Instances of this class are most commonly created by the factory methods of michael@0: * BreakIterator::createWordInstance(), BreakIterator::createLineInstance(), etc., michael@0: * and then used via the abstract API in class BreakIterator
michael@0: * michael@0: *See the ICU User Guide for information on Break Iterator Rules.
michael@0: * michael@0: *This class is not intended to be subclassed. (Class DictionaryBasedBreakIterator michael@0: * is a subclass, but that relationship is effectively internal to the ICU michael@0: * implementation. The subclassing interface to RulesBasedBreakIterator is michael@0: * not part of the ICU API, and may not remain stable.
michael@0: * michael@0: */ michael@0: class U_COMMON_API RuleBasedBreakIterator : public BreakIterator { michael@0: michael@0: protected: michael@0: /** michael@0: * The UText through which this BreakIterator accesses the text michael@0: * @internal michael@0: */ michael@0: UText *fText; michael@0: michael@0: /** michael@0: * A character iterator that refers to the same text as the UText, above. michael@0: * Only included for compatibility with old API, which was based on CharacterIterators. michael@0: * Value may be adopted from outside, or one of fSCharIter or fDCharIter, below. michael@0: */ michael@0: CharacterIterator *fCharIter; michael@0: michael@0: /** michael@0: * When the input text is provided by a UnicodeString, this will point to michael@0: * a characterIterator that wraps that data. Needed only for the michael@0: * implementation of getText(), a backwards compatibility issue. michael@0: */ michael@0: StringCharacterIterator *fSCharIter; michael@0: michael@0: /** michael@0: * When the input text is provided by a UText, this michael@0: * dummy CharacterIterator over an empty string will michael@0: * be returned from getText() michael@0: */ michael@0: UCharCharacterIterator *fDCharIter; michael@0: michael@0: /** michael@0: * The rule data for this BreakIterator instance michael@0: * @internal michael@0: */ michael@0: RBBIDataWrapper *fData; michael@0: michael@0: /** Index of the Rule {tag} values for the most recent match. michael@0: * @internal michael@0: */ michael@0: int32_t fLastRuleStatusIndex; michael@0: michael@0: /** michael@0: * Rule tag value valid flag. michael@0: * Some iterator operations don't intrinsically set the correct tag value. michael@0: * This flag lets us lazily compute the value if we are ever asked for it. michael@0: * @internal michael@0: */ michael@0: UBool fLastStatusIndexValid; michael@0: michael@0: /** michael@0: * Counter for the number of characters encountered with the "dictionary" michael@0: * flag set. michael@0: * @internal michael@0: */ michael@0: uint32_t fDictionaryCharCount; michael@0: michael@0: /** michael@0: * When a range of characters is divided up using the dictionary, the break michael@0: * positions that are discovered are stored here, preventing us from having michael@0: * to use either the dictionary or the state table again until the iterator michael@0: * leaves this range of text. Has the most impact for line breaking. michael@0: * @internal michael@0: */ michael@0: int32_t* fCachedBreakPositions; michael@0: michael@0: /** michael@0: * The number of elements in fCachedBreakPositions michael@0: * @internal michael@0: */ michael@0: int32_t fNumCachedBreakPositions; michael@0: michael@0: /** michael@0: * if fCachedBreakPositions is not null, this indicates which item in the michael@0: * cache the current iteration position refers to michael@0: * @internal michael@0: */ michael@0: int32_t fPositionInCache; michael@0: michael@0: /** michael@0: * michael@0: * If present, UStack of LanguageBreakEngine objects that might handle michael@0: * dictionary characters. Searched from top to bottom to find an object to michael@0: * handle a given character. michael@0: * @internal michael@0: */ michael@0: UStack *fLanguageBreakEngines; michael@0: michael@0: /** michael@0: * michael@0: * If present, the special LanguageBreakEngine used for handling michael@0: * characters that are in the dictionary set, but not handled by any michael@0: * LangugageBreakEngine. michael@0: * @internal michael@0: */ michael@0: UnhandledEngine *fUnhandledBreakEngine; michael@0: michael@0: /** michael@0: * michael@0: * The type of the break iterator, or -1 if it has not been set. michael@0: * @internal michael@0: */ michael@0: int32_t fBreakType; michael@0: michael@0: protected: michael@0: //======================================================================= michael@0: // constructors michael@0: //======================================================================= michael@0: michael@0: #ifndef U_HIDE_INTERNAL_API michael@0: /** michael@0: * Constant to be used in the constructor michael@0: * RuleBasedBreakIterator(RBBIDataHeader*, EDontAdopt, UErrorCode &); michael@0: * which does not adopt the memory indicated by the RBBIDataHeader* michael@0: * parameter. michael@0: * michael@0: * @internal michael@0: */ michael@0: enum EDontAdopt { michael@0: kDontAdopt michael@0: }; michael@0: michael@0: /** michael@0: * Constructor from a flattened set of RBBI data in malloced memory. michael@0: * RulesBasedBreakIterators built from a custom set of rules michael@0: * are created via this constructor; the rules are compiled michael@0: * into memory, then the break iterator is constructed here. michael@0: * michael@0: * The break iterator adopts the memory, and will michael@0: * free it when done. michael@0: * @internal michael@0: */ michael@0: RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode &status); michael@0: michael@0: /** michael@0: * Constructor from a flattened set of RBBI data in memory which need not michael@0: * be malloced (e.g. it may be a memory-mapped file, etc.). michael@0: * michael@0: * This version does not adopt the memory, and does not michael@0: * free it when done. michael@0: * @internal michael@0: */ michael@0: RuleBasedBreakIterator(const RBBIDataHeader* data, enum EDontAdopt dontAdopt, UErrorCode &status); michael@0: #endif /* U_HIDE_INTERNAL_API */ michael@0: michael@0: michael@0: friend class RBBIRuleBuilder; michael@0: /** @internal */ michael@0: friend class BreakIterator; michael@0: michael@0: michael@0: michael@0: public: michael@0: michael@0: /** Default constructor. Creates an empty shell of an iterator, with no michael@0: * rules or text to iterate over. Object can subsequently be assigned to. michael@0: * @stable ICU 2.2 michael@0: */ michael@0: RuleBasedBreakIterator(); michael@0: michael@0: /** michael@0: * Copy constructor. Will produce a break iterator with the same behavior, michael@0: * and which iterates over the same text, as the one passed in. michael@0: * @param that The RuleBasedBreakIterator passed to be copied michael@0: * @stable ICU 2.0 michael@0: */ michael@0: RuleBasedBreakIterator(const RuleBasedBreakIterator& that); michael@0: michael@0: /** michael@0: * Construct a RuleBasedBreakIterator from a set of rules supplied as a string. michael@0: * @param rules The break rules to be used. michael@0: * @param parseError In the event of a syntax error in the rules, provides the location michael@0: * within the rules of the problem. michael@0: * @param status Information on any errors encountered. michael@0: * @stable ICU 2.2 michael@0: */ michael@0: RuleBasedBreakIterator( const UnicodeString &rules, michael@0: UParseError &parseError, michael@0: UErrorCode &status); michael@0: michael@0: /** michael@0: * Contruct a RuleBasedBreakIterator from a set of precompiled binary rules. michael@0: * Binary rules are obtained from RulesBasedBreakIterator::getBinaryRules(). michael@0: * Construction of a break iterator in this way is substantially faster than michael@0: * constuction from source rules. michael@0: * michael@0: * Ownership of the storage containing the compiled rules remains with the michael@0: * caller of this function. The compiled rules must not be modified or michael@0: * deleted during the life of the break iterator. michael@0: * michael@0: * The compiled rules are not compatible across different major versions of ICU. michael@0: * The compiled rules are comaptible only between machines with the same michael@0: * byte ordering (little or big endian) and the same base character set family michael@0: * (ASCII or EBCDIC). michael@0: * michael@0: * @see #getBinaryRules michael@0: * @param compiledRules A pointer to the compiled break rules to be used. michael@0: * @param ruleLength The length of the compiled break rules, in bytes. This michael@0: * corresponds to the length value produced by getBinaryRules(). michael@0: * @param status Information on any errors encountered, including invalid michael@0: * binary rules. michael@0: * @stable ICU 4.8 michael@0: */ michael@0: RuleBasedBreakIterator(const uint8_t *compiledRules, michael@0: uint32_t ruleLength, michael@0: UErrorCode &status); michael@0: michael@0: /** michael@0: * This constructor uses the udata interface to create a BreakIterator michael@0: * whose internal tables live in a memory-mapped file. "image" is an michael@0: * ICU UDataMemory handle for the pre-compiled break iterator tables. michael@0: * @param image handle to the memory image for the break iterator data. michael@0: * Ownership of the UDataMemory handle passes to the Break Iterator, michael@0: * which will be responsible for closing it when it is no longer needed. michael@0: * @param status Information on any errors encountered. michael@0: * @see udata_open michael@0: * @see #getBinaryRules michael@0: * @stable ICU 2.8 michael@0: */ michael@0: RuleBasedBreakIterator(UDataMemory* image, UErrorCode &status); michael@0: michael@0: /** michael@0: * Destructor michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual ~RuleBasedBreakIterator(); michael@0: michael@0: /** michael@0: * Assignment operator. Sets this iterator to have the same behavior, michael@0: * and iterate over the same text, as the one passed in. michael@0: * @param that The RuleBasedBreakItertor passed in michael@0: * @return the newly created RuleBasedBreakIterator michael@0: * @stable ICU 2.0 michael@0: */ michael@0: RuleBasedBreakIterator& operator=(const RuleBasedBreakIterator& that); michael@0: michael@0: /** michael@0: * Equality operator. Returns TRUE if both BreakIterators are of the michael@0: * same class, have the same behavior, and iterate over the same text. michael@0: * @param that The BreakIterator to be compared for equality michael@0: * @return TRUE if both BreakIterators are of the michael@0: * same class, have the same behavior, and iterate over the same text. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual UBool operator==(const BreakIterator& that) const; michael@0: michael@0: /** michael@0: * Not-equal operator. If operator== returns TRUE, this returns FALSE, michael@0: * and vice versa. michael@0: * @param that The BreakIterator to be compared for inequality michael@0: * @return TRUE if both BreakIterators are not same. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: UBool operator!=(const BreakIterator& that) const; michael@0: michael@0: /** michael@0: * Returns a newly-constructed RuleBasedBreakIterator with the same michael@0: * behavior, and iterating over the same text, as this one. michael@0: * Differs from the copy constructor in that it is polymorphic, and michael@0: * will correctly clone (copy) a derived class. michael@0: * clone() is thread safe. Multiple threads may simultaeneously michael@0: * clone the same source break iterator. michael@0: * @return a newly-constructed RuleBasedBreakIterator michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual BreakIterator* clone() const; michael@0: michael@0: /** michael@0: * Compute a hash code for this BreakIterator michael@0: * @return A hash code michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t hashCode(void) const; michael@0: michael@0: /** michael@0: * Returns the description used to create this iterator michael@0: * @return the description used to create this iterator michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual const UnicodeString& getRules(void) const; michael@0: michael@0: //======================================================================= michael@0: // BreakIterator overrides michael@0: //======================================================================= michael@0: michael@0: /** michael@0: *michael@0: * Return a CharacterIterator over the text being analyzed. michael@0: * The returned character iterator is owned by the break iterator, and must michael@0: * not be deleted by the caller. Repeated calls to this function may michael@0: * return the same CharacterIterator. michael@0: *
michael@0: *michael@0: * The returned character iterator must not be used concurrently with michael@0: * the break iterator. If concurrent operation is needed, clone the michael@0: * returned character iterator first and operate on the clone. michael@0: *
michael@0: *michael@0: * When the break iterator is operating on text supplied via a UText, michael@0: * this function will fail. Lacking any way to signal failures, it michael@0: * returns an CharacterIterator containing no text. michael@0: * The function getUText() provides similar functionality, michael@0: * is reliable, and is more efficient. michael@0: *
michael@0: * michael@0: * TODO: deprecate this function? michael@0: * michael@0: * @return An iterator over the text being analyzed. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual CharacterIterator& getText(void) const; michael@0: michael@0: michael@0: /** michael@0: * Get a UText for the text being analyzed. michael@0: * The returned UText is a shallow clone of the UText used internally michael@0: * by the break iterator implementation. It can safely be used to michael@0: * access the text without impacting any break iterator operations, michael@0: * but the underlying text itself must not be altered. michael@0: * michael@0: * @param fillIn A UText to be filled in. If NULL, a new UText will be michael@0: * allocated to hold the result. michael@0: * @param status receives any error codes. michael@0: * @return The current UText for this break iterator. If an input michael@0: * UText was provided, it will always be returned. michael@0: * @stable ICU 3.4 michael@0: */ michael@0: virtual UText *getUText(UText *fillIn, UErrorCode &status) const; michael@0: michael@0: /** michael@0: * Set the iterator to analyze a new piece of text. This function resets michael@0: * the current iteration position to the beginning of the text. michael@0: * @param newText An iterator over the text to analyze. The BreakIterator michael@0: * takes ownership of the character iterator. The caller MUST NOT delete it! michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual void adoptText(CharacterIterator* newText); michael@0: michael@0: /** michael@0: * Set the iterator to analyze a new piece of text. This function resets michael@0: * the current iteration position to the beginning of the text. michael@0: * @param newText The text to analyze. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual void setText(const UnicodeString& newText); michael@0: michael@0: /** michael@0: * Reset the break iterator to operate over the text represented by michael@0: * the UText. The iterator position is reset to the start. michael@0: * michael@0: * This function makes a shallow clone of the supplied UText. This means michael@0: * that the caller is free to immediately close or otherwise reuse the michael@0: * Utext that was passed as a parameter, but that the underlying text itself michael@0: * must not be altered while being referenced by the break iterator. michael@0: * michael@0: * @param text The UText used to change the text. michael@0: * @param status Receives any error codes. michael@0: * @stable ICU 3.4 michael@0: */ michael@0: virtual void setText(UText *text, UErrorCode &status); michael@0: michael@0: /** michael@0: * Sets the current iteration position to the beginning of the text. michael@0: * @return The offset of the beginning of the text. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t first(void); michael@0: michael@0: /** michael@0: * Sets the current iteration position to the end of the text. michael@0: * @return The text's past-the-end offset. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t last(void); michael@0: michael@0: /** michael@0: * Advances the iterator either forward or backward the specified number of steps. michael@0: * Negative values move backward, and positive values move forward. This is michael@0: * equivalent to repeatedly calling next() or previous(). michael@0: * @param n The number of steps to move. The sign indicates the direction michael@0: * (negative is backwards, and positive is forwards). michael@0: * @return The character offset of the boundary position n boundaries away from michael@0: * the current one. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t next(int32_t n); michael@0: michael@0: /** michael@0: * Advances the iterator to the next boundary position. michael@0: * @return The position of the first boundary after this one. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t next(void); michael@0: michael@0: /** michael@0: * Moves the iterator backwards, to the last boundary preceding this one. michael@0: * @return The position of the last boundary position preceding this one. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t previous(void); michael@0: michael@0: /** michael@0: * Sets the iterator to refer to the first boundary position following michael@0: * the specified position. michael@0: * @param offset The position from which to begin searching for a break position. michael@0: * @return The position of the first break after the current position. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t following(int32_t offset); michael@0: michael@0: /** michael@0: * Sets the iterator to refer to the last boundary position before the michael@0: * specified position. michael@0: * @param offset The position to begin searching for a break from. michael@0: * @return The position of the last boundary before the starting position. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t preceding(int32_t offset); michael@0: michael@0: /** michael@0: * Returns true if the specfied position is a boundary position. As a side michael@0: * effect, leaves the iterator pointing to the first boundary position at michael@0: * or after "offset". michael@0: * @param offset the offset to check. michael@0: * @return True if "offset" is a boundary position. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual UBool isBoundary(int32_t offset); michael@0: michael@0: /** michael@0: * Returns the current iteration position. michael@0: * @return The current iteration position. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual int32_t current(void) const; michael@0: michael@0: michael@0: /** michael@0: * Return the status tag from the break rule that determined the most recently michael@0: * returned break position. For break rules that do not specify a michael@0: * status, a default value of 0 is returned. If more than one break rule michael@0: * would cause a boundary to be located at some position in the text, michael@0: * the numerically largest of the applicable status values is returned. michael@0: *michael@0: * Of the standard types of ICU break iterators, only word break and michael@0: * line break provide status values. The values are defined in michael@0: * the header file ubrk.h. For Word breaks, the status allows distinguishing between words michael@0: * that contain alphabetic letters, "words" that appear to be numbers, michael@0: * punctuation and spaces, words containing ideographic characters, and michael@0: * more. For Line Break, the status distinguishes between hard (mandatory) breaks michael@0: * and soft (potential) break positions. michael@0: *
michael@0: * getRuleStatus() can be called after obtaining a boundary
michael@0: * position from next(), previous(), or
michael@0: * any other break iterator functions that returns a boundary position.
michael@0: *
michael@0: * When creating custom break rules, one is free to define whatever michael@0: * status values may be convenient for the application. michael@0: *
michael@0: * Note: this function is not thread safe. It should not have been michael@0: * declared const, and the const remains only for compatibility michael@0: * reasons. (The function is logically const, but not bit-wise const). michael@0: *
michael@0: * @return the status from the break rule that determined the most recently michael@0: * returned break position. michael@0: * michael@0: * @see UWordBreak michael@0: * @stable ICU 2.2 michael@0: */ michael@0: virtual int32_t getRuleStatus() const; michael@0: michael@0: /** michael@0: * Get the status (tag) values from the break rule(s) that determined the most michael@0: * recently returned break position. michael@0: *
michael@0: * The returned status value(s) are stored into an array provided by the caller. michael@0: * The values are stored in sorted (ascending) order. michael@0: * If the capacity of the output array is insufficient to hold the data, michael@0: * the output will be truncated to the available length, and a michael@0: * U_BUFFER_OVERFLOW_ERROR will be signaled. michael@0: * michael@0: * @param fillInVec an array to be filled in with the status values. michael@0: * @param capacity the length of the supplied vector. A length of zero causes michael@0: * the function to return the number of status values, in the michael@0: * normal way, without attemtping to store any values. michael@0: * @param status receives error codes. michael@0: * @return The number of rule status values from rules that determined michael@0: * the most recent boundary returned by the break iterator. michael@0: * In the event of a U_BUFFER_OVERFLOW_ERROR, the return value michael@0: * is the total number of status values that were available, michael@0: * not the reduced number that were actually returned. michael@0: * @see getRuleStatus michael@0: * @stable ICU 3.0 michael@0: */ michael@0: virtual int32_t getRuleStatusVec(int32_t *fillInVec, int32_t capacity, UErrorCode &status); michael@0: michael@0: /** michael@0: * Returns a unique class ID POLYMORPHICALLY. Pure virtual override. michael@0: * This method is to implement a simple version of RTTI, since not all michael@0: * C++ compilers support genuine RTTI. Polymorphic operator==() and michael@0: * clone() methods call this method. michael@0: * michael@0: * @return The class ID for this object. All objects of a michael@0: * given class have the same class ID. Objects of michael@0: * other classes have different class IDs. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: virtual UClassID getDynamicClassID(void) const; michael@0: michael@0: /** michael@0: * Returns the class ID for this class. This is useful only for michael@0: * comparing to a return value from getDynamicClassID(). For example: michael@0: * michael@0: * Base* polymorphic_pointer = createPolymorphicObject(); michael@0: * if (polymorphic_pointer->getDynamicClassID() == michael@0: * Derived::getStaticClassID()) ... michael@0: * michael@0: * @return The class ID for all objects of this class. michael@0: * @stable ICU 2.0 michael@0: */ michael@0: static UClassID U_EXPORT2 getStaticClassID(void); michael@0: michael@0: /** michael@0: * Deprecated functionality. Use clone() instead. michael@0: * michael@0: * Create a clone (copy) of this break iterator in memory provided michael@0: * by the caller. The idea is to increase performance by avoiding michael@0: * a storage allocation. Use of this functoin is NOT RECOMMENDED. michael@0: * Performance gains are minimal, and correct buffer management is michael@0: * tricky. Use clone() instead. michael@0: * michael@0: * @param stackBuffer The pointer to the memory into which the cloned object michael@0: * should be placed. If NULL, allocate heap memory michael@0: * for the cloned object. michael@0: * @param BufferSize The size of the buffer. If zero, return the required michael@0: * buffer size, but do not clone the object. If the michael@0: * size was too small (but not zero), allocate heap michael@0: * storage for the cloned object. michael@0: * michael@0: * @param status Error status. U_SAFECLONE_ALLOCATED_WARNING will be michael@0: * returned if the the provided buffer was too small, and michael@0: * the clone was therefore put on the heap. michael@0: * michael@0: * @return Pointer to the clone object. This may differ from the stackBuffer michael@0: * address if the byte alignment of the stack buffer was not suitable michael@0: * or if the stackBuffer was too small to hold the clone. michael@0: * @deprecated ICU 52. Use clone() instead. michael@0: */ michael@0: virtual BreakIterator * createBufferClone(void *stackBuffer, michael@0: int32_t &BufferSize, michael@0: UErrorCode &status); michael@0: michael@0: michael@0: /** michael@0: * Return the binary form of compiled break rules, michael@0: * which can then be used to create a new break iterator at some michael@0: * time in the future. Creating a break iterator from pre-compiled rules michael@0: * is much faster than building one from the source form of the michael@0: * break rules. michael@0: * michael@0: * The binary data can only be used with the same version of ICU michael@0: * and on the same platform type (processor endian-ness) michael@0: * michael@0: * @param length Returns the length of the binary data. (Out paramter.) michael@0: * michael@0: * @return A pointer to the binary (compiled) rule data. The storage michael@0: * belongs to the RulesBasedBreakIterator object, not the michael@0: * caller, and must not be modified or deleted. michael@0: * @stable ICU 4.8 michael@0: */ michael@0: virtual const uint8_t *getBinaryRules(uint32_t &length); michael@0: michael@0: /** michael@0: * Set the subject text string upon which the break iterator is operating michael@0: * without changing any other aspect of the matching state. michael@0: * The new and previous text strings must have the same content. michael@0: * michael@0: * This function is intended for use in environments where ICU is operating on michael@0: * strings that may move around in memory. It provides a mechanism for notifying michael@0: * ICU that the string has been relocated, and providing a new UText to access the michael@0: * string in its new position. michael@0: * michael@0: * Note that the break iterator implementation never copies the underlying text michael@0: * of a string being processed, but always operates directly on the original text michael@0: * provided by the user. Refreshing simply drops the references to the old text michael@0: * and replaces them with references to the new. michael@0: * michael@0: * Caution: this function is normally used only by very specialized, michael@0: * system-level code. One example use case is with garbage collection that moves michael@0: * the text in memory. michael@0: * michael@0: * @param input The new (moved) text string. michael@0: * @param status Receives errors detected by this function. michael@0: * @return *this michael@0: * michael@0: * @stable ICU 49 michael@0: */ michael@0: virtual RuleBasedBreakIterator &refreshInputText(UText *input, UErrorCode &status); michael@0: michael@0: michael@0: protected: michael@0: //======================================================================= michael@0: // implementation michael@0: //======================================================================= michael@0: /** michael@0: * Dumps caches and performs other actions associated with a complete change michael@0: * in text or iteration position. michael@0: * @internal michael@0: */ michael@0: virtual void reset(void); michael@0: michael@0: #if 0 michael@0: /** michael@0: * Return true if the category lookup for this char michael@0: * indicates that it is in the set of dictionary lookup chars. michael@0: * This function is intended for use by dictionary based break iterators. michael@0: * @return true if the category lookup for this char michael@0: * indicates that it is in the set of dictionary lookup chars. michael@0: * @internal michael@0: */ michael@0: virtual UBool isDictionaryChar(UChar32); michael@0: michael@0: /** michael@0: * Get the type of the break iterator. michael@0: * @internal michael@0: */ michael@0: virtual int32_t getBreakType() const; michael@0: #endif michael@0: michael@0: /** michael@0: * Set the type of the break iterator. michael@0: * @internal michael@0: */ michael@0: virtual void setBreakType(int32_t type); michael@0: michael@0: #ifndef U_HIDE_INTERNAL_API michael@0: /** michael@0: * Common initialization function, used by constructors and bufferClone. michael@0: * @internal michael@0: */ michael@0: void init(); michael@0: #endif /* U_HIDE_INTERNAL_API */ michael@0: michael@0: private: michael@0: michael@0: /** michael@0: * This method backs the iterator back up to a "safe position" in the text. michael@0: * This is a position that we know, without any context, must be a break position. michael@0: * The various calling methods then iterate forward from this safe position to michael@0: * the appropriate position to return. (For more information, see the description michael@0: * of buildBackwardsStateTable() in RuleBasedBreakIterator.Builder.) michael@0: * @param statetable state table used of moving backwards michael@0: * @internal michael@0: */ michael@0: int32_t handlePrevious(const RBBIStateTable *statetable); michael@0: michael@0: /** michael@0: * This method is the actual implementation of the next() method. All iteration michael@0: * vectors through here. This method initializes the state machine to state 1 michael@0: * and advances through the text character by character until we reach the end michael@0: * of the text or the state machine transitions to state 0. We update our return michael@0: * value every time the state machine passes through a possible end state. michael@0: * @param statetable state table used of moving forwards michael@0: * @internal michael@0: */ michael@0: int32_t handleNext(const RBBIStateTable *statetable); michael@0: michael@0: protected: michael@0: michael@0: #ifndef U_HIDE_INTERNAL_API michael@0: /** michael@0: * This is the function that actually implements dictionary-based michael@0: * breaking. Covering at least the range from startPos to endPos, michael@0: * it checks for dictionary characters, and if it finds them determines michael@0: * the appropriate object to deal with them. It may cache found breaks in michael@0: * fCachedBreakPositions as it goes. It may well also look at text outside michael@0: * the range startPos to endPos. michael@0: * If going forward, endPos is the normal Unicode break result, and michael@0: * if goind in reverse, startPos is the normal Unicode break result michael@0: * @param startPos The start position of a range of text michael@0: * @param endPos The end position of a range of text michael@0: * @param reverse The call is for the reverse direction michael@0: * @internal michael@0: */ michael@0: int32_t checkDictionary(int32_t startPos, int32_t endPos, UBool reverse); michael@0: #endif /* U_HIDE_INTERNAL_API */ michael@0: michael@0: private: michael@0: michael@0: /** michael@0: * This function returns the appropriate LanguageBreakEngine for a michael@0: * given character c. michael@0: * @param c A character in the dictionary set michael@0: * @internal michael@0: */ michael@0: const LanguageBreakEngine *getLanguageBreakEngine(UChar32 c); michael@0: michael@0: /** michael@0: * @internal michael@0: */ michael@0: void makeRuleStatusValid(); michael@0: michael@0: }; michael@0: michael@0: //------------------------------------------------------------------------------ michael@0: // michael@0: // Inline Functions Definitions ... michael@0: // michael@0: //------------------------------------------------------------------------------ michael@0: michael@0: inline UBool RuleBasedBreakIterator::operator!=(const BreakIterator& that) const { michael@0: return !operator==(that); michael@0: } michael@0: michael@0: U_NAMESPACE_END michael@0: michael@0: #endif /* #if !UCONFIG_NO_BREAK_ITERATION */ michael@0: michael@0: #endif