icuSources/i18n/unicode/plurfmt.h

   1 // © 2016 and later: Unicode, Inc. and others.
   2 // License & terms of use: http://www.unicode.org/copyright.html
   3 /*
   4 *******************************************************************************
   5 * Copyright (C) 2007-2014, International Business Machines Corporation and
   6 * others. All Rights Reserved.
   7 *******************************************************************************
   8 *
   9
  10 * File PLURFMT.H
  11 ********************************************************************************
  12 */
  13
  14 #ifndef PLURFMT
  15 #define PLURFMT
  16
  17 #include "unicode/utypes.h"
  18
  19 #if U_SHOW_CPLUSPLUS_API
  20
  21 /**
  22  * \file
  23  * \brief C++ API: PluralFormat object
  24  */
  25
  26 #if !UCONFIG_NO_FORMATTING
  27
  28 #include "unicode/messagepattern.h"
  29 #include "unicode/numfmt.h"
  30 #include "unicode/plurrule.h"
  31
  32 U_NAMESPACE_BEGIN
  33
  34 class Hashtable;
  35 class NFRule;
  36
  37 /**
  38  * <p>
  39  * <code>PluralFormat</code> supports the creation of internationalized
  40  * messages with plural inflection. It is based on <i>plural
  41  * selection</i>, i.e. the caller specifies messages for each
  42  * plural case that can appear in the user's language and the
  43  * <code>PluralFormat</code> selects the appropriate message based on
  44  * the number.
  45  * </p>
  46  * <h4>The Problem of Plural Forms in Internationalized Messages</h4>
  47  * <p>
  48  * Different languages have different ways to inflect
  49  * plurals. Creating internationalized messages that include plural
  50  * forms is only feasible when the framework is able to handle plural
  51  * forms of <i>all</i> languages correctly. <code>ChoiceFormat</code>
  52  * doesn't handle this well, because it attaches a number interval to
  53  * each message and selects the message whose interval contains a
  54  * given number. This can only handle a finite number of
  55  * intervals. But in some languages, like Polish, one plural case
  56  * applies to infinitely many intervals (e.g., the plural case applies to
  57  * numbers ending with 2, 3, or 4 except those ending with 12, 13, or
  58  * 14). Thus <code>ChoiceFormat</code> is not adequate.
  59  * </p><p>
  60  * <code>PluralFormat</code> deals with this by breaking the problem
  61  * into two parts:
  62  * <ul>
  63  * <li>It uses <code>PluralRules</code> that can define more complex
  64  *     conditions for a plural case than just a single interval. These plural
  65  *     rules define both what plural cases exist in a language, and to
  66  *     which numbers these cases apply.
  67  * <li>It provides predefined plural rules for many languages. Thus, the programmer
  68  *     need not worry about the plural cases of a language and
  69  *     does not have to define the plural cases; they can simply
  70  *     use the predefined keywords. The whole plural formatting of messages can
  71  *     be done using localized patterns from resource bundles. For predefined plural
  72  *     rules, see the CLDR <i>Language Plural Rules</i> page at
  73  *    http://unicode.org/repos/cldr-tmp/trunk/diff/supplemental/language_plural_rules.html
  74  * </ul>
  75  * </p>
  76  * <h4>Usage of <code>PluralFormat</code></h4>
  77  * <p>Note: Typically, plural formatting is done via <code>MessageFormat</code>
  78  * with a <code>plural</code> argument type,
  79  * rather than using a stand-alone <code>PluralFormat</code>.
  80  * </p><p>
  81  * This discussion assumes that you use <code>PluralFormat</code> with
  82  * a predefined set of plural rules. You can create one using one of
  83  * the constructors that takes a <code>locale</code> object. To
  84  * specify the message pattern, you can either pass it to the
  85  * constructor or set it explicitly using the
  86  * <code>applyPattern()</code> method. The <code>format()</code>
  87  * method takes a number object and selects the message of the
  88  * matching plural case. This message will be returned.
  89  * </p>
  90  * <h5>Patterns and Their Interpretation</h5>
  91  * <p>
  92  * The pattern text defines the message output for each plural case of the
  93  * specified locale. Syntax:
  94  * <pre>
  95  * pluralStyle = [offsetValue] (selector '{' message '}')+
  96  * offsetValue = "offset:" number
  97  * selector = explicitValue | keyword
  98  * explicitValue = '=' number  // adjacent, no white space in between
  99  * keyword = [^[[:Pattern_Syntax:][:Pattern_White_Space:]]]+
 100  * message: see {@link MessageFormat}
 101  * </pre>
 102  * Pattern_White_Space between syntax elements is ignored, except
 103  * between the {curly braces} and their sub-message,
 104  * and between the '=' and the number of an explicitValue.
 105  *
 106  * </p><p>
 107  * There are 6 predefined casekeyword in CLDR/ICU - 'zero', 'one', 'two', 'few', 'many' and
 108  * 'other'. You always have to define a message text for the default plural case
 109  * <code>other</code> which is contained in every rule set.
 110  * If you do not specify a message text for a particular plural case, the
 111  * message text of the plural case <code>other</code> gets assigned to this
 112  * plural case.
 113  * </p><p>
 114  * When formatting, the input number is first matched against the explicitValue clauses.
 115  * If there is no exact-number match, then a keyword is selected by calling
 116  * the <code>PluralRules</code> with the input number <em>minus the offset</em>.
 117  * (The offset defaults to 0 if it is omitted from the pattern string.)
 118  * If there is no clause with that keyword, then the "other" clauses is returned.
 119  * </p><p>
 120  * An unquoted pound sign (<code>#</code>) in the selected sub-message
 121  * itself (i.e., outside of arguments nested in the sub-message)
 122  * is replaced by the input number minus the offset.
 123  * The number-minus-offset value is formatted using a
 124  * <code>NumberFormat</code> for the <code>PluralFormat</code>'s locale. If you
 125  * need special number formatting, you have to use a <code>MessageFormat</code>
 126  * and explicitly specify a <code>NumberFormat</code> argument.
 127  * <strong>Note:</strong> That argument is formatting without subtracting the offset!
 128  * If you need a custom format and have a non-zero offset, then you need to pass the
 129  * number-minus-offset value as a separate parameter.
 130  * </p>
 131  * For a usage example, see the {@link MessageFormat} class documentation.
 132  *
 133  * <h4>Defining Custom Plural Rules</h4>
 134  * <p>If you need to use <code>PluralFormat</code> with custom rules, you can
 135  * create a <code>PluralRules</code> object and pass it to
 136  * <code>PluralFormat</code>'s constructor. If you also specify a locale in this
 137  * constructor, this locale will be used to format the number in the message
 138  * texts.
 139  * </p><p>
 140  * For more information about <code>PluralRules</code>, see
 141  * {@link PluralRules}.
 142  * </p>
 143  *
 144  * ported from Java
 145  * @stable ICU 4.0
 146  */
 147
 148 class U_I18N_API PluralFormat : public Format {
 149 public:
 150
 151     /**
 152      * Creates a new cardinal-number <code>PluralFormat</code> for the default locale.
 153      * This locale will be used to get the set of plural rules and for standard
 154      * number formatting.
 155      * @param status  output param set to success/failure code on exit, which
 156      *                must not indicate a failure before the function call.
 157      * @stable ICU 4.0
 158      */
 159     PluralFormat(UErrorCode& status);
 160
 161     /**
 162      * Creates a new cardinal-number <code>PluralFormat</code> for a given locale.
 163      * @param locale the <code>PluralFormat</code> will be configured with
 164      *               rules for this locale. This locale will also be used for
 165      *               standard number formatting.
 166      * @param status output param set to success/failure code on exit, which
 167      *               must not indicate a failure before the function call.
 168      * @stable ICU 4.0
 169      */
 170     PluralFormat(const Locale& locale, UErrorCode& status);
 171
 172     /**
 173      * Creates a new <code>PluralFormat</code> for a given set of rules.
 174      * The standard number formatting will be done using the default locale.
 175      * @param rules   defines the behavior of the <code>PluralFormat</code>
 176      *                object.
 177      * @param status  output param set to success/failure code on exit, which
 178      *                must not indicate a failure before the function call.
 179      * @stable ICU 4.0
 180      */
 181     PluralFormat(const PluralRules& rules, UErrorCode& status);
 182
 183     /**
 184      * Creates a new <code>PluralFormat</code> for a given set of rules.
 185      * The standard number formatting will be done using the given locale.
 186      * @param locale  the default number formatting will be done using this
 187      *                locale.
 188      * @param rules   defines the behavior of the <code>PluralFormat</code>
 189      *                object.
 190      * @param status  output param set to success/failure code on exit, which
 191      *                must not indicate a failure before the function call.
 192      * @stable ICU 4.0
 193          * <p>
 194          * <h4>Sample code</h4>
 195          * \snippet samples/plurfmtsample/plurfmtsample.cpp PluralFormatExample1
 196          * \snippet samples/plurfmtsample/plurfmtsample.cpp PluralFormatExample
 197          * <p>
 198      */
 199     PluralFormat(const Locale& locale, const PluralRules& rules, UErrorCode& status);
 200
 201     /**
 202      * Creates a new <code>PluralFormat</code> for the plural type.
 203      * The standard number formatting will be done using the given locale.
 204      * @param locale  the default number formatting will be done using this
 205      *                locale.
 206      * @param type    The plural type (e.g., cardinal or ordinal).
 207      * @param status  output param set to success/failure code on exit, which
 208      *                must not indicate a failure before the function call.
 209      * @stable ICU 50
 210      */
 211     PluralFormat(const Locale& locale, UPluralType type, UErrorCode& status);
 212
 213     /**
 214      * Creates a new cardinal-number <code>PluralFormat</code> for a given pattern string.
 215      * The default locale will be used to get the set of plural rules and for
 216      * standard number formatting.
 217      * @param  pattern the pattern for this <code>PluralFormat</code>.
 218      *                 errors are returned to status if the pattern is invalid.
 219      * @param status   output param set to success/failure code on exit, which
 220      *                 must not indicate a failure before the function call.
 221      * @stable ICU 4.0
 222      */
 223     PluralFormat(const UnicodeString& pattern, UErrorCode& status);
 224
 225     /**
 226      * Creates a new cardinal-number <code>PluralFormat</code> for a given pattern string and
 227      * locale.
 228      * The locale will be used to get the set of plural rules and for
 229      * standard number formatting.
 230      * @param locale   the <code>PluralFormat</code> will be configured with
 231      *                 rules for this locale. This locale will also be used for
 232      *                 standard number formatting.
 233      * @param pattern  the pattern for this <code>PluralFormat</code>.
 234      *                 errors are returned to status if the pattern is invalid.
 235      * @param status   output param set to success/failure code on exit, which
 236      *                 must not indicate a failure before the function call.
 237      * @stable ICU 4.0
 238      */
 239     PluralFormat(const Locale& locale, const UnicodeString& pattern, UErrorCode& status);
 240
 241     /**
 242      * Creates a new <code>PluralFormat</code> for a given set of rules, a
 243      * pattern and a locale.
 244      * @param rules    defines the behavior of the <code>PluralFormat</code>
 245      *                 object.
 246      * @param pattern  the pattern for this <code>PluralFormat</code>.
 247      *                 errors are returned to status if the pattern is invalid.
 248      * @param status   output param set to success/failure code on exit, which
 249      *                 must not indicate a failure before the function call.
 250      * @stable ICU 4.0
 251      */
 252     PluralFormat(const PluralRules& rules,
 253                  const UnicodeString& pattern,
 254                  UErrorCode& status);
 255
 256     /**
 257      * Creates a new <code>PluralFormat</code> for a given set of rules, a
 258      * pattern and a locale.
 259      * @param locale  the <code>PluralFormat</code> will be configured with
 260      *                rules for this locale. This locale will also be used for
 261      *                standard number formatting.
 262      * @param rules   defines the behavior of the <code>PluralFormat</code>
 263      *                object.
 264      * @param pattern the pattern for this <code>PluralFormat</code>.
 265      *                errors are returned to status if the pattern is invalid.
 266      * @param status  output param set to success/failure code on exit, which
 267      *                must not indicate a failure before the function call.
 268      * @stable ICU 4.0
 269      */
 270     PluralFormat(const Locale& locale,
 271                  const PluralRules& rules,
 272                  const UnicodeString& pattern,
 273                  UErrorCode& status);
 274
 275     /**
 276      * Creates a new <code>PluralFormat</code> for a plural type, a
 277      * pattern and a locale.
 278      * @param locale  the <code>PluralFormat</code> will be configured with
 279      *                rules for this locale. This locale will also be used for
 280      *                standard number formatting.
 281      * @param type    The plural type (e.g., cardinal or ordinal).
 282      * @param pattern the pattern for this <code>PluralFormat</code>.
 283      *                errors are returned to status if the pattern is invalid.
 284      * @param status  output param set to success/failure code on exit, which
 285      *                must not indicate a failure before the function call.
 286      * @stable ICU 50
 287      */
 288     PluralFormat(const Locale& locale,
 289                  UPluralType type,
 290                  const UnicodeString& pattern,
 291                  UErrorCode& status);
 292
 293     /**
 294       * copy constructor.
 295       * @stable ICU 4.0
 296       */
 297     PluralFormat(const PluralFormat& other);
 298
 299     /**
 300      * Destructor.
 301      * @stable ICU 4.0
 302      */
 303     virtual ~PluralFormat();
 304
 305     /**
 306      * Sets the pattern used by this plural format.
 307      * The method parses the pattern and creates a map of format strings
 308      * for the plural rules.
 309      * Patterns and their interpretation are specified in the class description.
 310      *
 311      * @param pattern the pattern for this plural format
 312      *                errors are returned to status if the pattern is invalid.
 313      * @param status  output param set to success/failure code on exit, which
 314      *                must not indicate a failure before the function call.
 315      * @stable ICU 4.0
 316      */
 317     void applyPattern(const UnicodeString& pattern, UErrorCode& status);
 318
 319
 320     using Format::format;
 321
 322     /**
 323      * Formats a plural message for a given number.
 324      *
 325      * @param number  a number for which the plural message should be formatted
 326      *                for. If no pattern has been applied to this
 327      *                <code>PluralFormat</code> object yet, the formatted number
 328      *                will be returned.
 329      * @param status  output param set to success/failure code on exit, which
 330      *                must not indicate a failure before the function call.
 331      * @return        the string containing the formatted plural message.
 332      * @stable ICU 4.0
 333      */
 334     UnicodeString format(int32_t number, UErrorCode& status) const;
 335
 336     /**
 337      * Formats a plural message for a given number.
 338      *
 339      * @param number  a number for which the plural message should be formatted
 340      *                for. If no pattern has been applied to this
 341      *                PluralFormat object yet, the formatted number
 342      *                will be returned.
 343      * @param status  output param set to success or failure code on exit, which
 344      *                must not indicate a failure before the function call.
 345      * @return        the string containing the formatted plural message.
 346      * @stable ICU 4.0
 347      */
 348     UnicodeString format(double number, UErrorCode& status) const;
 349
 350     /**
 351      * Formats a plural message for a given number.
 352      *
 353      * @param number   a number for which the plural message should be formatted
 354      *                 for. If no pattern has been applied to this
 355      *                 <code>PluralFormat</code> object yet, the formatted number
 356      *                 will be returned.
 357      * @param appendTo output parameter to receive result.
 358      *                 result is appended to existing contents.
 359      * @param pos      On input: an alignment field, if desired.
 360      *                 On output: the offsets of the alignment field.
 361      * @param status   output param set to success/failure code on exit, which
 362      *                 must not indicate a failure before the function call.
 363      * @return         the string containing the formatted plural message.
 364      * @stable ICU 4.0
 365      */
 366     UnicodeString& format(int32_t number,
 367                           UnicodeString& appendTo,
 368                           FieldPosition& pos,
 369                           UErrorCode& status) const;
 370
 371     /**
 372      * Formats a plural message for a given number.
 373      *
 374      * @param number   a number for which the plural message should be formatted
 375      *                 for. If no pattern has been applied to this
 376      *                 PluralFormat object yet, the formatted number
 377      *                 will be returned.
 378      * @param appendTo output parameter to receive result.
 379      *                 result is appended to existing contents.
 380      * @param pos      On input: an alignment field, if desired.
 381      *                 On output: the offsets of the alignment field.
 382      * @param status   output param set to success/failure code on exit, which
 383      *                 must not indicate a failure before the function call.
 384      * @return         the string containing the formatted plural message.
 385      * @stable ICU 4.0
 386      */
 387     UnicodeString& format(double number,
 388                           UnicodeString& appendTo,
 389                           FieldPosition& pos,
 390                           UErrorCode& status) const;
 391
 392 #ifndef U_HIDE_DEPRECATED_API
 393     /**
 394      * Sets the locale used by this <code>PluraFormat</code> object.
 395      * Note: Calling this method resets this <code>PluraFormat</code> object,
 396      *     i.e., a pattern that was applied previously will be removed,
 397      *     and the NumberFormat is set to the default number format for
 398      *     the locale.  The resulting format behaves the same as one
 399      *     constructed from {@link #PluralFormat(const Locale& locale, UPluralType type, UErrorCode& status)}
 400      *     with UPLURAL_TYPE_CARDINAL.
 401      * @param locale  the <code>locale</code> to use to configure the formatter.
 402      * @param status  output param set to success/failure code on exit, which
 403      *                must not indicate a failure before the function call.
 404      * @deprecated ICU 50 This method clears the pattern and might create
 405      *             a different kind of PluralRules instance;
 406      *             use one of the constructors to create a new instance instead.
 407      */
 408     void setLocale(const Locale& locale, UErrorCode& status);
 409 #endif  /* U_HIDE_DEPRECATED_API */
 410
 411     /**
 412       * Sets the number format used by this formatter.  You only need to
 413       * call this if you want a different number format than the default
 414       * formatter for the locale.
 415       * @param format  the number format to use.
 416       * @param status  output param set to success/failure code on exit, which
 417       *                must not indicate a failure before the function call.
 418       * @stable ICU 4.0
 419       */
 420     void setNumberFormat(const NumberFormat* format, UErrorCode& status);
 421
 422     /**
 423        * Assignment operator
 424        *
 425        * @param other    the PluralFormat object to copy from.
 426        * @stable ICU 4.0
 427        */
 428     PluralFormat& operator=(const PluralFormat& other);
 429
 430     /**
 431       * Return true if another object is semantically equal to this one.
 432       *
 433       * @param other    the PluralFormat object to be compared with.
 434       * @return         true if other is semantically equal to this.
 435       * @stable ICU 4.0
 436       */
 437     virtual UBool operator==(const Format& other) const;
 438
 439     /**
 440      * Return true if another object is semantically unequal to this one.
 441      *
 442      * @param other    the PluralFormat object to be compared with.
 443      * @return         true if other is semantically unequal to this.
 444      * @stable ICU 4.0
 445      */
 446     virtual UBool operator!=(const Format& other) const;
 447
 448     /**
 449      * Clones this Format object polymorphically.  The caller owns the
 450      * result and should delete it when done.
 451      * @stable ICU 4.0
 452      */
 453     virtual PluralFormat* clone() const;
 454
 455    /**
 456     * Formats a plural message for a number taken from a Formattable object.
 457     *
 458     * @param obj       The object containing a number for which the
 459     *                  plural message should be formatted.
 460     *                  The object must be of a numeric type.
 461     * @param appendTo  output parameter to receive result.
 462     *                  Result is appended to existing contents.
 463     * @param pos       On input: an alignment field, if desired.
 464     *                  On output: the offsets of the alignment field.
 465     * @param status    output param filled with success/failure status.
 466     * @return          Reference to 'appendTo' parameter.
 467     * @stable ICU 4.0
 468     */
 469    UnicodeString& format(const Formattable& obj,
 470                          UnicodeString& appendTo,
 471                          FieldPosition& pos,
 472                          UErrorCode& status) const;
 473
 474    /**
 475     * Returns the pattern from applyPattern() or constructor().
 476     *
 477     * @param  appendTo  output parameter to receive result.
 478      *                  Result is appended to existing contents.
 479     * @return the UnicodeString with inserted pattern.
 480     * @stable ICU 4.0
 481     */
 482    UnicodeString& toPattern(UnicodeString& appendTo);
 483
 484    /**
 485     * This method is not yet supported by <code>PluralFormat</code>.
 486     * <P>
 487     * Before calling, set parse_pos.index to the offset you want to start
 488     * parsing at in the source. After calling, parse_pos.index is the end of
 489     * the text you parsed. If error occurs, index is unchanged.
 490     * <P>
 491     * When parsing, leading whitespace is discarded (with a successful parse),
 492     * while trailing whitespace is left as is.
 493     * <P>
 494     * See Format::parseObject() for more.
 495     *
 496     * @param source    The string to be parsed into an object.
 497     * @param result    Formattable to be set to the parse result.
 498     *                  If parse fails, return contents are undefined.
 499     * @param parse_pos The position to start parsing at. Upon return
 500     *                  this param is set to the position after the
 501     *                  last character successfully parsed. If the
 502     *                  source is not parsed successfully, this param
 503     *                  will remain unchanged.
 504     * @stable ICU 4.0
 505     */
 506    virtual void parseObject(const UnicodeString& source,
 507                             Formattable& result,
 508                             ParsePosition& parse_pos) const;
 509
 510     /**
 511      * ICU "poor man's RTTI", returns a UClassID for this class.
 512      *
 513      * @stable ICU 4.0
 514      *
 515      */
 516     static UClassID U_EXPORT2 getStaticClassID(void);
 517
 518     /**
 519      * ICU "poor man's RTTI", returns a UClassID for the actual class.
 520      *
 521      * @stable ICU 4.0
 522      */
 523      virtual UClassID getDynamicClassID() const;
 524
 525 private:
 526      /**
 527       * @internal (private)
 528       */
 529     class U_I18N_API PluralSelector : public UMemory {
 530       public:
 531         virtual ~PluralSelector();
 532         /**
 533          * Given a number, returns the appropriate PluralFormat keyword.
 534          *
 535          * @param context worker object for the selector.
 536          * @param number The number to be plural-formatted.
 537          * @param ec Error code.
 538          * @return The selected PluralFormat keyword.
 539          * @internal (private)
 540          */
 541         virtual UnicodeString select(void *context, double number, UErrorCode& ec) const = 0;
 542     };
 543
 544     class U_I18N_API PluralSelectorAdapter : public PluralSelector {
 545       public:
 546         PluralSelectorAdapter() : pluralRules(NULL) {
 547         }
 548
 549         virtual ~PluralSelectorAdapter();
 550
 551         virtual UnicodeString select(void *context, double number, UErrorCode& /*ec*/) const;
 552
 553         void reset();
 554
 555         PluralRules* pluralRules;
 556     };
 557
 558     Locale  locale;
 559     MessagePattern msgPattern;
 560     NumberFormat*  numberFormat;
 561     double offset;
 562     PluralSelectorAdapter pluralRulesWrapper;
 563
 564     PluralFormat();   // default constructor not implemented
 565     void init(const PluralRules* rules, UPluralType type, UErrorCode& status);
 566     /**
 567      * Copies dynamically allocated values (pointer fields).
 568      * Others are copied using their copy constructors and assignment operators.
 569      */
 570     void copyObjects(const PluralFormat& other);
 571
 572     UnicodeString& format(const Formattable& numberObject, double number,
 573                           UnicodeString& appendTo,
 574                           FieldPosition& pos,
 575                           UErrorCode& status) const;
 576
 577     /**
 578      * Finds the PluralFormat sub-message for the given number, or the "other" sub-message.
 579      * @param pattern A MessagePattern.
 580      * @param partIndex the index of the first PluralFormat argument style part.
 581      * @param selector the PluralSelector for mapping the number (minus offset) to a keyword.
 582      * @param context worker object for the selector.
 583      * @param number a number to be matched to one of the PluralFormat argument's explicit values,
 584      *        or mapped via the PluralSelector.
 585      * @param ec ICU error code.
 586      * @return the sub-message start part index.
 587      */
 588     static int32_t findSubMessage(
 589          const MessagePattern& pattern, int32_t partIndex,
 590          const PluralSelector& selector, void *context, double number, UErrorCode& ec); /**< @internal */
 591
 592     void parseType(const UnicodeString& source, const NFRule *rbnfLenientScanner,
 593         Formattable& result, FieldPosition& pos) const;
 594
 595     friend class MessageFormat;
 596     friend class NFRule;
 597 };
 598
 599 U_NAMESPACE_END
 600
 601 #endif /* #if !UCONFIG_NO_FORMATTING */
 602
 603 #endif /* U_SHOW_CPLUSPLUS_API */
 604
 605 #endif // _PLURFMT
 606 //eof