and the
 265  * arguments will be dynamically set at runtime.
 266  *
 267  * <p>Example 2:
 268  *
 269  * <pre>
 270  *  \code
 271  *     success = U_ZERO_ERROR;
 272  *     Formattable testArgs[] = {3L, "MyDisk"};
 273  *
 274  *     MessageFormat form(
 275  *         "The disk \"{1}\" contains {0} file(s).", success );
 276  *
 277  *     UnicodeString string;
 278  *     FieldPosition fpos = 0;
 279  *     cout << "format: " << form.format(testArgs, 2, string, fpos, success ) << endl;
 280  *
 281  *     // output, with different testArgs:
 282  *     // output: The disk "MyDisk" contains 0 file(s).
 283  *     // output: The disk "MyDisk" contains 1 file(s).
 284  *     // output: The disk "MyDisk" contains 1,273 file(s).
 285  *  \endcode
 286  *  </pre>
 287  *
 288  *
 289  * <p>For messages that include plural forms, you can use a plural argument:
 290  * <pre>
 291  * \code
 292  *  success = U_ZERO_ERROR;
 293  *  MessageFormat msgFmt(
 294  *       "{num_files, plural, "
 295  *       "=0{There are no files on disk \"{disk_name}\".}"
 296  *       "=1{There is one file on disk \"{disk_name}\".}"
 297  *       "other{There are # files on disk \"{disk_name}\".}}",
 298  *      Locale("en"),
 299  *      success);
 300  *  FieldPosition fpos = 0;
 301  *  Formattable testArgs[] = {0L, "MyDisk"};
 302  *  UnicodeString testArgsNames[] = {"num_files", "disk_name"};
 303  *  UnicodeString result;
 304  *  cout << msgFmt.format(testArgs, testArgsNames, 2, result, fpos, 0, success);
 305  *  testArgs[0] = 3L;
 306  *  cout << msgFmt.format(testArgs, testArgsNames, 2, result, fpos, 0, success);
 307  * \endcode
 308  * <em>output</em>:
 309  * There are no files on disk "MyDisk".
 310  * There are 3 files on "MyDisk".
 311  * </pre>
 312  * See {@link PluralFormat} and {@link PluralRules} for details.
 313  *
 314  * <h4><a name="synchronization">Synchronization</a></h4>
 315  *
 316  * <p>MessageFormats are not synchronized.
 317  * It is recommended to create separate format instances for each thread.
 318  * If multiple threads access a format concurrently, it must be synchronized
 319  * externally.
 320  *
 321  * @stable ICU 2.0
 322  */
 323 class U_I18N_API MessageFormat : public Format {
 324 public:
 325 #ifndef U_HIDE_OBSOLETE_API
 326     /**
 327      * Enum type for kMaxFormat.
 328      * @obsolete ICU 3.0.  The 10-argument limit was removed as of ICU 2.6,
 329      * rendering this enum type obsolete.
 330      */
 331     enum EFormatNumber {
 332         /**
 333          * The maximum number of arguments.
 334          * @obsolete ICU 3.0.  The 10-argument limit was removed as of ICU 2.6,
 335          * rendering this constant obsolete.
 336          */
 337         kMaxFormat = 10
 338     };
 339 #endif  /* U_HIDE_OBSOLETE_API */
 340
 341     /**
 342      * Constructs a new MessageFormat using the given pattern and the
 343      * default locale.
 344      *
 345      * @param pattern   Pattern used to construct object.
 346      * @param status    Input/output error code.  If the
 347      *                  pattern cannot be parsed, set to failure code.
 348      * @stable ICU 2.0
 349      */
 350     MessageFormat(const UnicodeString& pattern,
 351                   UErrorCode &status);
 352
 353     /**
 354      * Constructs a new MessageFormat using the given pattern and locale.
 355      * @param pattern   Pattern used to construct object.
 356      * @param newLocale The locale to use for formatting dates and numbers.
 357      * @param status    Input/output error code.  If the
 358      *                  pattern cannot be parsed, set to failure code.
 359      * @stable ICU 2.0
 360      */
 361     MessageFormat(const UnicodeString& pattern,
 362                   const Locale& newLocale,
 363                         UErrorCode& status);
 364     /**
 365      * Constructs a new MessageFormat using the given pattern and locale.
 366      * @param pattern   Pattern used to construct object.
 367      * @param newLocale The locale to use for formatting dates and numbers.
 368      * @param parseError Struct to receive information on the position
 369      *                   of an error within the pattern.
 370      * @param status    Input/output error code.  If the
 371      *                  pattern cannot be parsed, set to failure code.
 372      * @stable ICU 2.0
 373      */
 374     MessageFormat(const UnicodeString& pattern,
 375                   const Locale& newLocale,
 376                   UParseError& parseError,
 377                   UErrorCode& status);
 378     /**
 379      * Constructs a new MessageFormat from an existing one.
 380      * @stable ICU 2.0
 381      */
 382     MessageFormat(const MessageFormat&);
 383
 384     /**
 385      * Assignment operator.
 386      * @stable ICU 2.0
 387      */
 388     const MessageFormat& operator=(const MessageFormat&);
 389
 390     /**
 391      * Destructor.
 392      * @stable ICU 2.0
 393      */
 394     virtual ~MessageFormat();
 395
 396     /**
 397      * Clones this Format object polymorphically.  The caller owns the
 398      * result and should delete it when done.
 399      * @stable ICU 2.0
 400      */
 401     virtual Format* clone(void) const;
 402
 403     /**
 404      * Returns true if the given Format objects are semantically equal.
 405      * Objects of different subclasses are considered unequal.
 406      * @param other  the object to be compared with.
 407      * @return       true if the given Format objects are semantically equal.
 408      * @stable ICU 2.0
 409      */
 410     virtual UBool operator==(const Format& other) const;
 411
 412     /**
 413      * Sets the locale to be used for creating argument Format objects.
 414      * @param theLocale    the new locale value to be set.
 415      * @stable ICU 2.0
 416      */
 417     virtual void setLocale(const Locale& theLocale);
 418
 419     /**
 420      * Gets the locale used for creating argument Format objects.
 421      * format information.
 422      * @return    the locale of the object.
 423      * @stable ICU 2.0
 424      */
 425     virtual const Locale& getLocale(void) const;
 426
 427     /**
 428      * Applies the given pattern string to this message format.
 429      *
 430      * @param pattern   The pattern to be applied.
 431      * @param status    Input/output error code.  If the
 432      *                  pattern cannot be parsed, set to failure code.
 433      * @stable ICU 2.0
 434      */
 435     virtual void applyPattern(const UnicodeString& pattern,
 436                               UErrorCode& status);
 437     /**
 438      * Applies the given pattern string to this message format.
 439      *
 440      * @param pattern    The pattern to be applied.
 441      * @param parseError Struct to receive information on the position
 442      *                   of an error within the pattern.
 443      * @param status    Input/output error code.  If the
 444      *                  pattern cannot be parsed, set to failure code.
 445      * @stable ICU 2.0
 446      */
 447     virtual void applyPattern(const UnicodeString& pattern,
 448                              UParseError& parseError,
 449                              UErrorCode& status);
 450
 451     /**
 452      * Sets the UMessagePatternApostropheMode and the pattern used by this message format.
 453      * Parses the pattern and caches Format objects for simple argument types.
 454      * Patterns and their interpretation are specified in the
 455      * <a href="#patterns">class description</a>.
 456      * <p>
 457      * This method is best used only once on a given object to avoid confusion about the mode,
 458      * and after constructing the object with an empty pattern string to minimize overhead.
 459      *
 460      * @param pattern    The pattern to be applied.
 461      * @param aposMode   The new apostrophe mode.
 462      * @param parseError Struct to receive information on the position
 463      *                   of an error within the pattern.
 464      *                   Can be NULL.
 465      * @param status    Input/output error code.  If the
 466      *                  pattern cannot be parsed, set to failure code.
 467      * @stable ICU 4.8
 468      */
 469     virtual void applyPattern(const UnicodeString& pattern,
 470                               UMessagePatternApostropheMode aposMode,
 471                               UParseError* parseError,
 472                               UErrorCode& status);
 473
 474     /**
 475      * @return this instance's UMessagePatternApostropheMode.
 476      * @stable ICU 4.8
 477      */
 478     UMessagePatternApostropheMode getApostropheMode() const {
 479         return msgPattern.getApostropheMode();
 480     }
 481
 482     /**
 483      * Returns a pattern that can be used to recreate this object.
 484      *
 485      * @param appendTo  Output parameter to receive the pattern.
 486      *                  Result is appended to existing contents.
 487      * @return          Reference to 'appendTo' parameter.
 488      * @stable ICU 2.0
 489      */
 490     virtual UnicodeString& toPattern(UnicodeString& appendTo) const;
 491
 492     /**
 493      * Sets subformats.
 494      * See the class description about format numbering.
 495      * The caller should not delete the Format objects after this call.
 496      * <EM>The array formatsToAdopt is not itself adopted.</EM> Its
 497      * ownership is retained by the caller. If the call fails because
 498      * memory cannot be allocated, then the formats will be deleted
 499      * by this method, and this object will remain unchanged.
 500      *
 501      * <p>If this format uses named arguments, the new formats are discarded
 502      * and this format remains unchanged.
 503      *
 504      * @stable ICU 2.0
 505      * @param formatsToAdopt    the format to be adopted.
 506      * @param count             the size of the array.
 507      */
 508     virtual void adoptFormats(Format** formatsToAdopt, int32_t count);
 509
 510     /**
 511      * Sets subformats.
 512      * See the class description about format numbering.
 513      * Each item in the array is cloned into the internal array.
 514      * If the call fails because memory cannot be allocated, then this
 515      * object will remain unchanged.
 516      *
 517      * <p>If this format uses named arguments, the new formats are discarded
 518      * and this format remains unchanged.
 519      *
 520      * @stable ICU 2.0
 521      * @param newFormats the new format to be set.
 522      * @param cnt        the size of the array.
 523      */
 524     virtual void setFormats(const Format** newFormats, int32_t cnt);
 525
 526
 527     /**
 528      * Sets one subformat.
 529      * See the class description about format numbering.
 530      * The caller should not delete the Format object after this call.
 531      * If the number is over the number of formats already set,
 532      * the item will be deleted and ignored.
 533      *
 534      * <p>If this format uses named arguments, the new format is discarded
 535      * and this format remains unchanged.
 536      *
 537      * @stable ICU 2.0
 538      * @param formatNumber     index of the subformat.
 539      * @param formatToAdopt    the format to be adopted.
 540      */
 541     virtual void adoptFormat(int32_t formatNumber, Format* formatToAdopt);
 542
 543     /**
 544      * Sets one subformat.
 545      * See the class description about format numbering.
 546      * If the number is over the number of formats already set,
 547      * the item will be ignored.
 548      * @param formatNumber     index of the subformat.
 549      * @param format    the format to be set.
 550      * @stable ICU 2.0
 551      */
 552     virtual void setFormat(int32_t formatNumber, const Format& format);
 553
 554     /**
 555      * Gets format names. This function returns formatNames in StringEnumerations
 556      * which can be used with getFormat() and setFormat() to export formattable
 557      * array from current MessageFormat to another.  It is the caller's responsibility
 558      * to delete the returned formatNames.
 559      * @param status  output param set to success/failure code.
 560      * @stable ICU 4.0
 561      */
 562     virtual StringEnumeration* getFormatNames(UErrorCode& status);
 563
 564     /**
 565      * Gets subformat pointer for given format name.
 566      * This function supports both named and numbered
 567      * arguments. If numbered, the formatName is the
 568      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 569      * The returned Format object should not be deleted by the caller,
 570      * nor should the ponter of other object .  The pointer and its
 571      * contents remain valid only until the next call to any method
 572      * of this class is made with this object.
 573      * @param formatName the name or number specifying a format
 574      * @param status  output param set to success/failure code.
 575      * @stable ICU 4.0
 576      */
 577     virtual Format* getFormat(const UnicodeString& formatName, UErrorCode& status);
 578
 579     /**
 580      * Sets one subformat for given format name.
 581      * See the class description about format name.
 582      * This function supports both named and numbered
 583      * arguments-- if numbered, the formatName is the
 584      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 585      * If there is no matched formatName or wrong type,
 586      * the item will be ignored.
 587      * @param formatName  Name of the subformat.
 588      * @param format      the format to be set.
 589      * @param status  output param set to success/failure code.
 590      * @stable ICU 4.0
 591      */
 592     virtual void setFormat(const UnicodeString& formatName, const Format& format, UErrorCode& status);
 593
 594     /**
 595      * Sets one subformat for given format name.
 596      * See the class description about format name.
 597      * This function supports both named and numbered
 598      * arguments-- if numbered, the formatName is the
 599      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 600      * If there is no matched formatName or wrong type,
 601      * the item will be ignored.
 602      * The caller should not delete the Format object after this call.
 603      * @param formatName  Name of the subformat.
 604      * @param formatToAdopt  Format to be adopted.
 605      * @param status      output param set to success/failure code.
 606      * @stable ICU 4.0
 607      */
 608     virtual void adoptFormat(const UnicodeString& formatName, Format* formatToAdopt, UErrorCode& status);
 609
 610     /**
 611      * Gets an array of subformats of this object.  The returned array
 612      * should not be deleted by the caller, nor should the pointers
 613      * within the array.  The array and its contents remain valid only
 614      * until the next call to this format. See the class description
 615      * about format numbering.
 616      *
 617      * @param count output parameter to receive the size of the array
 618      * @return an array of count Format* objects, or NULL if out of
 619      * memory.  Any or all of the array elements may be NULL.
 620      * @stable ICU 2.0
 621      */
 622     virtual const Format** getFormats(int32_t& count) const;
 623
 624
 625     using Format::format;
 626
 627     /**
 628      * Formats the given array of arguments into a user-readable string.
 629      * Does not take ownership of the Formattable* array or its contents.
 630      *
 631      * <p>If this format uses named arguments, appendTo is unchanged and
 632      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 633      *
 634      * @param source    An array of objects to be formatted.
 635      * @param count     The number of elements of 'source'.
 636      * @param appendTo  Output parameter to receive result.
 637      *                  Result is appended to existing contents.
 638      * @param ignore    Not used; inherited from base class API.
 639      * @param status    Input/output error code.  If the
 640      *                  pattern cannot be parsed, set to failure code.
 641      * @return          Reference to 'appendTo' parameter.
 642      * @stable ICU 2.0
 643      */
 644     UnicodeString& format(const Formattable* source,
 645                           int32_t count,
 646                           UnicodeString& appendTo,
 647                           FieldPosition& ignore,
 648                           UErrorCode& status) const;
 649
 650     /**
 651      * Formats the given array of arguments into a user-readable string
 652      * using the given pattern.
 653      *
 654      * <p>If this format uses named arguments, appendTo is unchanged and
 655      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 656      *
 657      * @param pattern   The pattern.
 658      * @param arguments An array of objects to be formatted.
 659      * @param count     The number of elements of 'source'.
 660      * @param appendTo  Output parameter to receive result.
 661      *                  Result is appended to existing contents.
 662      * @param status    Input/output error code.  If the
 663      *                  pattern cannot be parsed, set to failure code.
 664      * @return          Reference to 'appendTo' parameter.
 665      * @stable ICU 2.0
 666      */
 667     static UnicodeString& format(const UnicodeString& pattern,
 668                                  const Formattable* arguments,
 669                                  int32_t count,
 670                                  UnicodeString& appendTo,
 671                                  UErrorCode& status);
 672
 673     /**
 674      * Formats the given array of arguments into a user-readable
 675      * string.  The array must be stored within a single Formattable
 676      * object of type kArray. If the Formattable object type is not of
 677      * type kArray, then returns a failing UErrorCode.
 678      *
 679      * <p>If this format uses named arguments, appendTo is unchanged and
 680      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 681      *
 682      * @param obj       A Formattable of type kArray containing
 683      *                  arguments to be formatted.
 684      * @param appendTo  Output parameter to receive result.
 685      *                  Result is appended to existing contents.
 686      * @param pos       On input: an alignment field, if desired.
 687      *                  On output: the offsets of the alignment field.
 688      * @param status    Input/output error code.  If the
 689      *                  pattern cannot be parsed, set to failure code.
 690      * @return          Reference to 'appendTo' parameter.
 691      * @stable ICU 2.0
 692      */
 693     virtual UnicodeString& format(const Formattable& obj,
 694                                   UnicodeString& appendTo,
 695                                   FieldPosition& pos,
 696                                   UErrorCode& status) const;
 697
 698     /**
 699      * Formats the given array of arguments into a user-defined argument name
 700      * array. This function supports both named and numbered
 701      * arguments-- if numbered, the formatName is the
 702      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 703      *
 704      * @param argumentNames argument name array
 705      * @param arguments An array of objects to be formatted.
 706      * @param count     The number of elements of 'argumentNames' and
 707      *                  arguments.  The number of argumentNames and arguments
 708      *                  must be the same.
 709      * @param appendTo  Output parameter to receive result.
 710      *                  Result is appended to existing contents.
 711      * @param status    Input/output error code.  If the
 712      *                  pattern cannot be parsed, set to failure code.
 713      * @return          Reference to 'appendTo' parameter.
 714      * @stable ICU 4.0
 715      */
 716     UnicodeString& format(const UnicodeString* argumentNames,
 717                           const Formattable* arguments,
 718                           int32_t count,
 719                           UnicodeString& appendTo,
 720                           UErrorCode& status) const;
 721     /**
 722      * Parses the given string into an array of output arguments.
 723      *
 724      * @param source    String to be parsed.
 725      * @param pos       On input, starting position for parse. On output,
 726      *                  final position after parse.  Unchanged if parse
 727      *                  fails.
 728      * @param count     Output parameter to receive the number of arguments
 729      *                  parsed.
 730      * @return an array of parsed arguments.  The caller owns both
 731      * the array and its contents.
 732      * @stable ICU 2.0
 733      */
 734     virtual Formattable* parse(const UnicodeString& source,
 735                                ParsePosition& pos,
 736                                int32_t& count) const;
 737
 738     /**
 739      * Parses the given string into an array of output arguments.
 740      *
 741      * <p>If this format uses named arguments, status is set to
 742      * U_ARGUMENT_TYPE_MISMATCH.
 743      *
 744      * @param source    String to be parsed.
 745      * @param count     Output param to receive size of returned array.
 746      * @param status    Input/output error code.  If the
 747      *                  pattern cannot be parsed, set to failure code.
 748      * @return an array of parsed arguments.  The caller owns both
 749      * the array and its contents. Returns NULL if status is not U_ZERO_ERROR.
 750      *
 751      * @stable ICU 2.0
 752      */
 753     virtual Formattable* parse(const UnicodeString& source,
 754                                int32_t& count,
 755                                UErrorCode& status) const;
 756
 757     /**
 758      * Parses the given string into an array of output arguments
 759      * stored within a single Formattable of type kArray.
 760      *
 761      * @param source    The string to be parsed into an object.
 762      * @param result    Formattable to be set to the parse result.
 763      *                  If parse fails, return contents are undefined.
 764      * @param pos       On input, starting position for parse. On output,
 765      *                  final position after parse.  Unchanged if parse
 766      *                  fails.
 767      * @stable ICU 2.0
 768      */
 769     virtual void parseObject(const UnicodeString& source,
 770                              Formattable& result,
 771                              ParsePosition& pos) const;
 772
 773     /**
 774      * Convert an 'apostrophe-friendly' pattern into a standard
 775      * pattern.  Standard patterns treat all apostrophes as
 776      * quotes, which is problematic in some languages, e.g.
 777      * French, where apostrophe is commonly used.  This utility
 778      * assumes that only an unpaired apostrophe immediately before
 779      * a brace is a true quote.  Other unpaired apostrophes are paired,
 780      * and the resulting standard pattern string is returned.
 781      *
 782      * <p><b>Note</b> it is not guaranteed that the returned pattern
 783      * is indeed a valid pattern.  The only effect is to convert
 784      * between patterns having different quoting semantics.
 785      *
 786      * @param pattern the 'apostrophe-friendly' patttern to convert
 787      * @param status    Input/output error code.  If the pattern
 788      *                  cannot be parsed, the failure code is set.
 789      * @return the standard equivalent of the original pattern
 790      * @stable ICU 3.4
 791      */
 792     static UnicodeString autoQuoteApostrophe(const UnicodeString& pattern,
 793         UErrorCode& status);
 794
 795
 796     /**
 797      * Returns true if this MessageFormat uses named arguments,
 798      * and false otherwise.  See class description.
 799      *
 800      * @return true if named arguments are used.
 801      * @stable ICU 4.0
 802      */
 803     UBool usesNamedArguments() const;
 804
 805
 806 #ifndef U_HIDE_INTERNAL_API
 807     /**
 808      * This API is for ICU internal use only.
 809      * Please do not use it.
 810      *
 811      * Returns argument types count in the parsed pattern.
 812      * Used to distinguish pattern "{0} d" and "d".
 813      *
 814      * @return           The number of formattable types in the pattern
 815      * @internal
 816      */
 817     int32_t getArgTypeCount() const;
 818 #endif  /* U_HIDE_INTERNAL_API */
 819
 820     /**
 821      * Returns a unique class ID POLYMORPHICALLY.  Pure virtual override.
 822      * This method is to implement a simple version of RTTI, since not all
 823      * C++ compilers support genuine RTTI.  Polymorphic operator==() and
 824      * clone() methods call this method.
 825      *
 826      * @return          The class ID for this object. All objects of a
 827      *                  given class have the same class ID.  Objects of
 828      *                  other classes have different class IDs.
 829      * @stable ICU 2.0
 830      */
 831     virtual UClassID getDynamicClassID(void) const;
 832
 833     /**
 834      * Return the class ID for this class.  This is useful only for
 835      * comparing to a return value from getDynamicClassID().  For example:
 836      * <pre>
 837      * .   Base* polymorphic_pointer = createPolymorphicObject();
 838      * .   if (polymorphic_pointer->getDynamicClassID() ==
 839      * .      Derived::getStaticClassID()) ...
 840      * </pre>
 841      * @return          The class ID for all objects of this class.
 842      * @stable ICU 2.0
 843      */
 844     static UClassID U_EXPORT2 getStaticClassID(void);
 845
 846 #ifndef U_HIDE_INTERNAL_API
 847     /**
 848      * Compares two Format objects. This is used for constructing the hash
 849      * tables.
 850      *
 851      * @param left pointer to a Format object. Must not be NULL.
 852      * @param right pointer to a Format object. Must not be NULL.
 853      *
 854      * @return whether the two objects are the same
 855      * @internal
 856      */
 857     static UBool equalFormats(const void* left, const void* right);
 858 #endif  /* U_HIDE_INTERNAL_API */
 859
 860 private:
 861
 862     Locale              fLocale;
 863     MessagePattern      msgPattern;
 864     Format**            formatAliases; // see getFormats
 865     int32_t             formatAliasesCapacity;
 866
 867     MessageFormat(); // default constructor not implemented
 868
 869      /**
 870       * This provider helps defer instantiation of a PluralRules object
 871       * until we actually need to select a keyword.
 872       * For example, if the number matches an explicit-value selector like "=1"
 873       * we do not need any PluralRules.
 874       */
 875     class U_I18N_API PluralSelectorProvider : public PluralFormat::PluralSelector {
 876     public:
 877         PluralSelectorProvider(const MessageFormat &mf, UPluralType type);
 878         virtual ~PluralSelectorProvider();
 879         virtual UnicodeString select(void *ctx, double number, UErrorCode& ec) const;
 880
 881         void reset();
 882     private:
 883         const MessageFormat &msgFormat;
 884         PluralRules* rules;
 885         UPluralType type;
 886     };
 887
 888     /**
 889      * A MessageFormat formats an array of arguments.  Each argument
 890      * has an expected type, based on the pattern.  For example, if
 891      * the pattern contains the subformat "{3,number,integer}", then
 892      * we expect argument 3 to have type Formattable::kLong.  This
 893      * array needs to grow dynamically if the MessageFormat is
 894      * modified.
 895      */
 896     Formattable::Type* argTypes;
 897     int32_t            argTypeCount;
 898     int32_t            argTypeCapacity;
 899
 900     /**
 901      * TRUE if there are different argTypes for the same argument.
 902      * This only matters when the MessageFormat is used in the plain C (umsg_xxx) API
 903      * where the pattern argTypes determine how the va_arg list is read.
 904      */
 905     UBool hasArgTypeConflicts;
 906
 907     // Variable-size array management
 908     UBool allocateArgTypes(int32_t capacity, UErrorCode& status);
 909
 910     /**
 911      * Default Format objects used when no format is specified and a
 912      * numeric or date argument is formatted.  These are volatile
 913      * cache objects maintained only for performance.  They do not
 914      * participate in operator=(), copy constructor(), nor
 915      * operator==().
 916      */
 917     NumberFormat* defaultNumberFormat;
 918     DateFormat*   defaultDateFormat;
 919
 920     UHashtable* cachedFormatters;
 921     UHashtable* customFormatArgStarts;
 922
 923     PluralSelectorProvider pluralProvider;
 924     PluralSelectorProvider ordinalProvider;
 925
 926     /**
 927      * Method to retrieve default formats (or NULL on failure).
 928      * These are semantically const, but may modify *this.
 929      */
 930     const NumberFormat* getDefaultNumberFormat(UErrorCode&) const;
 931     const DateFormat*   getDefaultDateFormat(UErrorCode&) const;
 932
 933     /**
 934      * Finds the word s, in the keyword list and returns the located index.
 935      * @param s the keyword to be searched for.
 936      * @param list the list of keywords to be searched with.
 937      * @return the index of the list which matches the keyword s.
 938      */
 939     static int32_t findKeyword( const UnicodeString& s,
 940                                 const UChar * const *list);
 941
 942     /**
 943      * Thin wrapper around the format(... AppendableWrapper ...) variant.
 944      * Wraps the destination UnicodeString into an AppendableWrapper and
 945      * supplies default values for some other parameters.
 946      */
 947     UnicodeString& format(const Formattable* arguments,
 948                           const UnicodeString *argumentNames,
 949                           int32_t cnt,
 950                           UnicodeString& appendTo,
 951                           FieldPosition* pos,
 952                           UErrorCode& status) const;
 953
 954     /**
 955      * Formats the arguments and writes the result into the
 956      * AppendableWrapper, updates the field position.
 957      *
 958      * @param msgStart      Index to msgPattern part to start formatting from.
 959      * @param plNumber      NULL except when formatting a plural argument sub-message
 960      *                      where a '#' is replaced by the format string for this number.
 961      * @param arguments     The formattable objects array. (Must not be NULL.)
 962      * @param argumentNames NULL if numbered values are used. Otherwise the same
 963      *                      length as "arguments", and each entry is the name of the
 964      *                      corresponding argument in "arguments".
 965      * @param cnt           The length of arguments (and of argumentNames if that is not NULL).
 966      * @param appendTo      Output parameter to receive the result.
 967      *                      The result string is appended to existing contents.
 968      * @param pos           Field position status.
 969      * @param success       The error code status.
 970      */
 971     void format(int32_t msgStart,
 972                 const void *plNumber,
 973                 const Formattable* arguments,
 974                 const UnicodeString *argumentNames,
 975                 int32_t cnt,
 976                 AppendableWrapper& appendTo,
 977                 FieldPosition* pos,
 978                 UErrorCode& success) const;
 979
 980     UnicodeString getArgName(int32_t partIndex);
 981
 982     void setArgStartFormat(int32_t argStart, Format* formatter, UErrorCode& status);
 983
 984     void setCustomArgStartFormat(int32_t argStart, Format* formatter, UErrorCode& status);
 985
 986     int32_t nextTopLevelArgStart(int32_t partIndex) const;
 987
 988     UBool argNameMatches(int32_t partIndex, const UnicodeString& argName, int32_t argNumber);
 989
 990     void cacheExplicitFormats(UErrorCode& status);
 991
 992     Format* createAppropriateFormat(UnicodeString& type,
 993                                     UnicodeString& style,
 994                                     Formattable::Type& formattableType,
 995                                     UParseError& parseError,
 996                                     UErrorCode& ec);
 997
 998     const Formattable* getArgFromListByName(const Formattable* arguments,
 999                                             const UnicodeString *argumentNames,
1000                                             int32_t cnt, UnicodeString& name) const;
1001
1002     Formattable* parse(int32_t msgStart,
1003                        const UnicodeString& source,
1004                        ParsePosition& pos,
1005                        int32_t& count,
1006                        UErrorCode& ec) const;
1007
1008     FieldPosition* updateMetaData(AppendableWrapper& dest, int32_t prevLength,
1009                                   FieldPosition* fp, const Formattable* argId) const;
1010
1011     /**
1012      * Finds the "other" sub-message.
1013      * @param partIndex the index of the first PluralFormat argument style part.
1014      * @return the "other" sub-message start part index.
1015      */
1016     int32_t findOtherSubMessage(int32_t partIndex) const;
1017
1018     /**
1019      * Returns the ARG_START index of the first occurrence of the plural number in a sub-message.
1020      * Returns -1 if it is a REPLACE_NUMBER.
1021      * Returns 0 if there is neither.
1022      */
1023     int32_t findFirstPluralNumberArg(int32_t msgStart, const UnicodeString &argName) const;
1024
1025     Format* getCachedFormatter(int32_t argumentNumber) const;
1026
1027     UnicodeString getLiteralStringUntilNextArgument(int32_t from) const;
1028
1029     void copyObjects(const MessageFormat& that, UErrorCode& ec);
1030
1031     void formatComplexSubMessage(int32_t msgStart,
1032                                  const void *plNumber,
1033                                  const Formattable* arguments,
1034                                  const UnicodeString *argumentNames,
1035                                  int32_t cnt,
1036                                  AppendableWrapper& appendTo,
1037                                  UErrorCode& success) const;
1038
1039     /**
1040      * Convenience method that ought to be in NumberFormat
1041      */
1042     NumberFormat* createIntegerFormat(const Locale& locale, UErrorCode& status) const;
1043
1044     /**
1045      * Returns array of argument types in the parsed pattern
1046      * for use in C API.  Only for the use of umsg_vformat().  Not
1047      * for public consumption.
1048      * @param listCount  Output parameter to receive the size of array
1049      * @return           The array of formattable types in the pattern
1050      */
1051     const Formattable::Type* getArgTypeList(int32_t& listCount) const {
1052         listCount = argTypeCount;
1053         return argTypes;
1054     }
1055
1056     /**
1057      * Resets the internal MessagePattern, and other associated caches.
1058      */
1059     void resetPattern();
1060
1061     /**
1062      * A DummyFormatter that we use solely to store a NULL value. UHash does
1063      * not support storing NULL values.
1064      */
1065     class U_I18N_API DummyFormat : public Format {
1066     public:
1067         virtual UBool operator==(const Format&) const;
1068         virtual Format* clone() const;
1069         virtual UnicodeString& format(const Formattable& obj,
1070                               UnicodeString& appendTo,
1071                               UErrorCode& status) const;
1072         virtual UnicodeString& format(const Formattable&,
1073                                       UnicodeString& appendTo,
1074                                       FieldPosition&,
1075                                       UErrorCode& status) const;
1076         virtual UnicodeString& format(const Formattable& obj,
1077                                       UnicodeString& appendTo,
1078                                       FieldPositionIterator* posIter,
1079                                       UErrorCode& status) const;
1080         virtual void parseObject(const UnicodeString&,
1081                                  Formattable&,
1082                                  ParsePosition&) const;
1083     };
1084
1085     friend class MessageFormatAdapter; // getFormatTypeList() access
1086 };
1087
1088 U_NAMESPACE_END
1089
1090 #endif /* #if !UCONFIG_NO_FORMATTING */
1091
1092 #endif // _MSGFMT
1093 //eof