* Important: The ICU collation service has been reimplemented * in order to achieve better performance and UCA compliance. * For details, see the * * collation design document. *
* RuleBasedCollator is a thin C++ wrapper over the C implementation. *
* For more information about the collation service see * the users guide. *
* Collation service provides correct sorting orders for most locales supported in ICU. * If specific data for a locale is not available, the orders eventually falls back * to the UCA sort order. *
* Sort ordering may be customized by providing your own set of rules. For more on * this subject see the * Collation customization section of the users guide. *
* Note, RuleBasedCollator is not to be subclassed. * @see Collator * @version 2.0 11/15/2001 */ class U_I18N_API RuleBasedCollator : public Collator { public: // constructor ------------------------------------------------------------- /** * RuleBasedCollator constructor. This takes the table rules and builds a * collation table out of them. Please see RuleBasedCollator class * description for more details on the collation rule syntax. * @param rules the collation rules to build the collation table from. * @param status reporting a success or an error. * @see Locale * @stable ICU 2.0 */ RuleBasedCollator(const UnicodeString& rules, UErrorCode& status); /** * RuleBasedCollator constructor. This takes the table rules and builds a * collation table out of them. Please see RuleBasedCollator class * description for more details on the collation rule syntax. * @param rules the collation rules to build the collation table from. * @param collationStrength default strength for comparison * @param status reporting a success or an error. * @see Locale * @stable ICU 2.0 */ RuleBasedCollator(const UnicodeString& rules, ECollationStrength collationStrength, UErrorCode& status); /** * RuleBasedCollator constructor. This takes the table rules and builds a * collation table out of them. Please see RuleBasedCollator class * description for more details on the collation rule syntax. * @param rules the collation rules to build the collation table from. * @param decompositionMode the normalisation mode * @param status reporting a success or an error. * @see Locale * @stable ICU 2.0 */ RuleBasedCollator(const UnicodeString& rules, UColAttributeValue decompositionMode, UErrorCode& status); /** * RuleBasedCollator constructor. This takes the table rules and builds a * collation table out of them. Please see RuleBasedCollator class * description for more details on the collation rule syntax. * @param rules the collation rules to build the collation table from. * @param collationStrength default strength for comparison * @param decompositionMode the normalisation mode * @param status reporting a success or an error. * @see Locale * @stable ICU 2.0 */ RuleBasedCollator(const UnicodeString& rules, ECollationStrength collationStrength, UColAttributeValue decompositionMode, UErrorCode& status); /** * Copy constructor. * @param the RuleBasedCollator object to be copied * @see Locale * @stable ICU 2.0 */ RuleBasedCollator(const RuleBasedCollator& other); // destructor -------------------------------------------------------------- /** * Destructor. * @stable ICU 2.0 */ virtual ~RuleBasedCollator(); // public methods ---------------------------------------------------------- /** * Assignment operator. * @param other other RuleBasedCollator object to compare with. * @stable ICU 2.0 */ RuleBasedCollator& operator=(const RuleBasedCollator& other); /** * Returns true if argument is the same as this object. * @param other Collator object to be compared. * @return true if arguments is the same as this object. * @stable ICU 2.0 */ virtual UBool operator==(const Collator& other) const; /** * Returns true if argument is not the same as this object. * @param other Collator object to be compared * @return returns true if argument is not the same as this object. * @stable ICU 2.0 */ virtual UBool operator!=(const Collator& other) const; /** * Makes a deep copy of the object. * The caller owns the returned object. * @return the cloned object. * @stable ICU 2.0 */ virtual Collator* clone(void) const; /** * Creates a collation element iterator for the source string. The caller of * this method is responsible for the memory management of the return * pointer. * @param source the string over which the CollationElementIterator will * iterate. * @return the collation element iterator of the source string using this as * the based Collator. * @stable ICU 2.2 */ virtual CollationElementIterator* createCollationElementIterator( const UnicodeString& source) const; /** * Creates a collation element iterator for the source. The caller of this * method is responsible for the memory management of the returned pointer. * @param source the CharacterIterator which produces the characters over * which the CollationElementItgerator will iterate. * @return the collation element iterator of the source using this as the * based Collator. * @stable ICU 2.2 */ virtual CollationElementIterator* createCollationElementIterator( const CharacterIterator& source) const; /** * Compares a range of character data stored in two different strings based * on the collation rules. Returns information about whether a string is * less than, greater than or equal to another string in a language. * This can be overriden in a subclass. * @param source the source string. * @param target the target string to be compared with the source string. * @return the comparison result. GREATER if the source string is greater * than the target string, LESS if the source is less than the * target. Otherwise, returns EQUAL. * @deprecated ICU 2.6 Use overload with UErrorCode& */ virtual EComparisonResult compare(const UnicodeString& source, const UnicodeString& target) const; /** * The comparison function compares the character data stored in two * different strings. Returns information about whether a string is less * than, greater than or equal to another string. * @param source the source string to be compared with. * @param target the string that is to be compared with the source string. * @param status possible error code * @return Returns an enum value. UCOL_GREATER if source is greater * than target; UCOL_EQUAL if source is equal to target; UCOL_LESS if source is less * than target * @draft ICU 2.6 **/ virtual UCollationResult compare(const UnicodeString& source, const UnicodeString& target, UErrorCode &status) const; /** * Compares a range of character data stored in two different strings based * on the collation rules up to the specified length. Returns information * about whether a string is less than, greater than or equal to another * string in a language. This can be overriden in a subclass. * @param source the source string. * @param target the target string to be compared with the source string. * @param length compares up to the specified length * @return the comparison result. GREATER if the source string is greater * than the target string, LESS if the source is less than the * target. Otherwise, returns EQUAL. * @deprecated ICU 2.6 Use overload with UErrorCode& */ virtual EComparisonResult compare(const UnicodeString& source, const UnicodeString& target, int32_t length) const; /** * Does the same thing as compare but limits the comparison to a specified * length * @param source the source string to be compared with. * @param target the string that is to be compared with the source string. * @param length the length the comparison is limited to * @param status possible error code * @return Returns an enum value. UCOL_GREATER if source (up to the specified * length) is greater than target; UCOL_EQUAL if source (up to specified * length) is equal to target; UCOL_LESS if source (up to the specified * length) is less than target. * @draft ICU 2.6 */ virtual UCollationResult compare(const UnicodeString& source, const UnicodeString& target, int32_t length, UErrorCode &status) const; /** * The comparison function compares the character data stored in two * different string arrays. Returns information about whether a string array * is less than, greater than or equal to another string array. *
Example of use: *
* . UChar ABC[] = {0x41, 0x42, 0x43, 0}; // = "ABC"
* . UChar abc[] = {0x61, 0x62, 0x63, 0}; // = "abc"
* . UErrorCode status = U_ZERO_ERROR;
* . Collator *myCollation =
* . Collator::createInstance(Locale::US, status);
* . if (U_FAILURE(status)) return;
* . myCollation->setStrength(Collator::PRIMARY);
* . // result would be Collator::EQUAL ("abc" == "ABC")
* . // (no primary difference between "abc" and "ABC")
* . Collator::EComparisonResult result =
* . myCollation->compare(abc, 3, ABC, 3);
* . myCollation->setStrength(Collator::TERTIARY);
* . // result would be Collator::LESS ("abc" <<< "ABC")
* . // (with tertiary difference between "abc" and "ABC")
* . result = myCollation->compare(abc, 3, ABC, 3);
*
* @param source the source string array to be compared with.
* @param sourceLength the length of the source string array. If this value
* is equal to -1, the string array is null-terminated.
* @param target the string that is to be compared with the source string.
* @param targetLength the length of the target string array. If this value
* is equal to -1, the string array is null-terminated.
* @return Returns a byte value. GREATER if source is greater than target;
* EQUAL if source is equal to target; LESS if source is less than
* target
* @deprecated ICU 2.6 Use overload with UErrorCode&
*/
virtual EComparisonResult compare(const UChar* source, int32_t sourceLength,
const UChar* target, int32_t targetLength)
const;
/**
* The comparison function compares the character data stored in two
* different string arrays. Returns information about whether a string array
* is less than, greater than or equal to another string array.
* @param source the source string array to be compared with.
* @param sourceLength the length of the source string array. If this value
* is equal to -1, the string array is null-terminated.
* @param target the string that is to be compared with the source string.
* @param targetLength the length of the target string array. If this value
* is equal to -1, the string array is null-terminated.
* @param status possible error code
* @return Returns an enum value. UCOL_GREATER if source is greater
* than target; UCOL_EQUAL if source is equal to target; UCOL_LESS if source is less
* than target
* @draft ICU 2.6
*/
virtual UCollationResult compare(const UChar* source, int32_t sourceLength,
const UChar* target, int32_t targetLength,
UErrorCode &status) const;
/**
* Transforms a specified region of the string into a series of characters
* that can be compared with CollationKey.compare. Use a CollationKey when
* you need to do repeated comparisions on the same string. For a single
* comparison the compare method will be faster.
* @param source the source string.
* @param key the transformed key of the source string.
* @param status the error code status.
* @return the transformed key.
* @see CollationKey
* @stable ICU 2.0
*/
virtual CollationKey& getCollationKey(const UnicodeString& source,
CollationKey& key,
UErrorCode& status) const;
/**
* Transforms a specified region of the string into a series of characters
* that can be compared with CollationKey.compare. Use a CollationKey when
* you need to do repeated comparisions on the same string. For a single
* comparison the compare method will be faster.
* @param source the source string.
* @param key the transformed key of the source string.
* @param status the error code status.
* @return the transformed key.
* @see CollationKey
* @stable ICU 2.0
*/
virtual CollationKey& getCollationKey(const UChar *source,
int32_t sourceLength,
CollationKey& key,
UErrorCode& status) const;
/**
* Generates the hash code for the rule-based collation object.
* @return the hash code.
* @stable ICU 2.0
*/
virtual int32_t hashCode(void) const;
/**
* Gets the locale of the Collator
* @param type can be either requested, valid or actual locale. For more
* information see the definition of ULocDataLocaleType in
* uloc.h
* @param status the error code status.
* @return locale where the collation data lives. If the collator
* was instantiated from rules, locale is empty.
* @draft ICU 2.1
*/
virtual const Locale getLocale(ULocDataLocaleType type, UErrorCode& status) const;
/**
* Gets the table-based rules for the collation object.
* @return returns the collation rules that the table collation object was
* created from.
* @stable ICU 2.0
*/
const UnicodeString& getRules(void) const;
/**
* Gets the version information for a Collator.
* @param info the version # information, the result will be filled in
* @stable ICU 2.0
*/
virtual void getVersion(UVersionInfo info) const;
/**
* Return the maximum length of any expansion sequences that end with the
* specified comparison order.
* @param order a collation order returned by previous or next.
* @return maximum size of the expansion sequences ending with the collation
* element or 1 if collation element does not occur at the end of
* any expansion sequence
* @see CollationElementIterator#getMaxExpansion
* @stable ICU 2.0
*/
int32_t getMaxExpansion(int32_t order) const;
/**
* Returns a unique class ID POLYMORPHICALLY. Pure virtual override. This
* method is to implement a simple version of RTTI, since not all C++
* compilers support genuine RTTI. Polymorphic operator==() and clone()
* methods call this method.
* @return The class ID for this object. All objects of a given class have
* the same class ID. Objects of other classes have different class
* IDs.
* @stable ICU 2.0
*/
virtual UClassID getDynamicClassID(void) const;
/**
* Returns the class ID for this class. This is useful only for comparing to
* a return value from getDynamicClassID(). For example:
* * Base* polymorphic_pointer = createPolymorphicObject(); * if (polymorphic_pointer->getDynamicClassID() == * Derived::getStaticClassID()) ... ** @return The class ID for all objects of this class. * @stable ICU 2.0 */ static UClassID getStaticClassID(void); /** * Returns the binary format of the class's rules. The format is that of * .col files. * @param length Returns the length of the data, in bytes * @param status the error code status. * @return memory, owned by the caller, of size 'length' bytes. * @stable ICU 2.2 */ uint8_t *cloneRuleData(int32_t &length, UErrorCode &status); /** * Returns current rules. Delta defines whether full rules are returned or * just the tailoring. * @param delta one of UCOL_TAILORING_ONLY, UCOL_FULL_RULES. * @param buffer UnicodeString to store the result rules * @stable ICU 2.2 */ void getRules(UColRuleOption delta, UnicodeString &buffer); /** * Universal attribute setter * @param attr attribute type * @param value attribute value * @param status to indicate whether the operation went on smoothly or there were errors * @stable ICU 2.2 */ virtual void setAttribute(UColAttribute attr, UColAttributeValue value, UErrorCode &status); /** * Universal attribute getter. * @param attr attribute type * @param status to indicate whether the operation went on smoothly or there were errors * @return attribute value * @stable ICU 2.2 */ virtual UColAttributeValue getAttribute(UColAttribute attr, UErrorCode &status); /** * Sets the variable top to a collation element value of a string supplied. * @param varTop one or more (if contraction) UChars to which the variable top should be set * @param len length of variable top string. If -1 it is considered to be zero terminated. * @param status error code. If error code is set, the return value is undefined. Errors set by this function are:
E.g. with strength == SECONDARY, the tertiary difference is ignored *
E.g. with strength == PRIMARY, the secondary and tertiary difference * are ignored. * @return the current comparison level. * @see RuleBasedCollator#setStrength * @deprecated ICU 2.6 Use getAttribute(UCOL_STRENGTH...) instead */ virtual ECollationStrength getStrength(void) const; /** * Sets the minimum strength to be used in comparison or transformation. * @see RuleBasedCollator#getStrength * @param newStrength the new comparison level. * @deprecated ICU 2.6 Use setAttribute(UCOL_STRENGTH...) instead */ virtual void setStrength(ECollationStrength newStrength); private: // private static constants ----------------------------------------------- static const int32_t UNMAPPED; static const int32_t CHARINDEX; // need look up in .commit() static const int32_t EXPANDCHARINDEX; // Expand index follows static const int32_t CONTRACTCHARINDEX; // contract indexes follow static const int32_t PRIMARYORDERINCREMENT; static const int32_t SECONDARYORDERINCREMENT; static const int32_t TERTIARYORDERINCREMENT; static const int32_t PRIMARYORDERMASK; static const int32_t SECONDARYORDERMASK; static const int32_t TERTIARYORDERMASK; static const int32_t IGNORABLEMASK; static const int32_t PRIMARYDIFFERENCEONLY; static const int32_t SECONDARYDIFFERENCEONLY; static const int32_t PRIMARYORDERSHIFT; static const int32_t SECONDARYORDERSHIFT; static const int32_t COLELEMENTSTART; static const int32_t PRIMARYLOWZEROMASK; static const int32_t RESETSECONDARYTERTIARY; static const int32_t RESETTERTIARY; static const int32_t PRIMIGNORABLE; static const int16_t FILEID; static const char kFilenameSuffix[]; // private static variables ----------------------------------------------- /** * static class id */ static const char fgClassID; // private data members --------------------------------------------------- UBool dataIsOwned; UBool isWriteThroughAlias; /** * c struct for collation. All initialisation for it has to be done through * setUCollator(). */ UCollator *ucollator; /** * Rule UnicodeString */ UnicodeString *urulestring; // friend classes -------------------------------------------------------- /** * Used to iterate over collation elements in a character source. */ friend class CollationElementIterator; /** * Collator ONLY needs access to RuleBasedCollator(const Locale&, * UErrorCode&) */ friend class Collator; /** * Searching over collation elements in a character source */ friend class StringSearch; // private constructors -------------------------------------------------- /** * Default constructor */ RuleBasedCollator(); /** * Constructor that takes in a UCollator struct * @param collator UCollator struct * @param rule the rule for the collator. */ RuleBasedCollator(UCollator *collator, UnicodeString *rule); /** * RuleBasedCollator constructor. This constructor takes a locale. The * only caller of this class should be Collator::createInstance(). If * createInstance() happens to know that the requested locale's collation is * implemented as a RuleBasedCollator, it can then call this constructor. * OTHERWISE IT SHOULDN'T, since this constructor ALWAYS RETURNS A VALID * COLLATION TABLE. It does this by falling back to defaults. * @param desiredLocale locale used * @param status error code status */ RuleBasedCollator(const Locale& desiredLocale, UErrorCode& status); /** * common constructor implementation * * @param rules the collation rules to build the collation table from. * @param collationStrength default strength for comparison * @param decompositionMode the normalisation mode * @param status reporting a success or an error. */ void construct(const UnicodeString& rules, UColAttributeValue collationStrength, UColAttributeValue decompositionMode, UErrorCode& status); // private methods ------------------------------------------------------- /** * Creates the c struct for ucollator * @param locale desired locale * @param status error status */ void setUCollator(const Locale& locale, UErrorCode& status); /** * Creates the c struct for ucollator * @param locale desired locale name * @param status error status */ void setUCollator(const char* locale, UErrorCode& status); /** * Creates the c struct for ucollator. This used internally by StringSearch. * Hence the responsibility of cleaning up the ucollator is not done by * this RuleBasedCollator. The isDataOwned flag is set to FALSE. * @param collator new ucollator data * @param rules corresponding collation rules */ void setUCollator(UCollator *collator, UnicodeString *rules); public: /** * Get UCollator data struct. Used only by StringSearch & intltest. * @return UCollator data struct * @internal */ const UCollator * getUCollator(); protected: /** * Used internally by registraton to define the requested and valid locales. * @param requestedLocale the requsted locale * @param validLocale the valid locale * @internal */ virtual void setLocales(const Locale& requestedLocale, const Locale& validLocale); private: // if not owned and not a write through alias, copy the ucollator void checkOwned(void); // utility to init rule string used by checkOwned and construct void setRuleStringFromCollator(UErrorCode& status); /** * Converts C's UCollationResult to EComparisonResult * @param result member of the enum UComparisonResult * @return EComparisonResult equivalent of UCollationResult * @deprecated ICU 2.6. We will not need it. */ Collator::EComparisonResult getEComparisonResult( const UCollationResult &result) const; /** * Converts C's UCollationStrength to ECollationStrength * @param strength member of the enum UCollationStrength * @return ECollationStrength equivalent of UCollationStrength */ Collator::ECollationStrength getECollationStrength( const UCollationStrength &strength) const; /** * Converts C++'s ECollationStrength to UCollationStrength * @param strength member of the enum ECollationStrength * @return UCollationStrength equivalent of ECollationStrength */ UCollationStrength getUCollationStrength( const Collator::ECollationStrength &strength) const; }; // inline method implementation --------------------------------------------- inline UClassID RuleBasedCollator::getStaticClassID(void) { return (UClassID)&fgClassID; } inline UClassID RuleBasedCollator::getDynamicClassID(void) const { return RuleBasedCollator::getStaticClassID(); } inline UBool RuleBasedCollator::operator!=(const Collator& other) const { return !(*this == other); } inline void RuleBasedCollator::setUCollator(const Locale &locale, UErrorCode &status) { setUCollator(locale.getName(), status); } inline void RuleBasedCollator::setUCollator(UCollator *collator, UnicodeString *rules) { if (ucollator && dataIsOwned) { ucol_close(ucollator); delete urulestring; } ucollator = collator; urulestring = rules; dataIsOwned = FALSE; isWriteThroughAlias = TRUE; } inline const UCollator * RuleBasedCollator::getUCollator() { return ucollator; } inline Collator::EComparisonResult RuleBasedCollator::getEComparisonResult( const UCollationResult &result) const { switch (result) { case UCOL_LESS : return Collator::LESS; case UCOL_EQUAL : return Collator::EQUAL; default : return Collator::GREATER; } } inline Collator::ECollationStrength RuleBasedCollator::getECollationStrength( const UCollationStrength &strength) const { switch (strength) { case UCOL_PRIMARY : return Collator::PRIMARY; case UCOL_SECONDARY : return Collator::SECONDARY; case UCOL_TERTIARY : return Collator::TERTIARY; case UCOL_QUATERNARY : return Collator::QUATERNARY; default : return Collator::IDENTICAL; } } inline UCollationStrength RuleBasedCollator::getUCollationStrength( const Collator::ECollationStrength &strength) const { switch (strength) { case Collator::PRIMARY : return UCOL_PRIMARY; case Collator::SECONDARY : return UCOL_SECONDARY; case Collator::TERTIARY : return UCOL_TERTIARY; case Collator::QUATERNARY : return UCOL_QUATERNARY; default : return UCOL_IDENTICAL; } } U_NAMESPACE_END #endif /* #if !UCONFIG_NO_COLLATION */ #endif