+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
/*
******************************************************************************
-* Copyright (C) 1996-2010, International Business Machines Corporation and
+* Copyright (C) 1996-2016, International Business Machines Corporation and
* others. All Rights Reserved.
******************************************************************************
*/
/**
- * \file
- * \brief C++ API: RuleBasedCollator class provides the simple implementation of Collator.
+ * \file
+ * \brief C++ API: The RuleBasedCollator class implements the Collator abstract base class.
*/
/**
* to implementation file.
* 01/29/01 synwee Modified into a C++ wrapper which calls C API
* (ucol.h)
+* 2012-2014 markus Rewritten in C++ again.
*/
#ifndef TBLCOLL_H
#include "unicode/utypes.h"
-
#if !UCONFIG_NO_COLLATION
#include "unicode/coll.h"
+#include "unicode/locid.h"
+#include "unicode/uiter.h"
#include "unicode/ucol.h"
-#include "unicode/sortkey.h"
-#include "unicode/normlzr.h"
+#if U_SHOW_CPLUSPLUS_API
U_NAMESPACE_BEGIN
+struct CollationCacheEntry;
+struct CollationData;
+struct CollationSettings;
+struct CollationTailoring;
/**
* @stable ICU 2.0
*/
* @stable ICU 2.0
*/
class CollationElementIterator;
+class CollationKey;
+class SortKeyByteSink;
+class UnicodeSet;
+class UnicodeString;
+class UVector64;
/**
- * The RuleBasedCollator class provides the simple implementation of
+ * The RuleBasedCollator class provides the implementation of
* Collator, using data-driven tables. The user can create a customized
* table-based collation.
- * <P>
- * <em>Important: </em>The ICU collation service has been reimplemented
- * in order to achieve better performance and UCA compliance.
- * For details, see the
- * <a href="http://source.icu-project.org/repos/icu/icuhtml/trunk/design/collation/ICU_collation_design.htm">
- * collation design document</a>.
* <p>
- * RuleBasedCollator is a thin C++ wrapper over the C implementation.
+ * For more information about the collation service see
+ * <a href="http://userguide.icu-project.org/collation">the User Guide</a>.
* <p>
- * For more information about the collation service see
- * <a href="http://icu-project.org/userguide/Collate_Intro.html">the users guide</a>.
- * <p>
- * Collation service provides correct sorting orders for most locales supported in ICU.
+ * 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 <a href="http://www.unicode.org/unicode/reports/tr10/">UCA sort order</a>.
+ * to the <a href="http://www.unicode.org/reports/tr35/tr35-collation.html#Root_Collation">CLDR root sort order</a>.
* <p>
* Sort ordering may be customized by providing your own set of rules. For more on
- * this subject see the <a href="http://icu-project.org/userguide/Collate_Customization.html">
- * Collation customization</a> section of the users guide.
+ * this subject see the <a href="http://userguide.icu-project.org/collation/customization">
+ * Collation Customization</a> section of the User Guide.
* <p>
* Note, RuleBasedCollator is not to be subclassed.
* @see Collator
- * @version 2.0 11/15/2001
*/
-class U_I18N_API RuleBasedCollator : public Collator
-{
+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);
* 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 collationStrength strength for comparison
* @param status reporting a success or an error.
- * @see Locale
* @stable ICU 2.0
*/
RuleBasedCollator(const UnicodeString& rules,
* @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,
* 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 collationStrength 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,
UColAttributeValue decompositionMode,
UErrorCode& status);
+#ifndef U_HIDE_INTERNAL_API
+ /**
+ * TODO: document & propose as public API
+ * @internal
+ */
+ RuleBasedCollator(const UnicodeString &rules,
+ UParseError &parseError, UnicodeString &reason,
+ UErrorCode &errorCode);
+#endif /* U_HIDE_INTERNAL_API */
+
/**
* Copy constructor.
* @param other the RuleBasedCollator object to be copied
- * @see Locale
* @stable ICU 2.0
*/
RuleBasedCollator(const RuleBasedCollator& other);
/** Opens a collator from a collator binary image created using
- * cloneBinary. Binary image used in instantiation of the
- * collator remains owned by the user and should stay around for
+ * cloneBinary. Binary image used in instantiation of the
+ * collator remains owned by the user and should stay around for
* the lifetime of the collator. The API also takes a base collator
- * which usualy should be UCA.
+ * which must be the root collator.
* @param bin binary image owned by the user and required through the
* lifetime of the collator
* @param length size of the image. If negative, the API will try to
* figure out the length of the image
- * @param base fallback collator, usually UCA. Base is required to be
- * present through the lifetime of the collator. Currently
- * it cannot be NULL.
+ * @param base Base collator, for lookup of untailored characters.
+ * Must be the root collator, must not be NULL.
+ * The base is required to be present through the lifetime of the collator.
* @param status for catching errors
* @return newly created collator
* @see cloneBinary
* @stable ICU 3.4
*/
- RuleBasedCollator(const uint8_t *bin, int32_t length,
- const RuleBasedCollator *base,
+ RuleBasedCollator(const uint8_t *bin, int32_t length,
+ const RuleBasedCollator *base,
UErrorCode &status);
- // destructor --------------------------------------------------------------
/**
* Destructor.
*/
virtual ~RuleBasedCollator();
- // public methods ----------------------------------------------------------
-
/**
* Assignment operator.
- * @param other other RuleBasedCollator object to compare with.
+ * @param other other RuleBasedCollator object to copy from.
* @stable ICU 2.0
*/
RuleBasedCollator& operator=(const RuleBasedCollator& other);
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.
+ * Makes a copy of this object.
+ * @return a copy of this object, owned by the caller
* @stable ICU 2.0
*/
virtual Collator* clone(void) const;
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;
-
+ // Make deprecated versions of Collator::compare() visible.
+ using Collator::compare;
/**
* The comparison function compares the character data stored in two
- * different strings. Returns information about whether a string is less
+ * 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.
* @stable 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;
+ const UnicodeString& target,
+ UErrorCode &status) const;
/**
- * Does the same thing as compare but limits the comparison to a specified
+ * 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
+ * @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.
* @stable 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.
- * <p>Example of use:
- * <pre>
- * . 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);
- * </pre>
- * @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;
+ 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
+ * 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
* than target
* @stable ICU 2.6
*/
- virtual UCollationResult compare(const UChar* source, int32_t sourceLength,
- const UChar* target, int32_t targetLength,
- UErrorCode &status) const;
+ virtual UCollationResult compare(const char16_t* source, int32_t sourceLength,
+ const char16_t* target, int32_t targetLength,
+ UErrorCode &status) const;
/**
* Compares two strings using the Collator.
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
- * @deprecated ICU 2.8 Use getSortKey(...) instead
- */
+ * Compares two UTF-8 strings using the Collator.
+ * Returns whether the first one compares less than/equal to/greater than
+ * the second one.
+ * This version takes UTF-8 input.
+ * Note that a StringPiece can be implicitly constructed
+ * from a std::string or a NUL-terminated const char * string.
+ * @param source the first UTF-8 string
+ * @param target the second UTF-8 string
+ * @param status ICU status
+ * @return UCOL_LESS, UCOL_EQUAL or UCOL_GREATER
+ * @stable ICU 51
+ */
+ virtual UCollationResult compareUTF8(const StringPiece &source,
+ const StringPiece &target,
+ UErrorCode &status) const;
+
+ /**
+ * Transforms the string into a series of characters
+ * that can be compared with CollationKey.compare().
+ *
+ * Note that sort keys are often less efficient than simply doing comparison.
+ * For more details, see the ICU User Guide.
+ *
+ * @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 sourceLength the length of the source string.
- * @param key the transformed key of the source string.
- * @param status the error code status.
- * @return the transformed key.
- * @see CollationKey
- * @deprecated ICU 2.8 Use getSortKey(...) instead
- */
- virtual CollationKey& getCollationKey(const UChar *source,
+ * Transforms a specified region of the string into a series of characters
+ * that can be compared with CollationKey.compare.
+ *
+ * Note that sort keys are often less efficient than simply doing comparison.
+ * For more details, see the ICU User Guide.
+ *
+ * @param source the source string.
+ * @param sourceLength the length of 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 char16_t *source,
int32_t sourceLength,
CollationKey& key,
UErrorCode& status) const;
* @return the hash code.
* @stable ICU 2.0
*/
- virtual int32_t hashCode(void) const;
+ virtual int32_t hashCode() const;
/**
* Gets the locale of the Collator
* was instantiated from rules, locale is empty.
* @deprecated ICU 2.8 likely to change in ICU 3.0, based on feedback
*/
- virtual const Locale getLocale(ULocDataLocaleType type, UErrorCode& status) const;
+ virtual 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.
+ * Gets the tailoring rules for this collator.
+ * @return the collation tailoring from which this collator was created
* @stable ICU 2.0
*/
- const UnicodeString& getRules(void) const;
+ const UnicodeString& getRules() const;
/**
* Gets the version information for a Collator.
*/
virtual void getVersion(UVersionInfo info) const;
+#ifndef U_HIDE_DEPRECATED_API
/**
- * Return the maximum length of any expansion sequences that end with the
+ * Returns the maximum length of any expansion sequences that end with the
* specified comparison order.
- * @param order a collation order returned by previous or next.
+ *
+ * This is specific to the kind of collation element values and sequences
+ * returned by the CollationElementIterator.
+ * Call CollationElementIterator::getMaxExpansion() instead.
+ *
+ * @param order a collation order returned by CollationElementIterator::previous
+ * or CollationElementIterator::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
+ * element, or 1 if the collation element does not occur at the end of
* any expansion sequence
* @see CollationElementIterator#getMaxExpansion
- * @stable ICU 2.0
+ * @deprecated ICU 51 Use CollationElementIterator::getMaxExpansion() instead.
*/
int32_t getMaxExpansion(int32_t order) const;
+#endif /* U_HIDE_DEPRECATED_API */
/**
* Returns a unique class ID POLYMORPHICALLY. Pure virtual override. This
*/
static UClassID U_EXPORT2 getStaticClassID(void);
+#ifndef U_HIDE_DEPRECATED_API
/**
- * Returns the binary format of the class's rules. The format is that of
- * .col files.
+ * Do not use this method: The caller and the ICU library might use different heaps.
+ * Use cloneBinary() instead which writes to caller-provided memory.
+ *
+ * Returns a binary format of this collator.
* @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
+ * @deprecated ICU 52. Use cloneBinary() instead.
*/
- uint8_t *cloneRuleData(int32_t &length, UErrorCode &status);
-
+ uint8_t *cloneRuleData(int32_t &length, UErrorCode &status) const;
+#endif /* U_HIDE_DEPRECATED_API */
- /** Creates a binary image of a collator. This binary image can be stored and
+ /** Creates a binary image of a collator. This binary image can be stored and
* later used to instantiate a collator using ucol_openBinary.
* This API supports preflighting.
* @param buffer a fill-in buffer to receive the binary image
* @see ucol_openBinary
* @stable ICU 3.4
*/
- int32_t cloneBinary(uint8_t *buffer, int32_t capacity, UErrorCode &status);
+ int32_t cloneBinary(uint8_t *buffer, int32_t capacity, UErrorCode &status) const;
/**
* Returns current rules. Delta defines whether full rules are returned or
* just the tailoring.
+ *
+ * getRules(void) should normally be used instead.
+ * See http://userguide.icu-project.org/collation/customization#TOC-Building-on-Existing-Locales
* @param delta one of UCOL_TAILORING_ONLY, UCOL_FULL_RULES.
* @param buffer UnicodeString to store the result rules
* @stable ICU 2.2
+ * @see UCOL_FULL_RULES
*/
- void getRules(UColRuleOption delta, UnicodeString &buffer);
+ void getRules(UColRuleOption delta, UnicodeString &buffer) const;
/**
* Universal attribute setter
* @stable ICU 2.2
*/
virtual UColAttributeValue getAttribute(UColAttribute attr,
- UErrorCode &status);
+ UErrorCode &status) const;
+
+ /**
+ * Sets the variable top to the top of the specified reordering group.
+ * The variable top determines the highest-sorting character
+ * which is affected by UCOL_ALTERNATE_HANDLING.
+ * If that attribute is set to UCOL_NON_IGNORABLE, then the variable top has no effect.
+ * @param group one of UCOL_REORDER_CODE_SPACE, UCOL_REORDER_CODE_PUNCTUATION,
+ * UCOL_REORDER_CODE_SYMBOL, UCOL_REORDER_CODE_CURRENCY;
+ * or UCOL_REORDER_CODE_DEFAULT to restore the default max variable group
+ * @param errorCode Standard ICU error code. Its input value must
+ * pass the U_SUCCESS() test, or else the function returns
+ * immediately. Check for U_FAILURE() on output or use with
+ * function chaining. (See User Guide for details.)
+ * @return *this
+ * @see getMaxVariable
+ * @stable ICU 53
+ */
+ virtual Collator &setMaxVariable(UColReorderCode group, UErrorCode &errorCode);
+
+ /**
+ * Returns the maximum reordering group whose characters are affected by UCOL_ALTERNATE_HANDLING.
+ * @return the maximum variable reordering group.
+ * @see setMaxVariable
+ * @stable ICU 53
+ */
+ virtual UColReorderCode getMaxVariable() const;
/**
- * 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
+ * Sets the variable top to the primary weight of the specified string.
+ *
+ * Beginning with ICU 53, the variable top is pinned to
+ * the top of one of the supported reordering groups,
+ * and it must not be beyond the last of those groups.
+ * See setMaxVariable().
+ * @param varTop one or more (if contraction) char16_ts 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: <br>
- * U_CE_NOT_FOUND_ERROR if more than one character was passed and there is no such a contraction<br>
- * U_PRIMARY_TOO_LONG_ERROR if the primary for the variable top has more than two bytes
- * @return a 32 bit value containing the value of the variable top in upper 16 bits. Lower 16 bits are undefined
- * @stable ICU 2.0
+ * U_CE_NOT_FOUND_ERROR if more than one character was passed and there is no such contraction<br>
+ * U_ILLEGAL_ARGUMENT_ERROR if the variable top is beyond
+ * the last reordering group supported by setMaxVariable()
+ * @return variable top primary weight
+ * @deprecated ICU 53 Call setMaxVariable() instead.
*/
- virtual uint32_t setVariableTop(const UChar *varTop, int32_t len, UErrorCode &status);
+ virtual uint32_t setVariableTop(const char16_t *varTop, int32_t len, UErrorCode &status);
/**
- * Sets the variable top to a collation element value of a string supplied.
- * @param varTop an UnicodeString size 1 or more (if contraction) of UChars to which the variable top should be set
+ * Sets the variable top to the primary weight of the specified string.
+ *
+ * Beginning with ICU 53, the variable top is pinned to
+ * the top of one of the supported reordering groups,
+ * and it must not be beyond the last of those groups.
+ * See setMaxVariable().
+ * @param varTop a UnicodeString size 1 or more (if contraction) of char16_ts to which the variable top should be set
* @param status error code. If error code is set, the return value is undefined. Errors set by this function are: <br>
- * U_CE_NOT_FOUND_ERROR if more than one character was passed and there is no such a contraction<br>
- * U_PRIMARY_TOO_LONG_ERROR if the primary for the variable top has more than two bytes
- * @return a 32 bit value containing the value of the variable top in upper 16 bits. Lower 16 bits are undefined
- * @stable ICU 2.0
+ * U_CE_NOT_FOUND_ERROR if more than one character was passed and there is no such contraction<br>
+ * U_ILLEGAL_ARGUMENT_ERROR if the variable top is beyond
+ * the last reordering group supported by setMaxVariable()
+ * @return variable top primary weight
+ * @deprecated ICU 53 Call setMaxVariable() instead.
*/
- virtual uint32_t setVariableTop(const UnicodeString varTop, UErrorCode &status);
+ virtual uint32_t setVariableTop(const UnicodeString &varTop, UErrorCode &status);
/**
- * Sets the variable top to a collation element value supplied. Variable top is set to the upper 16 bits.
- * Lower 16 bits are ignored.
- * @param varTop CE value, as returned by setVariableTop or ucol)getVariableTop
- * @param status error code (not changed by function)
- * @stable ICU 2.0
+ * Sets the variable top to the specified primary weight.
+ *
+ * Beginning with ICU 53, the variable top is pinned to
+ * the top of one of the supported reordering groups,
+ * and it must not be beyond the last of those groups.
+ * See setMaxVariable().
+ * @param varTop primary weight, as returned by setVariableTop or ucol_getVariableTop
+ * @param status error code
+ * @deprecated ICU 53 Call setMaxVariable() instead.
*/
- virtual void setVariableTop(const uint32_t varTop, UErrorCode &status);
+ virtual void setVariableTop(uint32_t varTop, UErrorCode &status);
/**
* Gets the variable top value of a Collator.
- * Lower 16 bits are undefined and should be ignored.
* @param status error code (not changed by function). If error code is set, the return value is undefined.
+ * @return the variable top primary weight
+ * @see getMaxVariable
* @stable ICU 2.0
*/
virtual uint32_t getVariableTop(UErrorCode &status) const;
/**
- * Get an UnicodeSet that contains all the characters and sequences tailored in
+ * Get a UnicodeSet that contains all the characters and sequences tailored in
* this collator.
* @param status error code of the operation
- * @return a pointer to a UnicodeSet object containing all the
+ * @return a pointer to a UnicodeSet object containing all the
* code points and sequences that may sort differently than
- * in the UCA. The object must be disposed of by using delete
+ * in the root collator. The object must be disposed of by using delete
* @stable ICU 2.4
*/
virtual UnicodeSet *getTailoredSet(UErrorCode &status) const;
/**
- * Thread safe cloning operation.
- * @return pointer to the new clone, user should remove it.
- * @stable ICU 2.2
- */
- virtual Collator* safeClone(void);
-
- /**
- * Get the sort key as an array of bytes from an UnicodeString.
+ * Get the sort key as an array of bytes from a UnicodeString.
+ *
+ * Note that sort keys are often less efficient than simply doing comparison.
+ * For more details, see the ICU User Guide.
+ *
* @param source string to be processed.
* @param result buffer to store result in. If NULL, number of bytes needed
* will be returned.
int32_t resultLength) const;
/**
- * Get the sort key as an array of bytes from an UChar buffer.
+ * Get the sort key as an array of bytes from a char16_t buffer.
+ *
+ * Note that sort keys are often less efficient than simply doing comparison.
+ * For more details, see the ICU User Guide.
+ *
* @param source string to be processed.
* @param sourceLength length of string to be processed. If -1, the string
* is 0 terminated and length will be decided by the function.
* @return Number of bytes needed for storing the sort key
* @stable ICU 2.2
*/
- virtual int32_t getSortKey(const UChar *source, int32_t sourceLength,
+ virtual int32_t getSortKey(const char16_t *source, int32_t sourceLength,
uint8_t *result, int32_t resultLength) const;
/**
- * Determines the minimum strength that will be use in comparison or
- * transformation.
- * <p>E.g. with strength == SECONDARY, the tertiary difference is ignored
- * <p>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);
-
- /**
- * Get the current reordering of scripts (if one has been set).
+ * Retrieves the reordering codes for this collator.
* @param dest The array to fill with the script ordering.
- * @param destCapacity The length of dest. If it is 0, then dest may be NULL and the function will only return the length of the result without writing any of the result string (pre-flighting).
- * @param pErrorCode Must be a valid pointer to an error code value, which must not indicate a failure before the function call.
- * @return The length of the array of the script ordering.
- * @see ucol_getReorderCodes
- * @internal
+ * @param destCapacity The length of dest. If it is 0, then dest may be NULL and the function
+ * will only return the length of the result without writing any codes (pre-flighting).
+ * @param status A reference to an error code value, which must not indicate
+ * a failure before the function call.
+ * @return The length of the script ordering array.
+ * @see ucol_setReorderCodes
+ * @see Collator#getEquivalentReorderCodes
+ * @see Collator#setReorderCodes
+ * @stable ICU 4.8
*/
- virtual int32_t getReorderCodes(int32_t* dest,
- int32_t destCapacity,
- UErrorCode& status) const;
+ virtual int32_t getReorderCodes(int32_t *dest,
+ int32_t destCapacity,
+ UErrorCode& status) const;
/**
- * Set the ordering of scripts for this collator.
- * @param reorderCodes An array of script codes in the new order.
+ * Sets the ordering of scripts for this collator.
+ * @param reorderCodes An array of script codes in the new order. This can be NULL if the
+ * length is also set to 0. An empty array will clear any reordering codes on the collator.
* @param reorderCodesLength The length of reorderCodes.
+ * @param status error code
* @see ucol_setReorderCodes
- * @internal
+ * @see Collator#getReorderCodes
+ * @see Collator#getEquivalentReorderCodes
+ * @stable ICU 4.8
+ */
+ virtual void setReorderCodes(const int32_t* reorderCodes,
+ int32_t reorderCodesLength,
+ UErrorCode& status) ;
+
+ /**
+ * Implements ucol_strcollUTF8().
+ * @internal
+ */
+ virtual UCollationResult internalCompareUTF8(
+ const char *left, int32_t leftLength,
+ const char *right, int32_t rightLength,
+ UErrorCode &errorCode) const;
+
+ /** Get the short definition string for a collator. This internal API harvests the collator's
+ * locale and the attribute set and produces a string that can be used for opening
+ * a collator with the same attributes using the ucol_openFromShortString API.
+ * This string will be normalized.
+ * The structure and the syntax of the string is defined in the "Naming collators"
+ * section of the users guide:
+ * http://userguide.icu-project.org/collation/concepts#TOC-Collator-naming-scheme
+ * This function supports preflighting.
+ *
+ * This is internal, and intended to be used with delegate converters.
+ *
+ * @param locale a locale that will appear as a collators locale in the resulting
+ * short string definition. If NULL, the locale will be harvested
+ * from the collator.
+ * @param buffer space to hold the resulting string
+ * @param capacity capacity of the buffer
+ * @param status for returning errors. All the preflighting errors are featured
+ * @return length of the resulting string
+ * @see ucol_openFromShortString
+ * @see ucol_normalizeShortDefinitionString
+ * @see ucol_getShortDefinitionString
+ * @internal
*/
- virtual void setReorderCodes(const int32_t* reorderCodes,
- int32_t reorderCodesLength,
- UErrorCode& status);
-
-
-private:
-
- // private static constants -----------------------------------------------
-
- enum {
- /* need look up in .commit() */
- CHARINDEX = 0x70000000,
- /* Expand index follows */
- EXPANDCHARINDEX = 0x7E000000,
- /* contract indexes follows */
- CONTRACTCHARINDEX = 0x7F000000,
- /* unmapped character values */
- UNMAPPED = 0xFFFFFFFF,
- /* primary strength increment */
- PRIMARYORDERINCREMENT = 0x00010000,
- /* secondary strength increment */
- SECONDARYORDERINCREMENT = 0x00000100,
- /* tertiary strength increment */
- TERTIARYORDERINCREMENT = 0x00000001,
- /* mask off anything but primary order */
- PRIMARYORDERMASK = 0xffff0000,
- /* mask off anything but secondary order */
- SECONDARYORDERMASK = 0x0000ff00,
- /* mask off anything but tertiary order */
- TERTIARYORDERMASK = 0x000000ff,
- /* mask off ignorable char order */
- IGNORABLEMASK = 0x0000ffff,
- /* use only the primary difference */
- PRIMARYDIFFERENCEONLY = 0xffff0000,
- /* use only the primary and secondary difference */
- SECONDARYDIFFERENCEONLY = 0xffffff00,
- /* primary order shift */
- PRIMARYORDERSHIFT = 16,
- /* secondary order shift */
- SECONDARYORDERSHIFT = 8,
- /* starting value for collation elements */
- COLELEMENTSTART = 0x02020202,
- /* testing mask for primary low element */
- PRIMARYLOWZEROMASK = 0x00FF0000,
- /* reseting value for secondaries and tertiaries */
- RESETSECONDARYTERTIARY = 0x00000202,
- /* reseting value for tertiaries */
- RESETTERTIARY = 0x00000002,
-
- PRIMIGNORABLE = 0x0202
- };
-
- // 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 --------------------------------------------------------
+ virtual int32_t internalGetShortDefinitionString(const char *locale,
+ char *buffer,
+ int32_t capacity,
+ UErrorCode &status) const;
/**
- * 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 --------------------------------------------------
+ * Implements ucol_nextSortKeyPart().
+ * @internal
+ */
+ virtual int32_t internalNextSortKeyPart(
+ UCharIterator *iter, uint32_t state[2],
+ uint8_t *dest, int32_t count, UErrorCode &errorCode) const;
+ // Do not enclose the default constructor with #ifndef U_HIDE_INTERNAL_API
/**
- * Default constructor
+ * Only for use in ucol_openRules().
+ * @internal
*/
RuleBasedCollator();
+#ifndef U_HIDE_INTERNAL_API
/**
- * 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
+ * Implements ucol_getLocaleByType().
+ * Needed because the lifetime of the locale ID string must match that of the collator.
+ * getLocale() returns a copy of a Locale, with minimal lifetime in a C wrapper.
+ * @internal
*/
- RuleBasedCollator(const Locale& desiredLocale, UErrorCode& status);
+ const char *internalGetLocaleID(ULocDataLocaleType type, UErrorCode &errorCode) const;
/**
- * 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.
+ * Implements ucol_getContractionsAndExpansions().
+ * Gets this collator's sets of contraction strings and/or
+ * characters and strings that map to multiple collation elements (expansions).
+ * If addPrefixes is TRUE, then contractions that are expressed as
+ * prefix/pre-context rules are included.
+ * @param contractions if not NULL, the set to hold the contractions
+ * @param expansions if not NULL, the set to hold the expansions
+ * @param addPrefixes include prefix contextual mappings
+ * @param errorCode in/out ICU error code
+ * @internal
*/
- void
- construct(const UnicodeString& rules,
- UColAttributeValue collationStrength,
- UColAttributeValue decompositionMode,
- UErrorCode& status);
-
- // private methods -------------------------------------------------------
+ void internalGetContractionsAndExpansions(
+ UnicodeSet *contractions, UnicodeSet *expansions,
+ UBool addPrefixes, UErrorCode &errorCode) const;
/**
- * Creates the c struct for ucollator
- * @param locale desired locale
- * @param status error status
- */
- void setUCollator(const Locale& locale, UErrorCode& status);
+ * Adds the contractions that start with character c to the set.
+ * Ignores prefixes. Used by AlphabeticIndex.
+ * @internal
+ */
+ void internalAddContractions(UChar32 c, UnicodeSet &set, UErrorCode &errorCode) const;
/**
- * Creates the c struct for ucollator
- * @param locale desired locale name
- * @param status error status
- */
- void setUCollator(const char* locale, UErrorCode& status);
+ * Implements from-rule constructors, and ucol_openRules().
+ * @internal
+ */
+ void internalBuildTailoring(
+ const UnicodeString &rules,
+ int32_t strength,
+ UColAttributeValue decompositionMode,
+ UParseError *outParseError, UnicodeString *outReason,
+ UErrorCode &errorCode);
- /**
- * 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);
+ /** @internal */
+ static inline RuleBasedCollator *rbcFromUCollator(UCollator *uc) {
+ return dynamic_cast<RuleBasedCollator *>(fromUCollator(uc));
+ }
+ /** @internal */
+ static inline const RuleBasedCollator *rbcFromUCollator(const UCollator *uc) {
+ return dynamic_cast<const RuleBasedCollator *>(fromUCollator(uc));
+ }
-public:
/**
- * Get UCollator data struct. Used only by StringSearch & intltest.
- * @return UCollator data struct
- * @internal
- */
- const UCollator * getUCollator();
+ * Appends the CEs for the string to the vector.
+ * @internal for tests & tools
+ */
+ void internalGetCEs(const UnicodeString &str, UVector64 &ces, UErrorCode &errorCode) const;
+#endif // U_HIDE_INTERNAL_API
protected:
/**
- * Used internally by registraton to define the requested and valid locales.
- * @param requestedLocale the requsted locale
+ * Used internally by registration to define the requested and valid locales.
+ * @param requestedLocale the requested locale
* @param validLocale the valid locale
* @param actualLocale the actual locale
* @internal
virtual void setLocales(const Locale& requestedLocale, const Locale& validLocale, const Locale& actualLocale);
private:
+ friend class CollationElementIterator;
+ friend class Collator;
- // 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();
+ RuleBasedCollator(const CollationCacheEntry *entry);
/**
- * 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;
+ * Enumeration of attributes that are relevant for short definition strings
+ * (e.g., ucol_getShortDefinitionString()).
+ * Effectively extends UColAttribute.
+ */
+ enum Attributes {
+ ATTR_VARIABLE_TOP = UCOL_ATTRIBUTE_COUNT,
+ ATTR_LIMIT
+ };
- /**
- * 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;
-};
+ void adoptTailoring(CollationTailoring *t, UErrorCode &errorCode);
-// inline method implementation ---------------------------------------------
+ // Both lengths must be <0 or else both must be >=0.
+ UCollationResult doCompare(const char16_t *left, int32_t leftLength,
+ const char16_t *right, int32_t rightLength,
+ UErrorCode &errorCode) const;
+ UCollationResult doCompare(const uint8_t *left, int32_t leftLength,
+ const uint8_t *right, int32_t rightLength,
+ UErrorCode &errorCode) const;
-inline void RuleBasedCollator::setUCollator(const Locale &locale,
- UErrorCode &status)
-{
- setUCollator(locale.getName(), status);
-}
+ void writeSortKey(const char16_t *s, int32_t length,
+ SortKeyByteSink &sink, UErrorCode &errorCode) const;
+ void writeIdenticalLevel(const char16_t *s, const char16_t *limit,
+ SortKeyByteSink &sink, UErrorCode &errorCode) const;
-inline void RuleBasedCollator::setUCollator(UCollator *collator)
-{
+ const CollationSettings &getDefaultSettings() const;
- if (ucollator && dataIsOwned) {
- ucol_close(ucollator);
+ void setAttributeDefault(int32_t attribute) {
+ explicitlySetAttributes &= ~((uint32_t)1 << attribute);
}
- ucollator = collator;
- dataIsOwned = FALSE;
- isWriteThroughAlias = TRUE;
- setRuleStringFromCollator();
-}
-
-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;
+ void setAttributeExplicitly(int32_t attribute) {
+ explicitlySetAttributes |= (uint32_t)1 << attribute;
}
-}
-
-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;
+ UBool attributeHasBeenSetExplicitly(int32_t attribute) const {
+ // assert(0 <= attribute < ATTR_LIMIT);
+ return (UBool)((explicitlySetAttributes & ((uint32_t)1 << attribute)) != 0);
}
-}
-
-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
+ /**
+ * Tests whether a character is "unsafe" for use as a collation starting point.
+ *
+ * @param c code point or code unit
+ * @return TRUE if c is unsafe
+ * @see CollationElementIterator#setOffset(int)
+ */
+ UBool isUnsafe(UChar32 c) const;
+
+ static void U_CALLCONV computeMaxExpansions(const CollationTailoring *t, UErrorCode &errorCode);
+ UBool initMaxExpansions(UErrorCode &errorCode) const;
+
+ void setFastLatinOptions(CollationSettings &ownedSettings) const;
-#endif /* #if !UCONFIG_NO_COLLATION */
+ const CollationData *data;
+ const CollationSettings *settings; // reference-counted
+ const CollationTailoring *tailoring; // alias of cacheEntry->tailoring
+ const CollationCacheEntry *cacheEntry; // reference-counted
+ Locale validLocale;
+ uint32_t explicitlySetAttributes;
+
+ UBool actualLocaleIsSameAsValid;
+};
+
+U_NAMESPACE_END
+#endif // U_SHOW_CPLUSPLUS_API
-#endif
+#endif // !UCONFIG_NO_COLLATION
+#endif // TBLCOLL_H
projects / apple / icu.git / blobdiff