icuSources/i18n/unicode/msgfmt.h

   1 // © 2016 and later: Unicode, Inc. and others.
   2 // License & terms of use: http://www.unicode.org/copyright.html
   3 /*
   4 * Copyright (C) 2007-2013, International Business Machines Corporation and
   5 * others. All Rights Reserved.
   6 ********************************************************************************
   7 *
   8 * File MSGFMT.H
   9 *
  10 * Modification History:
  11 *
  12 *   Date        Name        Description
  13 *   02/19/97    aliu        Converted from java.
  14 *   03/20/97    helena      Finished first cut of implementation.
  15 *   07/22/98    stephen     Removed operator!= (defined in Format)
  16 *   08/19/2002  srl         Removing Javaisms
  17 *******************************************************************************/
  18
  19 #ifndef MSGFMT_H
  20 #define MSGFMT_H
  21
  22 #include "unicode/utypes.h"
  23
  24 /**
  25  * \file
  26  * \brief C++ API: Formats messages in a language-neutral way.
  27  */
  28
  29 #if !UCONFIG_NO_FORMATTING
  30
  31 #include "unicode/format.h"
  32 #include "unicode/locid.h"
  33 #include "unicode/messagepattern.h"
  34 #include "unicode/parseerr.h"
  35 #include "unicode/plurfmt.h"
  36 #include "unicode/plurrule.h"
  37
  38 U_CDECL_BEGIN
  39 // Forward declaration.
  40 struct UHashtable;
  41 typedef struct UHashtable UHashtable; /**< @internal */
  42 U_CDECL_END
  43
  44 #if U_SHOW_CPLUSPLUS_API
  45 U_NAMESPACE_BEGIN
  46
  47 class AppendableWrapper;
  48 class DateFormat;
  49 class NumberFormat;
  50
  51 /**
  52  * <p>MessageFormat prepares strings for display to users,
  53  * with optional arguments (variables/placeholders).
  54  * The arguments can occur in any order, which is necessary for translation
  55  * into languages with different grammars.
  56  *
  57  * <p>A MessageFormat is constructed from a <em>pattern</em> string
  58  * with arguments in {curly braces} which will be replaced by formatted values.
  59  *
  60  * <p><code>MessageFormat</code> differs from the other <code>Format</code>
  61  * classes in that you create a <code>MessageFormat</code> object with one
  62  * of its constructors (not with a <code>createInstance</code> style factory
  63  * method). Factory methods aren't necessary because <code>MessageFormat</code>
  64  * itself doesn't implement locale-specific behavior. Any locale-specific
  65  * behavior is defined by the pattern that you provide and the
  66  * subformats used for inserted arguments.
  67  *
  68  * <p>Arguments can be named (using identifiers) or numbered (using small ASCII-digit integers).
  69  * Some of the API methods work only with argument numbers and throw an exception
  70  * if the pattern has named arguments (see {@link #usesNamedArguments()}).
  71  *
  72  * <p>An argument might not specify any format type. In this case,
  73  * a Number value is formatted with a default (for the locale) NumberFormat,
  74  * a Date value is formatted with a default (for the locale) DateFormat,
  75  * and for any other value its toString() value is used.
  76  *
  77  * <p>An argument might specify a "simple" type for which the specified
  78  * Format object is created, cached and used.
  79  *
  80  * <p>An argument might have a "complex" type with nested MessageFormat sub-patterns.
  81  * During formatting, one of these sub-messages is selected according to the argument value
  82  * and recursively formatted.
  83  *
  84  * <p>After construction, a custom Format object can be set for
  85  * a top-level argument, overriding the default formatting and parsing behavior
  86  * for that argument.
  87  * However, custom formatting can be achieved more simply by writing
  88  * a typeless argument in the pattern string
  89  * and supplying it with a preformatted string value.
  90  *
  91  * <p>When formatting, MessageFormat takes a collection of argument values
  92  * and writes an output string.
  93  * The argument values may be passed as an array
  94  * (when the pattern contains only numbered arguments)
  95  * or as an array of names and and an array of arguments (which works for both named
  96  * and numbered arguments).
  97  *
  98  * <p>Each argument is matched with one of the input values by array index or argument name
  99  * and formatted according to its pattern specification
 100  * (or using a custom Format object if one was set).
 101  * A numbered pattern argument is matched with an argument name that contains that number
 102  * as an ASCII-decimal-digit string (without leading zero).
 103  *
 104  * <h4><a name="patterns">Patterns and Their Interpretation</a></h4>
 105  *
 106  * <code>MessageFormat</code> uses patterns of the following form:
 107  * <pre>
 108  * message = messageText (argument messageText)*
 109  * argument = noneArg | simpleArg | complexArg
 110  * complexArg = choiceArg | pluralArg | selectArg | selectordinalArg
 111  *
 112  * noneArg = '{' argNameOrNumber '}'
 113  * simpleArg = '{' argNameOrNumber ',' argType [',' argStyle] '}'
 114  * choiceArg = '{' argNameOrNumber ',' "choice" ',' choiceStyle '}'
 115  * pluralArg = '{' argNameOrNumber ',' "plural" ',' pluralStyle '}'
 116  * selectArg = '{' argNameOrNumber ',' "select" ',' selectStyle '}'
 117  * selectordinalArg = '{' argNameOrNumber ',' "selectordinal" ',' pluralStyle '}'
 118  *
 119  * choiceStyle: see {@link ChoiceFormat}
 120  * pluralStyle: see {@link PluralFormat}
 121  * selectStyle: see {@link SelectFormat}
 122  *
 123  * argNameOrNumber = argName | argNumber
 124  * argName = [^[[:Pattern_Syntax:][:Pattern_White_Space:]]]+
 125  * argNumber = '0' | ('1'..'9' ('0'..'9')*)
 126  *
 127  * argType = "number" | "date" | "time" | "spellout" | "ordinal" | "duration"
 128  * argStyle = "short" | "medium" | "long" | "full" | "integer" | "currency" | "percent" | argStyleText
 129  * </pre>
 130  *
 131  * <ul>
 132  *   <li>messageText can contain quoted literal strings including syntax characters.
 133  *       A quoted literal string begins with an ASCII apostrophe and a syntax character
 134  *       (usually a {curly brace}) and continues until the next single apostrophe.
 135  *       A double ASCII apostrohpe inside or outside of a quoted string represents
 136  *       one literal apostrophe.
 137  *   <li>Quotable syntax characters are the {curly braces} in all messageText parts,
 138  *       plus the '#' sign in a messageText immediately inside a pluralStyle,
 139  *       and the '|' symbol in a messageText immediately inside a choiceStyle.
 140  *   <li>See also {@link #UMessagePatternApostropheMode}
 141  *   <li>In argStyleText, every single ASCII apostrophe begins and ends quoted literal text,
 142  *       and unquoted {curly braces} must occur in matched pairs.
 143  * </ul>
 144  *
 145  * <p>Recommendation: Use the real apostrophe (single quote) character
 146  * \htmlonly&#x2019;\endhtmlonly (U+2019) for
 147  * human-readable text, and use the ASCII apostrophe ' (U+0027)
 148  * only in program syntax, like quoting in MessageFormat.
 149  * See the annotations for U+0027 Apostrophe in The Unicode Standard.
 150  *
 151  * <p>The <code>choice</code> argument type is deprecated.
 152  * Use <code>plural</code> arguments for proper plural selection,
 153  * and <code>select</code> arguments for simple selection among a fixed set of choices.
 154  *
 155  * <p>The <code>argType</code> and <code>argStyle</code> values are used to create
 156  * a <code>Format</code> instance for the format element. The following
 157  * table shows how the values map to Format instances. Combinations not
 158  * shown in the table are illegal. Any <code>argStyleText</code> must
 159  * be a valid pattern string for the Format subclass used.
 160  *
 161  * <p><table border=1>
 162  *    <tr>
 163  *       <th>argType
 164  *       <th>argStyle
 165  *       <th>resulting Format object
 166  *    <tr>
 167  *       <td colspan=2><i>(none)</i>
 168  *       <td><code>null</code>
 169  *    <tr>
 170  *       <td rowspan=5><code>number</code>
 171  *       <td><i>(none)</i>
 172  *       <td><code>NumberFormat.createInstance(getLocale(), status)</code>
 173  *    <tr>
 174  *       <td><code>integer</code>
 175  *       <td><code>NumberFormat.createInstance(getLocale(), kNumberStyle, status)</code>
 176  *    <tr>
 177  *       <td><code>currency</code>
 178  *       <td><code>NumberFormat.createCurrencyInstance(getLocale(), status)</code>
 179  *    <tr>
 180  *       <td><code>percent</code>
 181  *       <td><code>NumberFormat.createPercentInstance(getLocale(), status)</code>
 182  *    <tr>
 183  *       <td><i>argStyleText</i>
 184  *       <td><code>new DecimalFormat(argStyleText, new DecimalFormatSymbols(getLocale(), status), status)</code>
 185  *    <tr>
 186  *       <td rowspan=6><code>date</code>
 187  *       <td><i>(none)</i>
 188  *       <td><code>DateFormat.createDateInstance(kDefault, getLocale(), status)</code>
 189  *    <tr>
 190  *       <td><code>short</code>
 191  *       <td><code>DateFormat.createDateInstance(kShort, getLocale(), status)</code>
 192  *    <tr>
 193  *       <td><code>medium</code>
 194  *       <td><code>DateFormat.createDateInstance(kDefault, getLocale(), status)</code>
 195  *    <tr>
 196  *       <td><code>long</code>
 197  *       <td><code>DateFormat.createDateInstance(kLong, getLocale(), status)</code>
 198  *    <tr>
 199  *       <td><code>full</code>
 200  *       <td><code>DateFormat.createDateInstance(kFull, getLocale(), status)</code>
 201  *    <tr>
 202  *       <td><i>argStyleText</i>
 203  *       <td><code>new SimpleDateFormat(argStyleText, getLocale(), status)
 204  *    <tr>
 205  *       <td rowspan=6><code>time</code>
 206  *       <td><i>(none)</i>
 207  *       <td><code>DateFormat.createTimeInstance(kDefault, getLocale(), status)</code>
 208  *    <tr>
 209  *       <td><code>short</code>
 210  *       <td><code>DateFormat.createTimeInstance(kShort, getLocale(), status)</code>
 211  *    <tr>
 212  *       <td><code>medium</code>
 213  *       <td><code>DateFormat.createTimeInstance(kDefault, getLocale(), status)</code>
 214  *    <tr>
 215  *       <td><code>long</code>
 216  *       <td><code>DateFormat.createTimeInstance(kLong, getLocale(), status)</code>
 217  *    <tr>
 218  *       <td><code>full</code>
 219  *       <td><code>DateFormat.createTimeInstance(kFull, getLocale(), status)</code>
 220  *    <tr>
 221  *       <td><i>argStyleText</i>
 222  *       <td><code>new SimpleDateFormat(argStyleText, getLocale(), status)
 223  *    <tr>
 224  *       <td><code>spellout</code>
 225  *       <td><i>argStyleText (optional)</i>
 226  *       <td><code>new RuleBasedNumberFormat(URBNF_SPELLOUT, getLocale(), status)
 227  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 228  *    <tr>
 229  *       <td><code>ordinal</code>
 230  *       <td><i>argStyleText (optional)</i>
 231  *       <td><code>new RuleBasedNumberFormat(URBNF_ORDINAL, getLocale(), status)
 232  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 233  *    <tr>
 234  *       <td><code>duration</code>
 235  *       <td><i>argStyleText (optional)</i>
 236  *       <td><code>new RuleBasedNumberFormat(URBNF_DURATION, getLocale(), status)
 237  *           <br/>&nbsp;&nbsp;&nbsp;&nbsp;.setDefaultRuleset(argStyleText, status);</code>
 238  * </table>
 239  * <p>
 240  *
 241  * <h4>Usage Information</h4>
 242  *
 243  * <p>Here are some examples of usage:
 244  * Example 1:
 245  *
 246  * <pre>
 247  * \code
 248  *     UErrorCode success = U_ZERO_ERROR;
 249  *     GregorianCalendar cal(success);
 250  *     Formattable arguments[] = {
 251  *         7L,
 252  *         Formattable( (Date) cal.getTime(success), Formattable::kIsDate),
 253  *         "a disturbance in the Force"
 254  *     };
 255  *
 256  *     UnicodeString result;
 257  *     MessageFormat::format(
 258  *          "At {1,time} on {1,date}, there was {2} on planet {0,number}.",
 259  *          arguments, 3, result, success );
 260  *
 261  *     cout << "result: " << result << endl;
 262  *     //<output>: At 4:34:20 PM on 23-Mar-98, there was a disturbance
 263  *     //             in the Force on planet 7.
 264  * \endcode
 265  * </pre>
 266  *
 267  * Typically, the message format will come from resources, and the
 268  * arguments will be dynamically set at runtime.
 269  *
 270  * <p>Example 2:
 271  *
 272  * <pre>
 273  *  \code
 274  *     success = U_ZERO_ERROR;
 275  *     Formattable testArgs[] = {3L, "MyDisk"};
 276  *
 277  *     MessageFormat form(
 278  *         "The disk \"{1}\" contains {0} file(s).", success );
 279  *
 280  *     UnicodeString string;
 281  *     FieldPosition fpos = 0;
 282  *     cout << "format: " << form.format(testArgs, 2, string, fpos, success ) << endl;
 283  *
 284  *     // output, with different testArgs:
 285  *     // output: The disk "MyDisk" contains 0 file(s).
 286  *     // output: The disk "MyDisk" contains 1 file(s).
 287  *     // output: The disk "MyDisk" contains 1,273 file(s).
 288  *  \endcode
 289  *  </pre>
 290  *
 291  *
 292  * <p>For messages that include plural forms, you can use a plural argument:
 293  * <pre>
 294  * \code
 295  *  success = U_ZERO_ERROR;
 296  *  MessageFormat msgFmt(
 297  *       "{num_files, plural, "
 298  *       "=0{There are no files on disk \"{disk_name}\".}"
 299  *       "=1{There is one file on disk \"{disk_name}\".}"
 300  *       "other{There are # files on disk \"{disk_name}\".}}",
 301  *      Locale("en"),
 302  *      success);
 303  *  FieldPosition fpos = 0;
 304  *  Formattable testArgs[] = {0L, "MyDisk"};
 305  *  UnicodeString testArgsNames[] = {"num_files", "disk_name"};
 306  *  UnicodeString result;
 307  *  cout << msgFmt.format(testArgs, testArgsNames, 2, result, fpos, 0, success);
 308  *  testArgs[0] = 3L;
 309  *  cout << msgFmt.format(testArgs, testArgsNames, 2, result, fpos, 0, success);
 310  * \endcode
 311  * <em>output</em>:
 312  * There are no files on disk "MyDisk".
 313  * There are 3 files on "MyDisk".
 314  * </pre>
 315  * See {@link PluralFormat} and {@link PluralRules} for details.
 316  *
 317  * <h4><a name="synchronization">Synchronization</a></h4>
 318  *
 319  * <p>MessageFormats are not synchronized.
 320  * It is recommended to create separate format instances for each thread.
 321  * If multiple threads access a format concurrently, it must be synchronized
 322  * externally.
 323  *
 324  * @stable ICU 2.0
 325  */
 326 class U_I18N_API MessageFormat : public Format {
 327 public:
 328 #ifndef U_HIDE_OBSOLETE_API
 329     /**
 330      * Enum type for kMaxFormat.
 331      * @obsolete ICU 3.0.  The 10-argument limit was removed as of ICU 2.6,
 332      * rendering this enum type obsolete.
 333      */
 334     enum EFormatNumber {
 335         /**
 336          * The maximum number of arguments.
 337          * @obsolete ICU 3.0.  The 10-argument limit was removed as of ICU 2.6,
 338          * rendering this constant obsolete.
 339          */
 340         kMaxFormat = 10
 341     };
 342 #endif  /* U_HIDE_OBSOLETE_API */
 343
 344     /**
 345      * Constructs a new MessageFormat using the given pattern and the
 346      * default locale.
 347      *
 348      * @param pattern   Pattern used to construct object.
 349      * @param status    Input/output error code.  If the
 350      *                  pattern cannot be parsed, set to failure code.
 351      * @stable ICU 2.0
 352      */
 353     MessageFormat(const UnicodeString& pattern,
 354                   UErrorCode &status);
 355
 356     /**
 357      * Constructs a new MessageFormat using the given pattern and locale.
 358      * @param pattern   Pattern used to construct object.
 359      * @param newLocale The locale to use for formatting dates and numbers.
 360      * @param status    Input/output error code.  If the
 361      *                  pattern cannot be parsed, set to failure code.
 362      * @stable ICU 2.0
 363      */
 364     MessageFormat(const UnicodeString& pattern,
 365                   const Locale& newLocale,
 366                         UErrorCode& status);
 367     /**
 368      * Constructs a new MessageFormat using the given pattern and locale.
 369      * @param pattern   Pattern used to construct object.
 370      * @param newLocale The locale to use for formatting dates and numbers.
 371      * @param parseError Struct to receive information on the position
 372      *                   of an error within the pattern.
 373      * @param status    Input/output error code.  If the
 374      *                  pattern cannot be parsed, set to failure code.
 375      * @stable ICU 2.0
 376      */
 377     MessageFormat(const UnicodeString& pattern,
 378                   const Locale& newLocale,
 379                   UParseError& parseError,
 380                   UErrorCode& status);
 381     /**
 382      * Constructs a new MessageFormat from an existing one.
 383      * @stable ICU 2.0
 384      */
 385     MessageFormat(const MessageFormat&);
 386
 387     /**
 388      * Assignment operator.
 389      * @stable ICU 2.0
 390      */
 391     const MessageFormat& operator=(const MessageFormat&);
 392
 393     /**
 394      * Destructor.
 395      * @stable ICU 2.0
 396      */
 397     virtual ~MessageFormat();
 398
 399     /**
 400      * Clones this Format object polymorphically.  The caller owns the
 401      * result and should delete it when done.
 402      * @stable ICU 2.0
 403      */
 404     virtual Format* clone(void) const;
 405
 406     /**
 407      * Returns true if the given Format objects are semantically equal.
 408      * Objects of different subclasses are considered unequal.
 409      * @param other  the object to be compared with.
 410      * @return       true if the given Format objects are semantically equal.
 411      * @stable ICU 2.0
 412      */
 413     virtual UBool operator==(const Format& other) const;
 414
 415     /**
 416      * Sets the locale to be used for creating argument Format objects.
 417      * @param theLocale    the new locale value to be set.
 418      * @stable ICU 2.0
 419      */
 420     virtual void setLocale(const Locale& theLocale);
 421
 422     /**
 423      * Gets the locale used for creating argument Format objects.
 424      * format information.
 425      * @return    the locale of the object.
 426      * @stable ICU 2.0
 427      */
 428     virtual const Locale& getLocale(void) const;
 429
 430     /**
 431      * Applies the given pattern string to this message format.
 432      *
 433      * @param pattern   The pattern to be applied.
 434      * @param status    Input/output error code.  If the
 435      *                  pattern cannot be parsed, set to failure code.
 436      * @stable ICU 2.0
 437      */
 438     virtual void applyPattern(const UnicodeString& pattern,
 439                               UErrorCode& status);
 440     /**
 441      * Applies the given pattern string to this message format.
 442      *
 443      * @param pattern    The pattern to be applied.
 444      * @param parseError Struct to receive information on the position
 445      *                   of an error within the pattern.
 446      * @param status    Input/output error code.  If the
 447      *                  pattern cannot be parsed, set to failure code.
 448      * @stable ICU 2.0
 449      */
 450     virtual void applyPattern(const UnicodeString& pattern,
 451                              UParseError& parseError,
 452                              UErrorCode& status);
 453
 454     /**
 455      * Sets the UMessagePatternApostropheMode and the pattern used by this message format.
 456      * Parses the pattern and caches Format objects for simple argument types.
 457      * Patterns and their interpretation are specified in the
 458      * <a href="#patterns">class description</a>.
 459      * <p>
 460      * This method is best used only once on a given object to avoid confusion about the mode,
 461      * and after constructing the object with an empty pattern string to minimize overhead.
 462      *
 463      * @param pattern    The pattern to be applied.
 464      * @param aposMode   The new apostrophe mode.
 465      * @param parseError Struct to receive information on the position
 466      *                   of an error within the pattern.
 467      *                   Can be NULL.
 468      * @param status    Input/output error code.  If the
 469      *                  pattern cannot be parsed, set to failure code.
 470      * @stable ICU 4.8
 471      */
 472     virtual void applyPattern(const UnicodeString& pattern,
 473                               UMessagePatternApostropheMode aposMode,
 474                               UParseError* parseError,
 475                               UErrorCode& status);
 476
 477     /**
 478      * @return this instance's UMessagePatternApostropheMode.
 479      * @stable ICU 4.8
 480      */
 481     UMessagePatternApostropheMode getApostropheMode() const {
 482         return msgPattern.getApostropheMode();
 483     }
 484
 485     /**
 486      * Returns a pattern that can be used to recreate this object.
 487      *
 488      * @param appendTo  Output parameter to receive the pattern.
 489      *                  Result is appended to existing contents.
 490      * @return          Reference to 'appendTo' parameter.
 491      * @stable ICU 2.0
 492      */
 493     virtual UnicodeString& toPattern(UnicodeString& appendTo) const;
 494
 495     /**
 496      * Sets subformats.
 497      * See the class description about format numbering.
 498      * The caller should not delete the Format objects after this call.
 499      * <EM>The array formatsToAdopt is not itself adopted.</EM> Its
 500      * ownership is retained by the caller. If the call fails because
 501      * memory cannot be allocated, then the formats will be deleted
 502      * by this method, and this object will remain unchanged.
 503      *
 504      * <p>If this format uses named arguments, the new formats are discarded
 505      * and this format remains unchanged.
 506      *
 507      * @stable ICU 2.0
 508      * @param formatsToAdopt    the format to be adopted.
 509      * @param count             the size of the array.
 510      */
 511     virtual void adoptFormats(Format** formatsToAdopt, int32_t count);
 512
 513     /**
 514      * Sets subformats.
 515      * See the class description about format numbering.
 516      * Each item in the array is cloned into the internal array.
 517      * If the call fails because memory cannot be allocated, then this
 518      * object will remain unchanged.
 519      *
 520      * <p>If this format uses named arguments, the new formats are discarded
 521      * and this format remains unchanged.
 522      *
 523      * @stable ICU 2.0
 524      * @param newFormats the new format to be set.
 525      * @param cnt        the size of the array.
 526      */
 527     virtual void setFormats(const Format** newFormats, int32_t cnt);
 528
 529
 530     /**
 531      * Sets one subformat.
 532      * See the class description about format numbering.
 533      * The caller should not delete the Format object after this call.
 534      * If the number is over the number of formats already set,
 535      * the item will be deleted and ignored.
 536      *
 537      * <p>If this format uses named arguments, the new format is discarded
 538      * and this format remains unchanged.
 539      *
 540      * @stable ICU 2.0
 541      * @param formatNumber     index of the subformat.
 542      * @param formatToAdopt    the format to be adopted.
 543      */
 544     virtual void adoptFormat(int32_t formatNumber, Format* formatToAdopt);
 545
 546     /**
 547      * Sets one subformat.
 548      * See the class description about format numbering.
 549      * If the number is over the number of formats already set,
 550      * the item will be ignored.
 551      * @param formatNumber     index of the subformat.
 552      * @param format    the format to be set.
 553      * @stable ICU 2.0
 554      */
 555     virtual void setFormat(int32_t formatNumber, const Format& format);
 556
 557     /**
 558      * Gets format names. This function returns formatNames in StringEnumerations
 559      * which can be used with getFormat() and setFormat() to export formattable
 560      * array from current MessageFormat to another.  It is the caller's responsibility
 561      * to delete the returned formatNames.
 562      * @param status  output param set to success/failure code.
 563      * @stable ICU 4.0
 564      */
 565     virtual StringEnumeration* getFormatNames(UErrorCode& status);
 566
 567     /**
 568      * Gets subformat pointer for given format name.
 569      * This function supports both named and numbered
 570      * arguments. If numbered, the formatName is the
 571      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 572      * The returned Format object should not be deleted by the caller,
 573      * nor should the ponter of other object .  The pointer and its
 574      * contents remain valid only until the next call to any method
 575      * of this class is made with this object.
 576      * @param formatName the name or number specifying a format
 577      * @param status  output param set to success/failure code.
 578      * @stable ICU 4.0
 579      */
 580     virtual Format* getFormat(const UnicodeString& formatName, UErrorCode& status);
 581
 582     /**
 583      * Sets one subformat for given format name.
 584      * See the class description about format name.
 585      * This function supports both named and numbered
 586      * arguments-- if numbered, the formatName is the
 587      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 588      * If there is no matched formatName or wrong type,
 589      * the item will be ignored.
 590      * @param formatName  Name of the subformat.
 591      * @param format      the format to be set.
 592      * @param status  output param set to success/failure code.
 593      * @stable ICU 4.0
 594      */
 595     virtual void setFormat(const UnicodeString& formatName, const Format& format, UErrorCode& status);
 596
 597     /**
 598      * Sets one subformat for given format name.
 599      * See the class description about format name.
 600      * This function supports both named and numbered
 601      * arguments-- if numbered, the formatName is the
 602      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 603      * If there is no matched formatName or wrong type,
 604      * the item will be ignored.
 605      * The caller should not delete the Format object after this call.
 606      * @param formatName  Name of the subformat.
 607      * @param formatToAdopt  Format to be adopted.
 608      * @param status      output param set to success/failure code.
 609      * @stable ICU 4.0
 610      */
 611     virtual void adoptFormat(const UnicodeString& formatName, Format* formatToAdopt, UErrorCode& status);
 612
 613     /**
 614      * Gets an array of subformats of this object.  The returned array
 615      * should not be deleted by the caller, nor should the pointers
 616      * within the array.  The array and its contents remain valid only
 617      * until the next call to this format. See the class description
 618      * about format numbering.
 619      *
 620      * @param count output parameter to receive the size of the array
 621      * @return an array of count Format* objects, or NULL if out of
 622      * memory.  Any or all of the array elements may be NULL.
 623      * @stable ICU 2.0
 624      */
 625     virtual const Format** getFormats(int32_t& count) const;
 626
 627
 628     using Format::format;
 629
 630     /**
 631      * Formats the given array of arguments into a user-readable string.
 632      * Does not take ownership of the Formattable* array or its contents.
 633      *
 634      * <p>If this format uses named arguments, appendTo is unchanged and
 635      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 636      *
 637      * @param source    An array of objects to be formatted.
 638      * @param count     The number of elements of 'source'.
 639      * @param appendTo  Output parameter to receive result.
 640      *                  Result is appended to existing contents.
 641      * @param ignore    Not used; inherited from base class API.
 642      * @param status    Input/output error code.  If the
 643      *                  pattern cannot be parsed, set to failure code.
 644      * @return          Reference to 'appendTo' parameter.
 645      * @stable ICU 2.0
 646      */
 647     UnicodeString& format(const Formattable* source,
 648                           int32_t count,
 649                           UnicodeString& appendTo,
 650                           FieldPosition& ignore,
 651                           UErrorCode& status) const;
 652
 653     /**
 654      * Formats the given array of arguments into a user-readable string
 655      * using the given pattern.
 656      *
 657      * <p>If this format uses named arguments, appendTo is unchanged and
 658      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 659      *
 660      * @param pattern   The pattern.
 661      * @param arguments An array of objects to be formatted.
 662      * @param count     The number of elements of 'source'.
 663      * @param appendTo  Output parameter to receive result.
 664      *                  Result is appended to existing contents.
 665      * @param status    Input/output error code.  If the
 666      *                  pattern cannot be parsed, set to failure code.
 667      * @return          Reference to 'appendTo' parameter.
 668      * @stable ICU 2.0
 669      */
 670     static UnicodeString& format(const UnicodeString& pattern,
 671                                  const Formattable* arguments,
 672                                  int32_t count,
 673                                  UnicodeString& appendTo,
 674                                  UErrorCode& status);
 675
 676     /**
 677      * Formats the given array of arguments into a user-readable
 678      * string.  The array must be stored within a single Formattable
 679      * object of type kArray. If the Formattable object type is not of
 680      * type kArray, then returns a failing UErrorCode.
 681      *
 682      * <p>If this format uses named arguments, appendTo is unchanged and
 683      * status is set to U_ILLEGAL_ARGUMENT_ERROR.
 684      *
 685      * @param obj       A Formattable of type kArray containing
 686      *                  arguments to be formatted.
 687      * @param appendTo  Output parameter to receive result.
 688      *                  Result is appended to existing contents.
 689      * @param pos       On input: an alignment field, if desired.
 690      *                  On output: the offsets of the alignment field.
 691      * @param status    Input/output error code.  If the
 692      *                  pattern cannot be parsed, set to failure code.
 693      * @return          Reference to 'appendTo' parameter.
 694      * @stable ICU 2.0
 695      */
 696     virtual UnicodeString& format(const Formattable& obj,
 697                                   UnicodeString& appendTo,
 698                                   FieldPosition& pos,
 699                                   UErrorCode& status) const;
 700
 701     /**
 702      * Formats the given array of arguments into a user-defined argument name
 703      * array. This function supports both named and numbered
 704      * arguments-- if numbered, the formatName is the
 705      * corresponding UnicodeStrings (e.g. "0", "1", "2"...).
 706      *
 707      * @param argumentNames argument name array
 708      * @param arguments An array of objects to be formatted.
 709      * @param count     The number of elements of 'argumentNames' and
 710      *                  arguments.  The number of argumentNames and arguments
 711      *                  must be the same.
 712      * @param appendTo  Output parameter to receive result.
 713      *                  Result is appended to existing contents.
 714      * @param status    Input/output error code.  If the
 715      *                  pattern cannot be parsed, set to failure code.
 716      * @return          Reference to 'appendTo' parameter.
 717      * @stable ICU 4.0
 718      */
 719     UnicodeString& format(const UnicodeString* argumentNames,
 720                           const Formattable* arguments,
 721                           int32_t count,
 722                           UnicodeString& appendTo,
 723                           UErrorCode& status) const;
 724     /**
 725      * Parses the given string into an array of output arguments.
 726      *
 727      * @param source    String to be parsed.
 728      * @param pos       On input, starting position for parse. On output,
 729      *                  final position after parse.  Unchanged if parse
 730      *                  fails.
 731      * @param count     Output parameter to receive the number of arguments
 732      *                  parsed.
 733      * @return an array of parsed arguments.  The caller owns both
 734      * the array and its contents.
 735      * @stable ICU 2.0
 736      */
 737     virtual Formattable* parse(const UnicodeString& source,
 738                                ParsePosition& pos,
 739                                int32_t& count) const;
 740
 741     /**
 742      * Parses the given string into an array of output arguments.
 743      *
 744      * <p>If this format uses named arguments, status is set to
 745      * U_ARGUMENT_TYPE_MISMATCH.
 746      *
 747      * @param source    String to be parsed.
 748      * @param count     Output param to receive size of returned array.
 749      * @param status    Input/output error code.  If the
 750      *                  pattern cannot be parsed, set to failure code.
 751      * @return an array of parsed arguments.  The caller owns both
 752      * the array and its contents. Returns NULL if status is not U_ZERO_ERROR.
 753      *
 754      * @stable ICU 2.0
 755      */
 756     virtual Formattable* parse(const UnicodeString& source,
 757                                int32_t& count,
 758                                UErrorCode& status) const;
 759
 760     /**
 761      * Parses the given string into an array of output arguments
 762      * stored within a single Formattable of type kArray.
 763      *
 764      * @param source    The string to be parsed into an object.
 765      * @param result    Formattable to be set to the parse result.
 766      *                  If parse fails, return contents are undefined.
 767      * @param pos       On input, starting position for parse. On output,
 768      *                  final position after parse.  Unchanged if parse
 769      *                  fails.
 770      * @stable ICU 2.0
 771      */
 772     virtual void parseObject(const UnicodeString& source,
 773                              Formattable& result,
 774                              ParsePosition& pos) const;
 775
 776     /**
 777      * Convert an 'apostrophe-friendly' pattern into a standard
 778      * pattern.  Standard patterns treat all apostrophes as
 779      * quotes, which is problematic in some languages, e.g.
 780      * French, where apostrophe is commonly used.  This utility
 781      * assumes that only an unpaired apostrophe immediately before
 782      * a brace is a true quote.  Other unpaired apostrophes are paired,
 783      * and the resulting standard pattern string is returned.
 784      *
 785      * <p><b>Note</b> it is not guaranteed that the returned pattern
 786      * is indeed a valid pattern.  The only effect is to convert
 787      * between patterns having different quoting semantics.
 788      *
 789      * @param pattern the 'apostrophe-friendly' patttern to convert
 790      * @param status    Input/output error code.  If the pattern
 791      *                  cannot be parsed, the failure code is set.
 792      * @return the standard equivalent of the original pattern
 793      * @stable ICU 3.4
 794      */
 795     static UnicodeString autoQuoteApostrophe(const UnicodeString& pattern,
 796         UErrorCode& status);
 797
 798
 799     /**
 800      * Returns true if this MessageFormat uses named arguments,
 801      * and false otherwise.  See class description.
 802      *
 803      * @return true if named arguments are used.
 804      * @stable ICU 4.0
 805      */
 806     UBool usesNamedArguments() const;
 807
 808
 809 #ifndef U_HIDE_INTERNAL_API
 810     /**
 811      * This API is for ICU internal use only.
 812      * Please do not use it.
 813      *
 814      * Returns argument types count in the parsed pattern.
 815      * Used to distinguish pattern "{0} d" and "d".
 816      *
 817      * @return           The number of formattable types in the pattern
 818      * @internal
 819      */
 820     int32_t getArgTypeCount() const;
 821 #endif  /* U_HIDE_INTERNAL_API */
 822
 823     /**
 824      * Returns a unique class ID POLYMORPHICALLY.  Pure virtual override.
 825      * This method is to implement a simple version of RTTI, since not all
 826      * C++ compilers support genuine RTTI.  Polymorphic operator==() and
 827      * clone() methods call this method.
 828      *
 829      * @return          The class ID for this object. All objects of a
 830      *                  given class have the same class ID.  Objects of
 831      *                  other classes have different class IDs.
 832      * @stable ICU 2.0
 833      */
 834     virtual UClassID getDynamicClassID(void) const;
 835
 836     /**
 837      * Return the class ID for this class.  This is useful only for
 838      * comparing to a return value from getDynamicClassID().  For example:
 839      * <pre>
 840      * .   Base* polymorphic_pointer = createPolymorphicObject();
 841      * .   if (polymorphic_pointer->getDynamicClassID() ==
 842      * .      Derived::getStaticClassID()) ...
 843      * </pre>
 844      * @return          The class ID for all objects of this class.
 845      * @stable ICU 2.0
 846      */
 847     static UClassID U_EXPORT2 getStaticClassID(void);
 848
 849 #ifndef U_HIDE_INTERNAL_API
 850     /**
 851      * Compares two Format objects. This is used for constructing the hash
 852      * tables.
 853      *
 854      * @param left pointer to a Format object. Must not be NULL.
 855      * @param right pointer to a Format object. Must not be NULL.
 856      *
 857      * @return whether the two objects are the same
 858      * @internal
 859      */
 860     static UBool equalFormats(const void* left, const void* right);
 861 #endif  /* U_HIDE_INTERNAL_API */
 862
 863 private:
 864
 865     Locale              fLocale;
 866     MessagePattern      msgPattern;
 867     Format**            formatAliases; // see getFormats
 868     int32_t             formatAliasesCapacity;
 869
 870     MessageFormat(); // default constructor not implemented
 871
 872      /**
 873       * This provider helps defer instantiation of a PluralRules object
 874       * until we actually need to select a keyword.
 875       * For example, if the number matches an explicit-value selector like "=1"
 876       * we do not need any PluralRules.
 877       */
 878     class U_I18N_API PluralSelectorProvider : public PluralFormat::PluralSelector {
 879     public:
 880         PluralSelectorProvider(const MessageFormat &mf, UPluralType type);
 881         virtual ~PluralSelectorProvider();
 882         virtual UnicodeString select(void *ctx, double number, UErrorCode& ec) const;
 883
 884         void reset();
 885     private:
 886         const MessageFormat &msgFormat;
 887         PluralRules* rules;
 888         UPluralType type;
 889     };
 890
 891     /**
 892      * A MessageFormat formats an array of arguments.  Each argument
 893      * has an expected type, based on the pattern.  For example, if
 894      * the pattern contains the subformat "{3,number,integer}", then
 895      * we expect argument 3 to have type Formattable::kLong.  This
 896      * array needs to grow dynamically if the MessageFormat is
 897      * modified.
 898      */
 899     Formattable::Type* argTypes;
 900     int32_t            argTypeCount;
 901     int32_t            argTypeCapacity;
 902
 903     /**
 904      * TRUE if there are different argTypes for the same argument.
 905      * This only matters when the MessageFormat is used in the plain C (umsg_xxx) API
 906      * where the pattern argTypes determine how the va_arg list is read.
 907      */
 908     UBool hasArgTypeConflicts;
 909
 910     // Variable-size array management
 911     UBool allocateArgTypes(int32_t capacity, UErrorCode& status);
 912
 913     /**
 914      * Default Format objects used when no format is specified and a
 915      * numeric or date argument is formatted.  These are volatile
 916      * cache objects maintained only for performance.  They do not
 917      * participate in operator=(), copy constructor(), nor
 918      * operator==().
 919      */
 920     NumberFormat* defaultNumberFormat;
 921     DateFormat*   defaultDateFormat;
 922
 923     UHashtable* cachedFormatters;
 924     UHashtable* customFormatArgStarts;
 925
 926     PluralSelectorProvider pluralProvider;
 927     PluralSelectorProvider ordinalProvider;
 928
 929     /**
 930      * Method to retrieve default formats (or NULL on failure).
 931      * These are semantically const, but may modify *this.
 932      */
 933     const NumberFormat* getDefaultNumberFormat(UErrorCode&) const;
 934     const DateFormat*   getDefaultDateFormat(UErrorCode&) const;
 935
 936     /**
 937      * Finds the word s, in the keyword list and returns the located index.
 938      * @param s the keyword to be searched for.
 939      * @param list the list of keywords to be searched with.
 940      * @return the index of the list which matches the keyword s.
 941      */
 942     static int32_t findKeyword( const UnicodeString& s,
 943                                 const char16_t * const *list);
 944
 945     /**
 946      * Thin wrapper around the format(... AppendableWrapper ...) variant.
 947      * Wraps the destination UnicodeString into an AppendableWrapper and
 948      * supplies default values for some other parameters.
 949      */
 950     UnicodeString& format(const Formattable* arguments,
 951                           const UnicodeString *argumentNames,
 952                           int32_t cnt,
 953                           UnicodeString& appendTo,
 954                           FieldPosition* pos,
 955                           UErrorCode& status) const;
 956
 957     /**
 958      * Formats the arguments and writes the result into the
 959      * AppendableWrapper, updates the field position.
 960      *
 961      * @param msgStart      Index to msgPattern part to start formatting from.
 962      * @param plNumber      NULL except when formatting a plural argument sub-message
 963      *                      where a '#' is replaced by the format string for this number.
 964      * @param arguments     The formattable objects array. (Must not be NULL.)
 965      * @param argumentNames NULL if numbered values are used. Otherwise the same
 966      *                      length as "arguments", and each entry is the name of the
 967      *                      corresponding argument in "arguments".
 968      * @param cnt           The length of arguments (and of argumentNames if that is not NULL).
 969      * @param appendTo      Output parameter to receive the result.
 970      *                      The result string is appended to existing contents.
 971      * @param pos           Field position status.
 972      * @param success       The error code status.
 973      */
 974     void format(int32_t msgStart,
 975                 const void *plNumber,
 976                 const Formattable* arguments,
 977                 const UnicodeString *argumentNames,
 978                 int32_t cnt,
 979                 AppendableWrapper& appendTo,
 980                 FieldPosition* pos,
 981                 UErrorCode& success) const;
 982
 983     UnicodeString getArgName(int32_t partIndex);
 984
 985     void setArgStartFormat(int32_t argStart, Format* formatter, UErrorCode& status);
 986
 987     void setCustomArgStartFormat(int32_t argStart, Format* formatter, UErrorCode& status);
 988
 989     int32_t nextTopLevelArgStart(int32_t partIndex) const;
 990
 991     UBool argNameMatches(int32_t partIndex, const UnicodeString& argName, int32_t argNumber);
 992
 993     void cacheExplicitFormats(UErrorCode& status);
 994
 995     Format* createAppropriateFormat(UnicodeString& type,
 996                                     UnicodeString& style,
 997                                     Formattable::Type& formattableType,
 998                                     UParseError& parseError,
 999                                     UErrorCode& ec);
1000
1001     const Formattable* getArgFromListByName(const Formattable* arguments,
1002                                             const UnicodeString *argumentNames,
1003                                             int32_t cnt, UnicodeString& name) const;
1004
1005     Formattable* parse(int32_t msgStart,
1006                        const UnicodeString& source,
1007                        ParsePosition& pos,
1008                        int32_t& count,
1009                        UErrorCode& ec) const;
1010
1011     FieldPosition* updateMetaData(AppendableWrapper& dest, int32_t prevLength,
1012                                   FieldPosition* fp, const Formattable* argId) const;
1013
1014     /**
1015      * Finds the "other" sub-message.
1016      * @param partIndex the index of the first PluralFormat argument style part.
1017      * @return the "other" sub-message start part index.
1018      */
1019     int32_t findOtherSubMessage(int32_t partIndex) const;
1020
1021     /**
1022      * Returns the ARG_START index of the first occurrence of the plural number in a sub-message.
1023      * Returns -1 if it is a REPLACE_NUMBER.
1024      * Returns 0 if there is neither.
1025      */
1026     int32_t findFirstPluralNumberArg(int32_t msgStart, const UnicodeString &argName) const;
1027
1028     Format* getCachedFormatter(int32_t argumentNumber) const;
1029
1030     UnicodeString getLiteralStringUntilNextArgument(int32_t from) const;
1031
1032     void copyObjects(const MessageFormat& that, UErrorCode& ec);
1033
1034     void formatComplexSubMessage(int32_t msgStart,
1035                                  const void *plNumber,
1036                                  const Formattable* arguments,
1037                                  const UnicodeString *argumentNames,
1038                                  int32_t cnt,
1039                                  AppendableWrapper& appendTo,
1040                                  UErrorCode& success) const;
1041
1042     /**
1043      * Convenience method that ought to be in NumberFormat
1044      */
1045     NumberFormat* createIntegerFormat(const Locale& locale, UErrorCode& status) const;
1046
1047     /**
1048      * Returns array of argument types in the parsed pattern
1049      * for use in C API.  Only for the use of umsg_vformat().  Not
1050      * for public consumption.
1051      * @param listCount  Output parameter to receive the size of array
1052      * @return           The array of formattable types in the pattern
1053      */
1054     const Formattable::Type* getArgTypeList(int32_t& listCount) const {
1055         listCount = argTypeCount;
1056         return argTypes;
1057     }
1058
1059     /**
1060      * Resets the internal MessagePattern, and other associated caches.
1061      */
1062     void resetPattern();
1063
1064     /**
1065      * A DummyFormatter that we use solely to store a NULL value. UHash does
1066      * not support storing NULL values.
1067      */
1068     class U_I18N_API DummyFormat : public Format {
1069     public:
1070         virtual UBool operator==(const Format&) const;
1071         virtual Format* clone() const;
1072         virtual UnicodeString& format(const Formattable& obj,
1073                               UnicodeString& appendTo,
1074                               UErrorCode& status) const;
1075         virtual UnicodeString& format(const Formattable&,
1076                                       UnicodeString& appendTo,
1077                                       FieldPosition&,
1078                                       UErrorCode& status) const;
1079         virtual UnicodeString& format(const Formattable& obj,
1080                                       UnicodeString& appendTo,
1081                                       FieldPositionIterator* posIter,
1082                                       UErrorCode& status) const;
1083         virtual void parseObject(const UnicodeString&,
1084                                  Formattable&,
1085                                  ParsePosition&) const;
1086     };
1087
1088     friend class MessageFormatAdapter; // getFormatTypeList() access
1089 };
1090
1091 U_NAMESPACE_END
1092 #endif // U_SHOW_CPLUSPLUS_API
1093
1094 #endif /* #if !UCONFIG_NO_FORMATTING */
1095
1096 #endif // _MSGFMT
1097 //eof