Home | History | Annotate | Download | only in unicode
      1 /*
      2 *******************************************************************************
      3 * Copyright (C) 2008-2011, International Business Machines Corporation and
      4 * others. All Rights Reserved.
      5 *******************************************************************************
      6 *
      7 *
      8 * File PLURRULE.H
      9 *
     10 * Modification History:*
     11 *   Date        Name        Description
     12 *
     13 ********************************************************************************
     14 */
     15 
     16 #ifndef PLURRULE
     17 #define PLURRULE
     18 
     19 #include "unicode/utypes.h"
     20 
     21 /**
     22  * \file
     23  * \brief C++ API: PluralRules object
     24  */
     25 
     26 #if !UCONFIG_NO_FORMATTING
     27 
     28 #include "unicode/format.h"
     29 
     30 /**
     31  * Value returned by PluralRules::getUniqueKeywordValue() when there is no
     32  * unique value to return.
     33  * @draft ICU 4.8
     34  */
     35 #define UPLRULES_NO_UNIQUE_VALUE ((double)-0.00123456777)
     36 
     37 U_NAMESPACE_BEGIN
     38 
     39 class Hashtable;
     40 class RuleChain;
     41 class RuleParser;
     42 class PluralKeywordEnumeration;
     43 
     44 /**
     45  * Defines rules for mapping non-negative numeric values onto a small set of
     46  * keywords. Rules are constructed from a text description, consisting
     47  * of a series of keywords and conditions.  The {@link #select} method
     48  * examines each condition in order and returns the keyword for the
     49  * first condition that matches the number.  If none match,
     50  * default rule(other) is returned.
     51  *
     52  * For more information, details, and tips for writing rules, see the
     53  * LDML spec, C.11 Language Plural Rules:
     54  * http://www.unicode.org/draft/reports/tr35/tr35.html#Language_Plural_Rules
     55  *
     56  * Examples:<pre>
     57  *   "one: n is 1; few: n in 2..4"</pre>
     58  *  This defines two rules, for 'one' and 'few'.  The condition for
     59  *  'one' is "n is 1" which means that the number must be equal to
     60  *  1 for this condition to pass.  The condition for 'few' is
     61  *  "n in 2..4" which means that the number must be between 2 and
     62  *  4 inclusive for this condition to pass.  All other numbers
     63  *  are assigned the keyword "other" by the default rule.
     64  *  </p><pre>
     65  *    "zero: n is 0; one: n is 1; zero: n mod 100 in 1..19"</pre>
     66  *  This illustrates that the same keyword can be defined multiple times.
     67  *  Each rule is examined in order, and the first keyword whose condition
     68  *  passes is the one returned.  Also notes that a modulus is applied
     69  *  to n in the last rule.  Thus its condition holds for 119, 219, 319...
     70  *  </p><pre>
     71  *    "one: n is 1; few: n mod 10 in 2..4 and n mod 100 not in 12..14"</pre>
     72  *  This illustrates conjunction and negation.  The condition for 'few'
     73  *  has two parts, both of which must be met: "n mod 10 in 2..4" and
     74  *  "n mod 100 not in 12..14".  The first part applies a modulus to n
     75  *  before the test as in the previous example.  The second part applies
     76  *  a different modulus and also uses negation, thus it matches all
     77  *  numbers _not_ in 12, 13, 14, 112, 113, 114, 212, 213, 214...
     78  *  </p>
     79  *  <p>
     80  * Syntax:<pre>
     81  * \code
     82  * rules         = rule (';' rule)*
     83  * rule          = keyword ':' condition
     84  * keyword       = <identifier>
     85  * condition     = and_condition ('or' and_condition)*
     86  * and_condition = relation ('and' relation)*
     87  * relation      = is_relation | in_relation | within_relation | 'n' <EOL>
     88  * is_relation   = expr 'is' ('not')? value
     89  * in_relation   = expr ('not')? 'in' range_list
     90  * within_relation = expr ('not')? 'within' range
     91  * expr          = 'n' ('mod' value)?
     92  * range_list    = (range | value) (',' range_list)*
     93  * value         = digit+
     94  * digit         = 0|1|2|3|4|5|6|7|8|9
     95  * range         = value'..'value
     96  * \endcode
     97  * </pre></p>
     98  * <p>
     99  * An "identifier" is a sequence of characters that do not have the
    100  * Unicode Pattern_Syntax or Pattern_White_Space properties.
    101  * <p>
    102  * The difference between 'in' and 'within' is that 'in' only includes
    103  * integers in the specified range, while 'within' includes all values.</p>
    104  * <p>
    105  * Keywords
    106  * could be defined by users or from ICU locale data. There are 6
    107  * predefined values in ICU - 'zero', 'one', 'two', 'few', 'many' and
    108  * 'other'. Callers need to check the value of keyword returned by
    109  * {@link #select} method.
    110  * </p>
    111  *
    112  * Examples:<pre>
    113  * UnicodeString keyword = pl->select(number);
    114  * if (keyword== UnicodeString("one") {
    115  *     ...
    116  * }
    117  * else if ( ... )
    118  * </pre>
    119  * <strong>Note:</strong><br>
    120  *  <p>
    121  *   ICU defines plural rules for many locales based on CLDR <i>Language Plural Rules</i>.
    122  *   For these predefined rules, see CLDR page at
    123  *    http://unicode.org/repos/cldr-tmp/trunk/diff/supplemental/language_plural_rules.html
    124  * </p>
    125  */
    126 class U_I18N_API PluralRules : public UObject {
    127 public:
    128 
    129     /**
    130      * Constructor.
    131      * @param status  Output param set to success/failure code on exit, which
    132      *                must not indicate a failure before the function call.
    133      *
    134      * @stable ICU 4.0
    135      */
    136     PluralRules(UErrorCode& status);
    137 
    138     /**
    139      * Copy constructor.
    140      * @stable ICU 4.0
    141      */
    142     PluralRules(const PluralRules& other);
    143 
    144     /**
    145      * Destructor.
    146      * @stable ICU 4.0
    147      */
    148     virtual ~PluralRules();
    149 
    150     /**
    151      * Clone
    152      * @stable ICU 4.0
    153      */
    154     PluralRules* clone() const;
    155 
    156     /**
    157       * Assignment operator.
    158       * @stable ICU 4.0
    159       */
    160     PluralRules& operator=(const PluralRules&);
    161 
    162     /**
    163      * Creates a PluralRules from a description if it is parsable, otherwise
    164      * returns NULL.
    165      *
    166      * @param description rule description
    167      * @param status      Output param set to success/failure code on exit, which
    168      *                    must not indicate a failure before the function call.
    169      * @return            new PluralRules pointer. NULL if there is an error.
    170      * @stable ICU 4.0
    171      */
    172     static PluralRules* U_EXPORT2 createRules(const UnicodeString& description,
    173                                               UErrorCode& status);
    174 
    175     /**
    176      * The default rules that accept any number.
    177      *
    178      * @param status  Output param set to success/failure code on exit, which
    179      *                must not indicate a failure before the function call.
    180      * @return        new PluralRules pointer. NULL if there is an error.
    181      * @stable ICU 4.0
    182      */
    183     static PluralRules* U_EXPORT2 createDefaultRules(UErrorCode& status);
    184 
    185     /**
    186      * Provides access to the predefined <code>PluralRules</code> for a given
    187      * locale.
    188      *
    189      * @param locale  The locale for which a <code>PluralRules</code> object is
    190      *                returned.
    191      * @param status  Output param set to success/failure code on exit, which
    192      *                must not indicate a failure before the function call.
    193      * @return        The predefined <code>PluralRules</code> object pointer for
    194      *                this locale. If there's no predefined rules for this locale,
    195      *                the rules for the closest parent in the locale hierarchy
    196      *                that has one will  be returned.  The final fallback always
    197      *                returns the default 'other' rules.
    198      * @stable ICU 4.0
    199      */
    200     static PluralRules* U_EXPORT2 forLocale(const Locale& locale, UErrorCode& status);
    201 
    202     /**
    203      * Given a number, returns the keyword of the first rule that applies to
    204      * the number.  This function can be used with isKeyword* functions to
    205      * determine the keyword for default plural rules.
    206      *
    207      * @param number  The number for which the rule has to be determined.
    208      * @return        The keyword of the selected rule.
    209      * @stable ICU 4.0
    210      */
    211     UnicodeString select(int32_t number) const;
    212 
    213     /**
    214      * Given a number, returns the keyword of the first rule that applies to
    215      * the number.  This function can be used with isKeyword* functions to
    216      * determine the keyword for default plural rules.
    217      *
    218      * @param number  The number for which the rule has to be determined.
    219      * @return        The keyword of the selected rule.
    220      * @stable ICU 4.0
    221      */
    222     UnicodeString select(double number) const;
    223 
    224     /**
    225      * Returns a list of all rule keywords used in this <code>PluralRules</code>
    226      * object.  The rule 'other' is always present by default.
    227      *
    228      * @param status Output param set to success/failure code on exit, which
    229      *               must not indicate a failure before the function call.
    230      * @return       StringEnumeration with the keywords.
    231      *               The caller must delete the object.
    232      * @stable ICU 4.0
    233      */
    234     StringEnumeration* getKeywords(UErrorCode& status) const;
    235 
    236     /**
    237      * Returns a unique value for this keyword if it exists, else the constant
    238      * UPLRULES_NO_UNIQUE_VALUE.
    239      *
    240      * @param keyword The keyword.
    241      * @return        The unique value that generates the keyword, or
    242      *                UPLRULES_NO_UNIQUE_VALUE if the keyword is undefined or there is no
    243      *                unique value that generates this keyword.
    244      * @draft ICU 4.8
    245      */
    246     double getUniqueKeywordValue(const UnicodeString& keyword);
    247 
    248     /**
    249      * Returns all the values for which select() would return the keyword.  If
    250      * the keyword is unknown, returns no values, but this is not an error.  If
    251      * the number of values is unlimited, returns no values and -1 as the
    252      * count.
    253      *
    254      * The number of returned values is typically small.
    255      *
    256      * @param keyword      The keyword.
    257      * @param dest         Array into which to put the returned values.  May
    258      *                     be NULL if destCapacity is 0.
    259      * @param destCapacity The capacity of the array, must be at least 0.
    260      * @param status       The error code.
    261      * @return             The count of values available, or -1.  This count
    262      *                     can be larger than destCapacity, but no more than
    263      *                     destCapacity values will be written.
    264      * @draft ICU 4.8
    265      */
    266     int32_t getAllKeywordValues(const UnicodeString &keyword,
    267                                 double *dest, int32_t destCapacity,
    268                                 UErrorCode& status);
    269 
    270     /**
    271      * Returns sample values for which select() would return the keyword.  If
    272      * the keyword is unknown, returns no values, but this is not an error.
    273      *
    274      * The number of returned values is typically small.
    275      *
    276      * @param keyword      The keyword.
    277      * @param dest         Array into which to put the returned values.  May
    278      *                     be NULL if destCapacity is 0.
    279      * @param destCapacity The capacity of the array, must be at least 0.
    280      * @param status       The error code.
    281      * @return             The count of values written.
    282      *                     If more than destCapacity samples are available, then
    283      *                     only destCapacity are written, and destCapacity is returned as the count,
    284      *                     rather than setting a U_BUFFER_OVERFLOW_ERROR.
    285      *                     (The actual number of keyword values could be unlimited.)
    286      * @draft ICU 4.8
    287      */
    288     int32_t getSamples(const UnicodeString &keyword,
    289                        double *dest, int32_t destCapacity,
    290                        UErrorCode& status);
    291 
    292     /**
    293      * Returns TRUE if the given keyword is defined in this
    294      * <code>PluralRules</code> object.
    295      *
    296      * @param keyword  the input keyword.
    297      * @return         TRUE if the input keyword is defined.
    298      *                 Otherwise, return FALSE.
    299      * @stable ICU 4.0
    300      */
    301     UBool isKeyword(const UnicodeString& keyword) const;
    302 
    303 
    304     /**
    305      * Returns keyword for default plural form.
    306      *
    307      * @return         keyword for default plural form.
    308      * @internal 4.0
    309      * @stable ICU 4.0
    310      */
    311     UnicodeString getKeywordOther() const;
    312 
    313     /**
    314      * Compares the equality of two PluralRules objects.
    315      *
    316      * @param other The other PluralRules object to be compared with.
    317      * @return      True if the given PluralRules is the same as this
    318      *              PluralRules; false otherwise.
    319      * @stable ICU 4.0
    320      */
    321     virtual UBool operator==(const PluralRules& other) const;
    322 
    323     /**
    324      * Compares the inequality of two PluralRules objects.
    325      *
    326      * @param other The PluralRules object to be compared with.
    327      * @return      True if the given PluralRules is not the same as this
    328      *              PluralRules; false otherwise.
    329      * @stable ICU 4.0
    330      */
    331     UBool operator!=(const PluralRules& other) const  {return !operator==(other);}
    332 
    333 
    334     /**
    335      * ICU "poor man's RTTI", returns a UClassID for this class.
    336      *
    337      * @stable ICU 4.0
    338      *
    339     */
    340     static UClassID U_EXPORT2 getStaticClassID(void);
    341 
    342     /**
    343      * ICU "poor man's RTTI", returns a UClassID for the actual class.
    344      *
    345      * @stable ICU 4.0
    346      */
    347     virtual UClassID getDynamicClassID() const;
    348 
    349 
    350 private:
    351     RuleChain  *mRules;
    352     RuleParser *mParser;
    353     double     *mSamples;
    354     int32_t    *mSampleInfo;
    355     int32_t    mSampleInfoCount;
    356 
    357     PluralRules();   // default constructor not implemented
    358     int32_t getRepeatLimit() const;
    359     void parseDescription(UnicodeString& ruleData, RuleChain& rules, UErrorCode &status);
    360     void getNextLocale(const UnicodeString& localeData, int32_t* curIndex, UnicodeString& localeName);
    361     void addRules(RuleChain& rules);
    362     int32_t getNumberValue(const UnicodeString& token) const;
    363     UnicodeString getRuleFromResource(const Locale& locale, UErrorCode& status);
    364 
    365     static const int32_t MAX_SAMPLES = 3;
    366 
    367     int32_t getSamplesInternal(const UnicodeString &keyword, double *dest,
    368                                int32_t destCapacity, UBool includeUnlimited,
    369                                UErrorCode& status);
    370     int32_t getKeywordIndex(const UnicodeString& keyword,
    371                             UErrorCode& status) const;
    372     void initSamples(UErrorCode& status);
    373 
    374 };
    375 
    376 U_NAMESPACE_END
    377 
    378 #endif /* #if !UCONFIG_NO_FORMATTING */
    379 
    380 #endif // _PLURRULE
    381 //eof
    382